Browse the corpus

Spina bifida (SB), a congenital neural tube defect, causes extensive health problems including hydrocephalus, Chiari II malformation, impaired sensation, muscle weakness, and paralysis, orthopedic problems such as hip and knee contractures, neurogenic bladder and bowel dysfunction, seizure disorders, and neuropsychological difficulties limiting self-management (Kelly, Zebracki, Holmbeck, & Gershenson, 2008; Mitchell et al., 2004; Tarazi, Zabel, & Mahone, 2008; Verhoef, Bark, van Asbeck, Gooskens, & Prevo, 2004). Although biomedical advances have dramatically increased life expectancy for individuals with SB (Bowman, McLone, Grant, Tomita, & Ito, 2001; Davis et al., 2005), less attention has been placed on supporting psychosocial functioning as these and other young adults with a chronic health condition (CHC) assume the roles and responsibilities of adulthood (Arnett, 1998, 2004; Betz, 2004; Betz & Redcay, 2005; Kinavey, 2007; Liptak, 2003; Reiss & Gibson, 2002; Tarazi, Mahone, & Zabel, 2007). Discrepancies in education, independent living, employment, and autonomy are regularly noted between young adults with a CHC and their peers (Blackorby & Wagner, 1996; Davis, Shurtleff, Walter, Seidel, & Duguay, 2006; Geenen, Powers, & Sells, 2003; Stam, Hartman, Deurloo, Groothoff, & Grootenhuis, 2006).

Zabel, 2007). Discrepancies in education, independent living, employment, and autonomy are regularly noted between young adults with a CHC and their peers (Blackorby & Wagner, 1996; Davis, Shurtleff, Walter, Seidel, & Duguay, 2006; Geenen, Powers, & Sells, 2003; Stam, Hartman, Deurloo, Groothoff, & Grootenhuis, 2006). These gaps in achieving the functional expectations of early adulthood may predispose young adults with a chronic condition like SB to poor psychological functioning (Arnett, 1998, 2004; McDonnell & McCann, 2000; Taleporos & McCabe, 2005; Zashikhina & Hagglof, 2007). Although an elevated risk for depressive symptoms has been documented for individuals with SB in the adolescent period (Appleton et al., 1997; Holmbeck et al., 2009), the prevalence of psychological symptoms in young adults with SB is an understudied area, and knowledge of factors associated with poor psychological functioning is particularly limited (Liptak, 2003). Since mood disorders have previously been shown to further restrict the ability of vulnerable populations to achieve self-management of their health condition (Gadalla, 2008), detecting and understanding factors associated with depression and anxiety in young adults with SB is highly significant. Based on social ecology theory (Bronfenbrenner, 1979, 2004), the intent of this multi-center study was to advance knowledge of risk and protective correlates of psychological symptoms in young adults living with SB.

erstanding factors associated with depression and anxiety in young adults with SB is highly significant. Based on social ecology theory (Bronfenbrenner, 1979, 2004), the intent of this multi-center study was to advance knowledge of risk and protective correlates of psychological symptoms in young adults living with SB. Social ecological theory proposes that human behavior is shaped by both individual and contextual factors (Bronfenbrenner, 1979, 2004; Fraser, 2004). This model further suggests that risk factors heightening vulnerability to poor psychological functioning and protective factors that mitigate the effects of adverse experiences on developing youths are embedded in individual, family, and community social systems (Fraser, Kirby & Smokowski, 2004). A strength-of-association model related to the relative impact of the ecological factors is also described, whereby risk and protective factors more proximal to youths (e.g., individual, family) are considered to exert greater influences on developmental trajectories than distal environmental factors (Friedman, Holmbeck, Jandasek, Zukerman, & Abad, 2004). Social ecology theory has long been proposed for use in understanding adaptation to childhood chronic illness and disability (Kazak, 1989, 1992). More recently, Holmbeck and colleagues expanded the social ecological theoretical framework to account for the impact of SB clinical factors on the developmental outcomes of affected youths (Holmbeck & Shapera, 1999; Kelly, et al., 2008).

nding adaptation to childhood chronic illness and disability (Kazak, 1989, 1992). More recently, Holmbeck and colleagues expanded the social ecological theoretical framework to account for the impact of SB clinical factors on the developmental outcomes of affected youths (Holmbeck & Shapera, 1999; Kelly, et al., 2008). The utility of a social–ecological framework for understanding psychological symptoms in young adults is supported by prior studies of adjustment in school-age youths and adolescents with SB. Modest support for relationships among SB clinical factors and psychological functioning has been reported. SB severity has been identified as a risk factor for low self-esteem (Sawin, Buran, Brei, & Fastenau, 2003), poor social competence (Hommeyer, Holmbeck, Wills, & Coers, 1999), and restricted quality of life (Cate, Kennedy, & Stevenson, 2002) in youths with SB. The prevalence and experience of pain in individuals with SB has also gained attention in recent years (Roebroeck, Jahnsen, Carona, Kent, & Chamberlain, 2009). Oddson, Clancy, and McGrath (2006), for example, observed a direct correlation between the experience of pain and depressive symptoms in a sample of 68 school-age youths with SB.

experience of pain in individuals with SB has also gained attention in recent years (Roebroeck, Jahnsen, Carona, Kent, & Chamberlain, 2009). Oddson, Clancy, and McGrath (2006), for example, observed a direct correlation between the experience of pain and depressive symptoms in a sample of 68 school-age youths with SB. Other proximal individual and family factors have also been shown to influence psychological functioning in youths with SB. Sawin et al. (2003) observed a protective influence of a positive attitude toward SB on self-esteem and interpersonal competence in a sample of 60 adolescents with SB. Their program of research also found a significant relationship between adolescent attitude toward SB and health-related quality of life (Sawin, Brei, Buran, & Fastenau, 2002). Family factors have likewise emerged as important correlates of psychological functioning in youths with SB. One of the earliest studies of the interrelationships among family functioning variables and child outcomes identified family conflict as a risk factor for depression and anxiety in adolescents with SB (Murch & Cohen, 1989). More recent research has highlighted relationships between over-protective parenting styles and depressive symptoms in preadolescents with SB (Holmbeck et al., 2002). However, Sawin and colleagues (2003) identified a protective influence of family cohesion and adolescent satisfaction with family functioning on adolescent psychological functioning.

lighted relationships between over-protective parenting styles and depressive symptoms in preadolescents with SB (Holmbeck et al., 2002). However, Sawin and colleagues (2003) identified a protective influence of family cohesion and adolescent satisfaction with family functioning on adolescent psychological functioning. At the more distal healthcare system level, the Chronic Care Model (CCM) is recognized as an important clinical framework to enhance patient care and health outcomes (Wagner, et al., 2001). It advances a patient-centered approach to service delivery for individuals with a CHC as reflected by optimizing the organization of health care, clinical information systems, delivery system design, decision support, self-management support, and linkages to community resources (Glasgow, Wagner, et al., 2005). Research with other CHC populations, including diabetes, heart disease, and asthma, has revealed important associations between the receipt of care services based on CCM principles and adaptive health outcomes (Glasgow, Wagner, et al., 2005; Glasgow, Whitesides, Nelson, & King, 2005; Schmittdiel, et al., 2008).

2005). Research with other CHC populations, including diabetes, heart disease, and asthma, has revealed important associations between the receipt of care services based on CCM principles and adaptive health outcomes (Glasgow, Wagner, et al., 2005; Glasgow, Whitesides, Nelson, & King, 2005; Schmittdiel, et al., 2008). In summary, the psychological functioning of individuals with SB appears to be impacted by diverse ecological factors. However, previous investigations failed to account for the influence of the healthcare system, namely the nature of care delivery, on psychological outcomes. Furthermore, prior studies generally included samples comprised of school-age youths and adolescents living with SB. Less is known about risk and protective correlates of psychological symptoms in young adults with SB. The current study advances our understanding of this population by testing the following hypotheses based on past literature and social ecology theory. We proposed that the combined effects of select SB clinical (SB severity and pain), individual (attitude toward SB), family (satisfaction with family functioning), and healthcare system (CCM services) factors would explain variability in depressive and anxiety symptoms. Additionally, following the work of Friedman and colleagues (2004), a strength-of-association model was explored, whereby it was expected that the proximal individual (attitude toward SB) and family (satisfaction with family functioning) ecological factors would be more strongly related to psychological symptoms than the distal healthcare system factor (CCM services).

olleagues (2004), a strength-of-association model was explored, whereby it was expected that the proximal individual (attitude toward SB) and family (satisfaction with family functioning) ecological factors would be more strongly related to psychological symptoms than the distal healthcare system factor (CCM services). Methods Participants Participants were part of a larger longitudinal study examining the trajectory of health outcomes and psychosocial adaptation (psychological functioning, self-management, bowel and bladder continence, and quality of life) in young adults with SB (Bellin, 2008). The current study presents the first wave of data (Time 1) collected on psychological functioning. Sixty-one young adults with SB were recruited from five geographically diverse SB clinic sites. Three clinics served individuals with SB from birth through adulthood, while two sites only provided clinical care to an adult population (18 and older).1 Study eligibility criteria included the following: (a) primary diagnosis of SB; (b) 18–25 years of age; (c) residence in catchment areas of participating sites; and (d) capacity to understand study instruments. The selected age range was informed by current theory on emerging adulthood (Arnett, 2004) and is consistent with prior research on the assumption of the roles and responsibilities of young adulthood for individuals with SB (Davis et al., 2006).

of participating sites; and (d) capacity to understand study instruments. The selected age range was informed by current theory on emerging adulthood (Arnett, 2004) and is consistent with prior research on the assumption of the roles and responsibilities of young adulthood for individuals with SB (Davis et al., 2006). Since individuals affected by SB may present with a range of neurocognitive deficits, from mild executive functioning difficulties to profound intellectual impairments (Rose & Holmbeck, 2007), all eligible participants were screened by study staff for capacity to provide informed consent. An adapted version of the MacArthur Competence Assessment Tool was administered to measure a subject's understanding of the purpose of the project (e.g., What is the purpose of the research), activities involved in study participation (e.g., How many study visits are you asked to participate in), benefits of participation (e.g., In what way might you benefit by volunteering to participate in this study), risks and discomforts associated with participation (e.g., Tell me about the possible risks associated with participating in this project), and procedure to withdraw from the study (e.g., What will you do if you decide that you no longer want to participate in this study) (Appelbaum & Grisso, 2001). Responses to the five domains of questions were scored on a 0–2 range (0 = inadequate understanding; 1 = partial understanding; 2 = adequate understanding). To be enrolled in the study, participants must have received a total score of 8 or higher, out of a possible score of 10, on the measure.

ppelbaum & Grisso, 2001). Responses to the five domains of questions were scored on a 0–2 range (0 = inadequate understanding; 1 = partial understanding; 2 = adequate understanding). To be enrolled in the study, participants must have received a total score of 8 or higher, out of a possible score of 10, on the measure. Of the 168 eligible individuals with SB between the ages of 18 and 25 years who received medical services at the participating sites, 64 (38%) agreed to participate. Three individuals failed the competence screening, resulting in a final sample of 61 young adults with SB. Participants reported a mean age of 21.05 years (SD = 2.11), range 18–25 years. A majority was female (n = 37, 60.7%) and Caucasian (n = 47, 77.0%). Over two-thirds of the young adults with SB had hydrocephalus requiring shunt placement (n = 42, 68.9%). The average number of surgical revisions to the shunt was 2.95 (SD = 2.68). The vast majority of participants had a primary diagnosis of myelomeningocele, the most severe form of SB2 (n = 51, 81.6%). A lumbar level of lesion (LOL) was most frequently reported in the medical chart (n = 34, 55.7%), followed by a sacral LOL (n = 13, 31.1%), and thoracic LOL (n = 8, 13.1%).

t was 2.95 (SD = 2.68). The vast majority of participants had a primary diagnosis of myelomeningocele, the most severe form of SB2 (n = 51, 81.6%). A lumbar level of lesion (LOL) was most frequently reported in the medical chart (n = 34, 55.7%), followed by a sacral LOL (n = 13, 31.1%), and thoracic LOL (n = 8, 13.1%). Procedure The study was reviewed and approved by the Institutional Review Boards associated with the participating SB clinic sites and by the Professional Advisory Council of the Spina Bifida Association. Participants were recruited through mailed letter of invitation and by face-to-face contact during routine SB clinic visits. Once informed consent was obtained, participants completed a self-report questionnaire comprised of demographic (e.g., living and employment status) and health-related questions (e.g., pain) followed by the standardized instruments described below. Participants received a $35.00 gift-card as an acknowledgement of their time. Research staff performed a chart review to obtain SB clinical data. A copy of all de-identified study materials was sent to the project Principal Investigator for data management and analysis.

by the standardized instruments described below. Participants received a $35.00 gift-card as an acknowledgement of their time. Research staff performed a chart review to obtain SB clinical data. A copy of all de-identified study materials was sent to the project Principal Investigator for data management and analysis. Measures SB Clinical Factors: Spina Bifida Severity and Pain Based on the work of Hommeyer et al. (1999), a SB severity composite was formed from the following variables: (a) shunt status (1 = no, yes = 2); (b) myelomeningocele (1 = no, yes = 2); (c) lesion level (sacral = 1, lumbar = 2, thoracic = 3); and, (d) ambulation status (no assistance = 1, needs assistive devices to walk = 2, wheelchair use = 3). Scores range from 4 to 10, with higher levels reflecting greater severity. The validity of the severity composite was previously established by Hommeyer et al. (1999) who observed a significant association with health professionals’ rating of SB severity (r =.60, p <.001). Internal consistency of the composite in this sample (α =.68) is comparable to that reported by Hommeyer and colleagues (α =.70). As an index of pain, participants rated their worst pain in the last week using a 10cm horizontal visual analogue scale (1 = no pain to 10 = extreme amount of pain). Previous research on pain in individuals with SB has found the worst pain in the last week, but not the current level of pain, to correlate with depressive symptoms (r =.51, p <.01) (Oddson et al., 2006).

y factor on step three, and the healthcare system factor on step four. The total variance accounted for by the ecological factors and the change in explained variance associated with each step of the model were examined. Confidence intervals around R2 were constructed based on the guidelines outlined by Dattalo (2008). An a priori power analysis indicated that a sample size of N = 58 was required for the proposed analysis based on the following parameters: (a) α =.05; (b) β =.20; (c) five predictors in the model; and (d) a medium to large effect size of f2 =.25 (Dattalo, 2008; Faul, Erdfelder, Lang, & Buchner, 2007). Results The young adults with SB generally reported restricted experiences with employment and independent living. The majority were unemployed (n = 37, 60.7%) or employed in part-time, low wage positions (e.g., cashier, food services provider) (n = 14, 23.0%). They primarily resided at home with a parent/caregiver (n = 41, 68.3%) or in a supervised environment such as an assisted living setting (n = 3, 5.0%). A sub-set of the young adults lived alone (n = 8, 13.3%), with a spouse/partner (n = 3, 4.9%), or with a roommate (n = 3, 4.9%).

t pain in the last week using a 10cm horizontal visual analogue scale (1 = no pain to 10 = extreme amount of pain). Previous research on pain in individuals with SB has found the worst pain in the last week, but not the current level of pain, to correlate with depressive symptoms (r =.51, p <.01) (Oddson et al., 2006). Individual-Level Factor: Attitude Toward Spina Bifida The 13-item Child Attitude Toward Illness Scale was developed by Austin and Huberty (1993) to capture feelings and attitudes about a health condition from the perspective of the affected individual (e.g., “How often do you feel different from others because you have spina bifida; How often do you feel sad about having spina bifida”). Higher participant scores reflected a more positive attitude toward SB. Construct validity of the measure was supported by significant relationships with self-esteem in adolescents with SB (r =.62, p <.05) (Sawin et al., 2003) and depression in adolescents with epilepsy (r = −.55, p <.01) (Dunn, Austin, & Huster, 1999). Following review of the scale by SB expert clinicians, the item “How often do you feel spina bifida is your fault” was dropped. Since SB is a congenital birth defect, as opposed to other chronic conditions that may develop across the lifespan, this item was considered to be conceptually irrelevant to the SB population. The internal consistency of the 12-item scale administered in this study (α =.86) was comparable to what is reported for full scale (α =.89) (Heimlich, Westbrook, Austin, Cramer, & Devinsky, 2000).

ditions that may develop across the lifespan, this item was considered to be conceptually irrelevant to the SB population. The internal consistency of the 12-item scale administered in this study (α =.86) was comparable to what is reported for full scale (α =.89) (Heimlich, Westbrook, Austin, Cramer, & Devinsky, 2000). Family-level Factor: Satisfaction with Family Functioning The Family APGAR provided an assessment of how satisfied participants were with family interaction (Smilkstein, 1978). The scale measures five dimensions of family functioning: Adaptation, Partnership, Growth, Affection, and Resolve (5 items; e.g., “I am satisfied that I can turn to my family for help when something is troubling me”). Higher scores on the Family APGAR (items range from 1 = Never to 5 = Always) reflect greater levels of family satisfaction. Moderate test-retest reliability (r =.73) and internal consistency (α =.71) have been reported (Austin & Huberty, 1989). The measure also has established reliability and validity for use with individuals who have SB (Sawin, et al., 2002, 2003).

rom 1 = Never to 5 = Always) reflect greater levels of family satisfaction. Moderate test-retest reliability (r =.73) and internal consistency (α =.71) have been reported (Austin & Huberty, 1989). The measure also has established reliability and validity for use with individuals who have SB (Sawin, et al., 2002, 2003). Healthcare System Factor: Chronic Care Model Services Participants completed the Patient Assessment of Chronic Illness Care (PACIC) to measure receipt of CCM services (20 items; e.g., “Over the past 12 months when I received care for spina bifida, I was asked for my ideas when we made a treatment plan”) (MacColl Institute for Healthcare Innovation, 2004). Participants rate the characteristics of health services on a 5-point Likert-type scale (1 = None of the time to 5 = Always), with higher scores reflecting services consistent with the principles of the CCM. The PACIC has documented reliability (α =.96) and concurrent and construct validity, and has been established for use in a range of chronic conditions (Glasgow, Wagner, et al., 2005; Glasgow, Whitesides, et al., 2005).

ime to 5 = Always), with higher scores reflecting services consistent with the principles of the CCM. The PACIC has documented reliability (α =.96) and concurrent and construct validity, and has been established for use in a range of chronic conditions (Glasgow, Wagner, et al., 2005; Glasgow, Whitesides, et al., 2005). Psychological Symptoms The Hopkins Symptom Checklist (HSCL-25) was administered as a self-report index of depressive and anxiety symptoms (Hesbacher, Rickels, Morris, Newman, & Rosenfeld, 1980). The HSCL-25 is derived from the 90-item Hopkins Symptom Checklist (SCL-90) (Derogatis, Lipman, & Covi, 1973) and includes a 15-item depressive symptoms scale and a 10-item anxiety symptoms scale. Items are scored on a Likert scale ranging from 1 (Not at all) to 4 (Extremely). A mean score of ≥1.75 is used as a cut-point for each of the scales (Winokur, Winokur, Rickels, & Cox, 1984). Relative to other screening instruments, the HSCL-25 has been found to reflect the urgency with which treatment services are needed (Sandanger et al., 1999), and has a moderate degree of sensitivity and specificity to formal psychiatric diagnostic criteria (Veijola et al., 2003). The HSCL-25 has been validated for use as a screening instrument for psychological symptoms in a range of CHC populations, and has previously been administered to adults with SB (Kalfoss & Merkens, 2006). A moderate association between the depressive and anxiety symptoms factors was observed in the current sample (r =.61, p <.001).

as been validated for use as a screening instrument for psychological symptoms in a range of CHC populations, and has previously been administered to adults with SB (Kalfoss & Merkens, 2006). A moderate association between the depressive and anxiety symptoms factors was observed in the current sample (r =.61, p <.001). Data Analysis Data were screened using SPSS 16.0 Missing Value Analysis program. Less than one percent of data were missing, and no patterns related to the nature of missing data were found. To maximize retention of cases for the analysis, values for randomly missing data dispersed throughout the observations were estimated via regression imputation. Hierarchical multiple regression analysis was performed to examine the unique contributions of the SB clinical (SB severity and pain), individual (Attitude Toward SB), family (Satisfaction with Family Functioning), and health care (CCM services) factors in explaining variance in depressive and anxiety symptoms. The SB clinical factors were entered on step one of each model. A proximal-to-distal approach was subsequently used to inform the order of forced entry of factors (Friedman et al., 2004): the individual-level factor was entered on step two, the family factor on step three, and the healthcare system factor on step four. The total variance accounted for by the ecological factors and the change in explained variance associated with each step of the model were examined. Confidence intervals around R2 were constructed based on the guidelines outlined by Dattalo (2008).

ood services provider) (n = 14, 23.0%). They primarily resided at home with a parent/caregiver (n = 41, 68.3%) or in a supervised environment such as an assisted living setting (n = 3, 5.0%). A sub-set of the young adults lived alone (n = 8, 13.3%), with a spouse/partner (n = 3, 4.9%), or with a roommate (n = 3, 4.9%). Participants averaged 1.73 (SD = 2.44, range 0–10) hospitalizations for SB related complications within the last three years and 1.05 (SD = 1.74, range 0–10) emergency room visits during the previous 12 months. Urinary tract infections and pressure ulcers were also fairly common in this group of young adults with SB. Participants experienced an average of 3.49 (SD = 5.10, range 0–24) urinary tract infections and 1.02 (SD = 1.43, range 0–5) pressure ulcers within the last 3 years.

ncy room visits during the previous 12 months. Urinary tract infections and pressure ulcers were also fairly common in this group of young adults with SB. Participants experienced an average of 3.49 (SD = 5.10, range 0–24) urinary tract infections and 1.02 (SD = 1.43, range 0–5) pressure ulcers within the last 3 years. Descriptive data on study instruments are presented in Table I. In each case, a higher score reflects higher levels of the concept being measured. In this sample of young adults with SB, family satisfaction was fairly high as indicated by a mean item score of 4.03 out of a possible score of 5 on the family functioning measure. In general, participants rated the nature of health services to be moderately consistent with the principles of the CCM, as reflected by a mean item score of 3.43 out a possible score of 5 on the PACIC. However, the self-reported feelings and attitudes about SB were less positive and slightly lower than those reported by adolescents with SB (Sawin et al., 2003). Intercorrelations among the explanatory variables revealed no evidence of multicollinearity. Simple correlations ranged from a low of r =.01, p >.05 (Attitude toward SB and CCM Services) to a high of r =.43, p =.001 (Attitude toward SB and Satisfaction with Family Functioning). Table I. Descriptive Analysis of Ecological Factors and Outcome Measures (n = 61)

variables revealed no evidence of multicollinearity. Simple correlations ranged from a low of r =.01, p >.05 (Attitude toward SB and CCM Services) to a high of r =.43, p =.001 (Attitude toward SB and Satisfaction with Family Functioning). Table I. Descriptive Analysis of Ecological Factors and Outcome Measures (n = 61) M SD Scale range α No. of items SB severity 7.64 1.77 4–10 .68 4 Pain 5.11 3.21 1–10 1 Attitude toward SBa 37.15 (3.10) 8.54 (.71) 12–60 (1–5) .86 12 Family satisfactionb 20.16 (4.03) 4.36 (.87) 5–25 (1–5) .91 5 CCM servicesc 68.69 (3.43) 16.48 (.82) 20–100 (1–5) .92 20 Depressive symptomsd 25.58 (1.71) 7.78 (.52) 15–60 (1–4) .90 15 Anxiety symptomsd 15.88 (1.59) 4.48 (.46) 10–40 (1–4) .80 10 The total scale score is presented first in each cell, followed by the mean item score in parenthesis to further contextualize findings. aAttitude Toward Illness (Austin & Huberty, 1993). bFamily APGAR (Austin & Huberty, 1989). cPatient Assessment of Chronic Illness Care (MacColl Institute for Healthcare Innovation, 2004). dHopkins Symptoms Checklist (Hesbacher et al., 1980).

M SD Scale range α No. of items SB severity 7.64 1.77 4–10 .68 4 Pain 5.11 3.21 1–10 1 Attitude toward SBa 37.15 (3.10) 8.54 (.71) 12–60 (1–5) .86 12 Family satisfactionb 20.16 (4.03) 4.36 (.87) 5–25 (1–5) .91 5 CCM servicesc 68.69 (3.43) 16.48 (.82) 20–100 (1–5) .92 20 Depressive symptomsd 25.58 (1.71) 7.78 (.52) 15–60 (1–4) .90 15 Anxiety symptomsd 15.88 (1.59) 4.48 (.46) 10–40 (1–4) .80 10 The total scale score is presented first in each cell, followed by the mean item score in parenthesis to further contextualize findings. aAttitude Toward Illness (Austin & Huberty, 1993). bFamily APGAR (Austin & Huberty, 1989). cPatient Assessment of Chronic Illness Care (MacColl Institute for Healthcare Innovation, 2004). dHopkins Symptoms Checklist (Hesbacher et al., 1980). With regard to the psychological functioning variables, nearly half of the young adults with SB reported psychological symptoms above the clinical cut-off (n = 30, 49.2%). In total, twenty-five individuals (41.0%) fell in the clinical range for depressive symptoms and nineteen (31.1%) reported scores above the clinical cut-off for anxiety symptoms. Of the 30 participants who were above the cut-off for psychological symptoms, 16 (53.3%) had scores above the cut point for both depressive and anxiety symptoms, ten (33.3%) had scores in the clinical range for depressive symptoms only, and four (13.3%) had scores in the clinical range for anxiety symptoms only. Following the study protocol, participants who scored in the clinical range were referred to local mental health services.

int for both depressive and anxiety symptoms, ten (33.3%) had scores in the clinical range for depressive symptoms only, and four (13.3%) had scores in the clinical range for anxiety symptoms only. Following the study protocol, participants who scored in the clinical range were referred to local mental health services. Although young women with SB may be an especially vulnerable group (Appleton et al., 1997; Sawin et al., 2009; Holmbeck et al., 2009), female gender was not associated with an increased risk for clinical levels of depressive symptoms, χ2(1, N = 61) =.39, p >.05, or anxiety symptoms, χ2(1, N = 61) =.07, p >.05. Employment status (employed versus not employed) likewise did not differentiate individuals above the clinical cut-point for depressive symptoms, χ2(1, N = 61) =.01, p >.05, or anxiety symptoms, χ2(1, N = 61) =.07, p >.05. Relationships between living status (supervised living environment vs independent living) and clinical levels of depressive symptoms, χ2(1, N = 61) =.04, p >.05, and anxiety symptoms, χ2(1, N = 61) = 1.47, p >.05, were also nonsignificant.

ms, χ2(1, N = 61) =.01, p >.05, or anxiety symptoms, χ2(1, N = 61) =.07, p >.05. Relationships between living status (supervised living environment vs independent living) and clinical levels of depressive symptoms, χ2(1, N = 61) =.04, p >.05, and anxiety symptoms, χ2(1, N = 61) = 1.47, p >.05, were also nonsignificant. Depressive Symptoms Model As reported in Table II, the overall model inclusive of the SB clinical, individual, family, and healthcare system factors explained a significant amount of variance in depressive symptoms [Adjusted R2 =.35, 95% CI =.18 to 0.53, F(5, 60) = 7.52, p <.001]. Based on the benchmarks established by Cohen (1988) for f2, where f2 of 0.02 = small, 0.15 = medium, and 0.35 = large, a large effect size was noted for the depressive symptoms model (f2 =.54). The SB clinical factors (severity, pain) accounted for a small but significant percentage of variance in depressive symptoms. As predicted, the addition of the proximal individual level factor (Attitude toward SB) to the model on step 2 [R2Δ =.22, F(1, 57) = 18.54, p <.001; f2 =.28] and family factor (Satisfaction with Family Functioning) on step 3 [R2Δ =.10, F(1, 56) = 9.04, p =.004; f2 =.11] were supported. However, the distal healthcare system factor (CCM services) was non-significant [R2Δ =.00, F(1, 55) = 0.00, p >.05]. In the final model, a main effect was observed for attitude toward SB (β = −.33, p =.006), satisfaction with family functioning (β = −.34, p =.005), and the experience of pain (β =.29, p =.008). Specifically, a more positive attitude toward SB and greater satisfaction with family functioning were associated with fewer depressive symptoms. However, pain was a risk factor for depressive symptoms in the young adults living with SB. Table II. Hierarchical Multiple Regression Results for Depressive Symptoms Model (n = 61)

cally, a more positive attitude toward SB and greater satisfaction with family functioning were associated with fewer depressive symptoms. However, pain was a risk factor for depressive symptoms in the young adults living with SB. Table II. Hierarchical Multiple Regression Results for Depressive Symptoms Model (n = 61) Predictor Total R2 Adjusted Δ R2 F Δ F df B SE β Step 1 Spina Bifida severity .05 .09 2.72 2.72* 2, 58 −.20 .46 −.05 Pain .71 .26 .29* Step 2 Attitude toward SB .27 .22 8.54** 18.54** 1, 57 −.30 .11 −.33* Step 3 Family satisfaction .36 .10 9.57** 9.04* 1, 56 −.61 .21 −.34* Step 4 CCM services .35 .00 7.52** 0.00 1, 55 .00 .05 .00 The reported unstandardized and standardized coefficients are from the final regression model. *p <.05; **p <.001

Predictor Total R2 Adjusted Δ R2 F Δ F df B SE β Step 1 Spina Bifida severity .05 .09 2.72 2.72* 2, 58 −.20 .46 −.05 Pain .71 .26 .29* Step 2 Attitude toward SB .27 .22 8.54** 18.54** 1, 57 −.30 .11 −.33* Step 3 Family satisfaction .36 .10 9.57** 9.04* 1, 56 −.61 .21 −.34* Step 4 CCM services .35 .00 7.52** 0.00 1, 55 .00 .05 .00 The reported unstandardized and standardized coefficients are from the final regression model. *p <.05; **p <.001 Anxiety Symptoms Model Less support was found for the combined effects of the ecological factors in explaining variance in anxiety symptoms (Table III). The overall model inclusive of the SB clinical, individual, family, and healthcare system factors was significant [Adjusted R2 =.26, 95% CI =.03 to 0.48, F(5, 60) = 5.11 p =.001], and was in the range of a medium-to large effect size (f2 =.33). However, the change in explained variance associated with the addition of the individual [R2Δ =.05, F(1, 57) = 3.80, p =.056; f2 =.05], family [R2Δ =.04, F(1, 56) = 3.41, p =.07; f2 =.04], and healthcare system factors [R2Δ =.01, F(1, 55) = 0.44, p =.51] to the model were all nonsignificant. In the final model, a main effect was only observed for pain (β =.46, p <.001), with pain level positively associated with anxiety symptoms. Table III. Hierarchical Multiple Regression Results for Anxiety Symptoms Model (n = 61)

re system factors [R2Δ =.01, F(1, 55) = 0.44, p =.51] to the model were all nonsignificant. In the final model, a main effect was only observed for pain (β =.46, p <.001), with pain level positively associated with anxiety symptoms. Table III. Hierarchical Multiple Regression Results for Anxiety Symptoms Model (n = 61) Predictor Total R2 Adjusted Δ R2 F Δ F df B SE β Step 1 Spina Bifida severity .20 .22 8.24** 8.24** 2, 58 −.28 .29 −.11 Pain .66 .16 .46** Step 2 Attitude toward SB .23 .05 7.03** 3.80 1, 57 −.06 .907 −.12 Step 3 Family satisfaction .26 .04 6.35** 3.41 1, 56 −.25 .13 −.24 Step 4 CCM services .26 .01 5.11* 0.44 1, 55 .02 .03 .08 The reported unstandardized and standardized coefficients are from the final regression model. *p <.05; **p <.001 Discussion With increased numbers of individuals with SB surviving into adulthood, it is paramount to address and support both their physical care needs and psychosocial health. This study investigated multi-level risk and protective correlates of psychological symptoms in young adults living with SB. Specifically, it was hypothesized that the combined effects of select SB clinical (SB severity, pain), individual (attitude toward SB), family (satisfaction with family functioning), and healthcare system (CCM services) factors would explain variability in psychological symptoms. Furthermore, a strength-of-association model was tested, whereby it was hypothesized that the more proximal ecological factors (individual, family) would be more strongly related to psychological symptoms than the distal healthcare system factor.

CM services) factors would explain variability in psychological symptoms. Furthermore, a strength-of-association model was tested, whereby it was hypothesized that the more proximal ecological factors (individual, family) would be more strongly related to psychological symptoms than the distal healthcare system factor. In general, the model tested in this research was supported. The combined effects of the ecological factors accounted for a significant amount of variance in psychological symptoms. A large effect size was noted for the depressive symptoms model (f2 =.54), while the anxiety symptoms model was in the range of a medium-to-large effect size (f2 =.33) (Cohen, 1988). The magnitude of change in explained variance associated with each step of the models was more modest in nature. In the depressive symptoms model, a medium-to-large effect size (f2 =.28) was noted for the individual factor (Attitude toward SB), and a small-to-medium effect size (f2 =.11) was observed for the family factor (Satisfaction with Family Functioning). However, in the anxiety symptoms model, a small effect size was found for the change in explained variance associated with the individual (f2 =.05) and family (f2 =.04) factors. Also, consistent with the predicted direction of relationships among the ecological factors and psychological symptoms, the proximal individual and family factors had stronger associations with depressive symptoms than the distal healthcare system factor (CCM services).

ndividual (f2 =.05) and family (f2 =.04) factors. Also, consistent with the predicted direction of relationships among the ecological factors and psychological symptoms, the proximal individual and family factors had stronger associations with depressive symptoms than the distal healthcare system factor (CCM services). A notable contribution of this research is our enhanced understanding of salient risk and protective factors to address in clinical intervention with young adults living with SB. Findings lend tentative support for a protective influence of a positive attitude toward SB and satisfaction with family functioning on the experience of depressive symptoms. In some respects, the associations are not surprising, as individuals with a CHC who positively perceive proximal aspects of life functioning (e.g., health condition, family) might be expected to report less distress. These observed relationships are consistent with long-standing theory that suggests the adjustment of individuals with a CHC is influenced by how they feel about having a chronic condition, as well as how responsive the surrounding family environment is to their developmental needs (Austin & Huberty, 1993; McCubbin & Patterson, 1983). Relationships between attitude toward disability, family functioning, and psychological adaptation have been previously documented in younger populations with SB (Sawin et al., 2002, 2003). These associations merit further investigation, as they may highlight factors relevant to the prevention and treatment of psychological symptoms in individuals with SB in early adulthood. Since prior research suggests that family functioning is consistently associated with the adjustment of youths with SB (Holmbeck et al., 2002; Sawin et al., 2002, 2003), it is also important to examine whether interventions aimed at improving family interactions during childhood and adolescence influence the subsequent psychological functioning of young adults with SB.

ng is consistently associated with the adjustment of youths with SB (Holmbeck et al., 2002; Sawin et al., 2002, 2003), it is also important to examine whether interventions aimed at improving family interactions during childhood and adolescence influence the subsequent psychological functioning of young adults with SB. Additional implications relate to the associations between the SB clinical factors and psychological symptoms. Although Wallander and Varni (1995) proposed a direct association between condition severity and adjustment in their theoretical model of adaptation to chronic illness and disability, contrary to our expectations, there was no relationship between SB severity and psychological symptoms. While SB severity variables such as lesion level and shunt status have been previously linked to child adjustment (Holmbeck & Faier-Routman, 1995) and neuropsychological presentation (Dennis, Landry, Barnes, & Fletcher, 2006; Fletcher et al., 2005), there is little evidence from the current study to suggest that these severity variables would be of use in identifying young adults at risk for psychological symptoms. Given prior findings of a robust relationship between characteristics of the family environment and adjustment in individuals with SB (Holmbeck et al., 2002; Sawin et al., 2002, 2003), it is possible that family functioning mediates the relationship between SB severity and psychological symptoms.

sychological symptoms. Given prior findings of a robust relationship between characteristics of the family environment and adjustment in individuals with SB (Holmbeck et al., 2002; Sawin et al., 2002, 2003), it is possible that family functioning mediates the relationship between SB severity and psychological symptoms. However, comparable to the findings of Oddson and colleagues (2006), self-report of recently experienced pain was strongly related to depressive and anxiety symptoms. The young adults identified varied causes of pain, although headaches and back, shoulder and foot discomfort were most frequently reported. Since previous research suggests that clinically significant pain in individuals affected by SB is often under-recognized (Clancy, McGrath, & Oddson, 2005), our findings lend support to regular screening of diverse sources of pain in young adults with SB.

ack, shoulder and foot discomfort were most frequently reported. Since previous research suggests that clinically significant pain in individuals affected by SB is often under-recognized (Clancy, McGrath, & Oddson, 2005), our findings lend support to regular screening of diverse sources of pain in young adults with SB. Our findings also add to the growing body of evidence indicating high rates of psychological distress in adults with SB. The self-report of clinically significant symptoms of depression (41%) and anxiety (31%) in our participants closely matched previous symptom reports of depression (47%) and anxiety (23.5%) in a sample of slightly older adults (mean age 29.5 years) with SB (Kalfoss & Merkens, 2006). However, these estimates in young adults with SB are considerably higher than comparable self-reports of serious psychological distress in adults with disabilities reporting assistive device use (5.4%), activity limitations (11.4%), or both (16.5%) (Okoro et al., 2009). While it is unclear if the high prevalence of psychological distress in this sample occurs secondary to CNS damage, increased vulnerability to stress, or environmental influences (Kalfoss & Alve, 2003), it seems likely that the contributing factors are active in some form before at-risk individuals with SB reach young adulthood. As such, the current data support efforts to increase clinic-based education to foster positive adjustment to SB and adaptive family functioning, as well as to expand routine screening for depression and anxiety in SB clinic visits. However, it remains to be seen if general treatment models based upon evidence-based practices (e.g., cognitive behavioral therapy, functional family therapy) are adequate for young adults with SB, or if condition-specific interventions are necessary. Disability scholars have previously raised concerns about the validity of traditional models of clinical intervention for individuals with a cognitive impairment (Dykens, 2007).

ve behavioral therapy, functional family therapy) are adequate for young adults with SB, or if condition-specific interventions are necessary. Disability scholars have previously raised concerns about the validity of traditional models of clinical intervention for individuals with a cognitive impairment (Dykens, 2007). Study findings are limited by several methodological considerations. It is important to note that the observed associations between the ecological factors and psychological symptoms are restricted due to the shared methodologies (e.g., self-report questionnaires) that preclude our ability to rule out common-method variance as an explanation for the significant relationships (Kelly et al., 2008). Furthermore, the directionality of observed relationships cannot be established due to the cross-sectional nature of the data. Longitudinal data, which is presently being collected on this cohort of young adults with SB, may help clarify whether negative perceptions of SB and family life increase vulnerability to psychological symptoms or if the presence of psychological symptoms causes young adults with SB to report negatively upon proximal aspects of their life (health condition, family).

ected on this cohort of young adults with SB, may help clarify whether negative perceptions of SB and family life increase vulnerability to psychological symptoms or if the presence of psychological symptoms causes young adults with SB to report negatively upon proximal aspects of their life (health condition, family). The response rate and sample of convenience present additional methodological concerns. The relatively low response rate of 38% reflects limited participant recruitment via mailed letter of invitation. Three sites exclusively relied on face-to-face recruitment in the spina bifida clinics and successfully enrolled 25 of 33 eligible participants (75.8% response rate). The poor response to mailed study invitations is not surprising given the executive functioning deficits that may create barriers to initiation and follow-through in individuals with SB (Tarazi et al., 2007). Although participants were enrolled from five geographically diverse SB clinic sites, it is possible that the sample characteristics are not representative of the larger population of young adults living with SB, particularly those who have a severe cognitive impairment. However, the clinical presentation of SB in this sample (e.g., level of lesion, shunt status) is comparable to what is reported in other recent studies with young adults with SB (Boudos & Mukherjee, 2008; Verhoef, et al., 2006, 2007) and is consistent with available data from participating clinic sites (Dicianno, Gaines, Collins & Lee, 2009).

esentation of SB in this sample (e.g., level of lesion, shunt status) is comparable to what is reported in other recent studies with young adults with SB (Boudos & Mukherjee, 2008; Verhoef, et al., 2006, 2007) and is consistent with available data from participating clinic sites (Dicianno, Gaines, Collins & Lee, 2009). The modest sample size also limited the number of variables entering the regression analysis. A post-hoc power analysis was performed with the following parameters: (a) N = 61; (b) α =.05; (c) β =.20; (d) five predictors in the model; and (e) a medium to large effect size of f2 =.25 (Dattalo, 2008; Faul et al., 2007). Although the analysis confirmed that the study indeed had ample power to test the main effects hierarchical regression model (1 − β =.8325), a larger sample would enable meaningful testing of moderating effects of clinical factors (e.g., SB severity) and key demographics (e.g., gender) on the observed relationships between the ecological factors and psychological symptoms (Holmbeck, 1997). Exploratory regression models were run with the SB clinical factors on step 1, the centered ecological factors (Attitude Toward SB, Satisfaction with Family Functioning, Chronic Care Model services) and Gender on step 2, and the interaction terms (e.g., Attitude Toward SB × Gender) on step 3. However, the change in explained variance associated with the interaction terms was nonsignificant in both the depressive symptoms and anxiety symptoms models.

atisfaction with Family Functioning, Chronic Care Model services) and Gender on step 2, and the interaction terms (e.g., Attitude Toward SB × Gender) on step 3. However, the change in explained variance associated with the interaction terms was nonsignificant in both the depressive symptoms and anxiety symptoms models. Finally, while the combined effects of the ecological factors in explaining psychological symptoms was supported, additional variance may be accounted for by other individual and contextual factors not included in this study. Future research might explore the effect of cognitive functioning and social perception on psychological symptoms (Dicianno, et al., 2008). Despite these limitations, the unique variance accounted for by the self-report of pain, attitude toward SB, and family satisfaction provides a potential foundation for multi-factor screening of young adults with SB who are vulnerable to psychological symptoms. Identifying mechanisms that elevate risk or protect against poor psychological functioning is essential to foster positive outcomes for young adults with SB. Funding This study was supported by the Spina Bifida Association Young Investigator Award and the University of Maryland Designated Research Initiative Funds (PI Bellin). Conflicts of interest: None declared.

Finally, while the combined effects of the ecological factors in explaining psychological symptoms was supported, additional variance may be accounted for by other individual and contextual factors not included in this study. Future research might explore the effect of cognitive functioning and social perception on psychological symptoms (Dicianno, et al., 2008). Despite these limitations, the unique variance accounted for by the self-report of pain, attitude toward SB, and family satisfaction provides a potential foundation for multi-factor screening of young adults with SB who are vulnerable to psychological symptoms. Identifying mechanisms that elevate risk or protect against poor psychological functioning is essential to foster positive outcomes for young adults with SB. Funding This study was supported by the Spina Bifida Association Young Investigator Award and the University of Maryland Designated Research Initiative Funds (PI Bellin). Conflicts of interest: None declared. Acknowledgments We thank Grayson Holmbeck, Kathy Sawin, and Patrick Dattalo for their helpful reviews of this article. We also thank Karen Rice, Rebecca Solomon, and Peggy Beall for their assistance in data entry. Most importantly, we acknowledge the young adults with spina bifida who participated in this research. 1 No differences in key demographics or study measures were found, so participants were combined for the analysis. 2 No differences in model results were found when the analysis was run with the myelomeningocele group alone.

e for convergent and discriminant validity when compared to other well-established PTSS measures as well as measures of depression and general anxiety. A value of 44 or above on the full scale has been suggested as a clinical cut off suggesting a diagnosis of PTSD (Blanchard, Jones-Alexander, Buckley & Forneris, 1996). Procedure Participants were included within 2 weeks after their child's diagnosis at four pediatric oncology centers. Potential participants were approached by a nurse who provided written and oral information about participation. The same nurse obtained oral informed consent to participate and to be contacted via telephone by a research assistant. The research assistant then, via telephone, conducted the interview where the PCL-C and other instruments (not reported herein) were administered. Permission to be contacted again was obtained at the end of the interview. The procedure was approved by the ethical review board at each respective faculty of medicine.

Accurately identifying mothers of children with disability (CWD) who are at greater risk for poor psychological well-being creates an opportunity to increase the specificity of supports, and has the potential to improve outcomes for mothers and CWD. Disability is defined as a long-term motor, language, adaptive/cognitive, or personal/social impairment (McDougall & Miller, 2003). Childhood disability often imposes a social and emotional burden for children and their families (Farmer, Marien, Clark, Sherman, & Selva, 2004; Webster, Majnemer, Platt, & Shevell, 2008), including considerable costs for health and social services (Newacheck, Inkelas, & Kim, 2004). Collectively, parents of CWD are often resilient in the face of managing their child (Flaherty & Masters Glidden, 2000; Glidden & Schoolcraft, 2003; Hastings, Beck, & Hill, 2005; Scorgie & Sobsey, 2000). However, the process model of stress and coping (Lazarus & Folkman, 1984) suggests that some subgroups may be at greater risk for clinically significant psychological distress (Baker, Blacher, & Olsson, 2005; Brehaut et al., 2004; Mulvihill et al., 2005; Neely-Barnes & Marcenko, 2004; Plant & Sanders, 2007; Smith, Oliver, & Innocenti, 2001; Webster et al., 2008) and impaired coping (Grant & Whittell, 2000; Patenaude & Kupst, 2005).

eater risk for clinically significant psychological distress (Baker, Blacher, & Olsson, 2005; Brehaut et al., 2004; Mulvihill et al., 2005; Neely-Barnes & Marcenko, 2004; Plant & Sanders, 2007; Smith, Oliver, & Innocenti, 2001; Webster et al., 2008) and impaired coping (Grant & Whittell, 2000; Patenaude & Kupst, 2005). The process model of stress and coping posits that a stressor (i.e., CWD) is mediated by coping resources and cognitive appraisal of the stressor to predict adaptation (Lazarus & Folkman, 1984). Park (1998) provides a review of the model that suggests internal coping resources (i.e., parenting morale; Trute & Hiebert-Murphy, 2005) and cognitive appraisal (i.e., perceptions of the impact of the CWD on the family; Trute, Hiebert-Murphy, & Levine, 2007) predict adaptation (i.e., psychological well-being). For the purpose of this article, psychological well-being includes depression, parenting stress, family resilience, and family adjustment. Unless there is a heritable component to the child’s disability, most of the time, families of CWD are typical families with a special child (Seligman & Darling, 1997). These families have diverse adaptational profiles (Ferguson, 2002), and therefore diverse needs for additional supports and services to care for the child (Farmer et al., 2004). While it is acknowledged that CWD live in, and are cared for by families, typically mothers are the primary caregiver and are therefore the focus of this study.

e diverse adaptational profiles (Ferguson, 2002), and therefore diverse needs for additional supports and services to care for the child (Farmer et al., 2004). While it is acknowledged that CWD live in, and are cared for by families, typically mothers are the primary caregiver and are therefore the focus of this study. With current measures, it is difficult to identify early in the child’s interaction with service providers those mothers whose psychosocial well-being will enable them to mobilize their own internal coping resources to provide care for their child, and those who will require additional intensive emotional and social supports. Most often the initial assessment to access child disability services is based on an open-ended interview between the family and a service coordinator (Summers et al., 1990). These interviews may be time consuming and are generally without a standard protocol. When standardized maternal and family assessment measures are used, they tend to be long with limited immediate relevance to service planning. Some mothers find these measures inconvenient and inconsistent with their experiences and needs (Slentz & Bricker, 1992). Thus, assessments can vary greatly in the quantity and quality of information upon which to base decisions about service requirements. The end result may be a mismatch between needs and the services provided (Krauss, Wells, Gulley, & Anderson, 2001). While it is clear that standardized approaches to assessment at intake to services are needed, brief and psychometrically sound measures are not readily available. The overall aim of this study was to assess the psychometric properties of two brief measures of psychological well-being in mothers of CWD: Parenting Morale Index (PMI) and Family Impact of Childhood Disability (FICD) scale. Scores on the PMI and FICD may be able to (a) reliably identify mothers of CWD at risk for poor psychological well-being, (b) increase the specificity of psychosocial supports, (c) more effectively allocate services within an environment of limited resources, and (d) potentially improve outcomes for mothers and CWD.

e. Scores on the PMI and FICD may be able to (a) reliably identify mothers of CWD at risk for poor psychological well-being, (b) increase the specificity of psychosocial supports, (c) more effectively allocate services within an environment of limited resources, and (d) potentially improve outcomes for mothers and CWD. Trute and colleagues developed the PMI and FICD to provide health and social service professionals with brief, easy-to-score, and interpret measures of psychological well-being in mothers of CWD (Trute & Hiebert-Murphy, 2002, 2005; Trute et al., 2007). Preliminary psychometric testing in one sample (N = 103) of Canadian mothers of CWD suggested that the PMI and FICD show promise. Exploratory factor analyses suggested stable factor structures in both measures and acceptable initial reliability and validity data.

Trute & Hiebert-Murphy, 2002, 2005; Trute et al., 2007). Preliminary psychometric testing in one sample (N = 103) of Canadian mothers of CWD suggested that the PMI and FICD show promise. Exploratory factor analyses suggested stable factor structures in both measures and acceptable initial reliability and validity data. This study contributes to family assessment in pediatric psychology by reporting on the psychometric properties of two measures of maternal psychological well-being. In order to be confident in recommending these measures to assess psychological well-being in mothers of CWD at intake to services, it was critical to confirm the factor structures and better establish reliability and validity across samples that vary by geographic and sociodemographic characteristics. If the PMI and FICD show adequate psychometric strength, they hold the potential to serve as a standardized, brief, and convenient package of measures to augment clinical interview findings in the determination of maternal psychological well-being and service needs in mothers of CWD. If the PMI and FICD are effective in specifying service needs based on potential maternal outcomes, then the use of these measures may result in more efficient allocation of limited resources.

ment clinical interview findings in the determination of maternal psychological well-being and service needs in mothers of CWD. If the PMI and FICD are effective in specifying service needs based on potential maternal outcomes, then the use of these measures may result in more efficient allocation of limited resources. The purpose of this study was to (a) assess the factor structure of the PMI and FICD, (b) evaluate their internal consistency and temporal stability, (c) test the construct validity using instruments of similar and divergent concepts, (d) test the predictive validity over 1 year, and (e) examine social desirability response bias. First, we hypothesized that factor analyses would confirm a uni-dimensional structure of the PMI and a two-dimensional (Positive and Negative subscales) structure of the FICD. Second, we expected that the PMI and FICD would demonstrate acceptable internal consistency and temporal stability over 4 weeks. Third, the PMI was conceptualized as a unique measure of parenting morale, so we expected positive relationships with measures of global well-being, positive affect, and self-efficacy. We expected negative relationships with measures of family adjustment and negative affect. Fourth, the FICD was conceptualized as a unique measure of cognitive appraisal of the family consequences of having a CWD. We expected positive relationships between the FICD Positive subscale and measures of global well-being, positive affect, and self-efficacy. We expected negative relationships between the FICD Negative subscale and measures of family adjustment, global well-being, positive affect, and self-efficacy. Finally, we hypothesized that together, the PMI and FICD Positive and Negative subscales would predict maternal depressive symptoms, parenting stress, family hardiness, and family adjustment over a 1-year interval.

gative subscale and measures of family adjustment, global well-being, positive affect, and self-efficacy. Finally, we hypothesized that together, the PMI and FICD Positive and Negative subscales would predict maternal depressive symptoms, parenting stress, family hardiness, and family adjustment over a 1-year interval. Method Recruitment We recruited participants with the assistance of Family Support for Children with Disabilities (FSCD), Alberta Children and Youth Services. FSCD is a government-sponsored support program that is offered to all families of CWD or complex health conditions. Family support services are provided without fee and include a key or dedicated worker who coordinates community based health and social services for CWD and their family members. We created a sampling frame (N = 1,019) that included all families of CWD, with first entry to disability services in the previous 3–12 months. To preserve confidentiality, we used passive recruitment methods. FSCD mailed an invitation to participate in the study with a reminder to non-respondents 6-weeks later. This resulted in an estimated response rate of 29% (N = 296), which is typical for single-mode survey designs (Dillman, Smyth, & Christian, 2009). This is conservative estimate because the response rate calculation could not account for non-respondents who were ineligible (indeterminates; Allison & Yoshida, 1989).

ter. This resulted in an estimated response rate of 29% (N = 296), which is typical for single-mode survey designs (Dillman, Smyth, & Christian, 2009). This is conservative estimate because the response rate calculation could not account for non-respondents who were ineligible (indeterminates; Allison & Yoshida, 1989). Inclusion criteria were (a) over the age of 18 years, (b) English sufficient to complete a telephone interview, and (c) CWD living with the respondent. Duplicate initial contacts, unavailability for interview, insufficient English proficiency, inability to contact, and insufficient access to a telephone further reduced the eligible respondents to 286. Of those, 237 completed a telephone survey. Only mother (N = 195) survey information was used. There were two reasons for this. First, mothers constituted the largest proportion of the overall sample (195/237). Second, there are important gender differences in parental psychological response to childhood disability (Hastings et al., 2005; Trute, 1995), in mothers’ and fathers’ coping with stress (Nagy & Ungerer, 1990) and in their assessment of family needs (Bailey, Blasco, & Simeonsson, 1992). There were no statistically significant differences on maternal age, child age, child sex, or disability characteristics of the child between mother respondents who participated in the study (n = 195) and those who did not (n = 23).

er, 1990) and in their assessment of family needs (Bailey, Blasco, & Simeonsson, 1992). There were no statistically significant differences on maternal age, child age, child sex, or disability characteristics of the child between mother respondents who participated in the study (n = 195) and those who did not (n = 23). Participants Participants were 195 mothers of CWD. See Table I for sociodemographic characteristics of mothers and CWD. Nearly one-quarter (23.1%) of mothers reported an annual household income <$40,000CDN, which approximates the Canadian before-tax, low-income cut-off (LICO; $39,399) for a family of four in 2006 (Statistics Canada, 2006). LICO is a proxy measure of poverty in Canada. Geographically, mothers were representative of both rural and urban areas. Table I. Characteristics of Mothers and Their Child with Disability (N = 195) Mean SD Frequency Percentage  (%) Mother Age (years) 37.6 6.5 Married/cohabiting 161 82.5 Completed high school 176 90.3 Employed 118 60.5 Low-income family 44a 23.1 Child with disability Age (years) 7.92 4.72 Gender (% males) 138 70.8 Child age at diagnosis Prenatal 28 14.4 Neonatal (<28 days) 12 6.2 Infant (<1 year) 15 7.7 Toddler (1–3 years) 49 25.1 Preschool (4–5 years) 42 21.5 School age (6–12 years) 43 22.1 Adolescent (13–17 years) 6 3.1 Diagnostic categories Developmental conditions 107 55.7 Physical/motor impairments 12 6.3 Mental health disorder 36 18.8 Sensory impairment 4 2.1 Complex health condition 27 14.1 Unconfirmed conditions 6 3.1 aData are missing for five participants.

42 21.5 School age (6–12 years) 43 22.1 Adolescent (13–17 years) 6 3.1 Diagnostic categories Developmental conditions 107 55.7 Physical/motor impairments 12 6.3 Mental health disorder 36 18.8 Sensory impairment 4 2.1 Complex health condition 27 14.1 Unconfirmed conditions 6 3.1 aData are missing for five participants. Procedure Between May and September 2007, mothers completed the PMI and FICD, and validation measures, via computer assisted telephone interviews (CATI). CATI is an interactive computer system that aids interviewers to ask questions over the telephone and immediately key answers into a data file. Telephone interviewers were trained to ensure sensitivity to the mothers and were monitored for interview quality throughout the study. To prevent respondent burden, we randomly selected 51 mothers (26.2%) who completed the interview again 4-weeks later to test temporal stability. All mothers selected agreed to participate. To assess predictive validity 1 year after the first interview, 154 mothers completed the PMI and FICD again, along with other validation measures used in these analyses. There were no statistically significant differences on maternal age, family income, child age, or child sex between mothers who completed the longitudinal follow-up (n = 154) and those who did not (n = 41). We obtained informed consent verbally during the CATI. Two university institutional review boards approved the study. We mailed a gift certificate ($40CDN) to recognize mothers’ contributions to the study.

e, or child sex between mothers who completed the longitudinal follow-up (n = 154) and those who did not (n = 41). We obtained informed consent verbally during the CATI. Two university institutional review boards approved the study. We mailed a gift certificate ($40CDN) to recognize mothers’ contributions to the study. Target Measures PMI The PMI (Trute & Hiebert-Murphy, 2005) is a 10-item measure designed to capture positive spirits, psychological energy, and enthusiasm for parenting a CWD. Item (e.g., “When you think of your daily life as a parent, how often do you feel optimistic?”) responses range from 1 (not at all) to 5 (very often). Six items were scored in reverse so all items on the scale were pointed in the same direction; all items were summed to create a total score. Higher scores indicate higher parenting morale. A Canadian study with a sample of 111 mothers of CWD (Trute & Hiebert-Murphy, 2005) reported moderate correlations between scores on the PMI and Parenting Stress Index-Short Form (PSI-SF; r = −.59; Abidin, 1995), and the PMI and Family Assessment Measure (FAM) Brief Form (r = −.50; Skinner, Steinhauer, & Santa-Barbara, 1995). A principal components analysis with varimax rotation yielded a solution with one underlying factor, and a Cronbach’s alpha for mothers of .86 (Trute & Hiebert-Murphy, 2005).

I-SF; r = −.59; Abidin, 1995), and the PMI and Family Assessment Measure (FAM) Brief Form (r = −.50; Skinner, Steinhauer, & Santa-Barbara, 1995). A principal components analysis with varimax rotation yielded a solution with one underlying factor, and a Cronbach’s alpha for mothers of .86 (Trute & Hiebert-Murphy, 2005). FICD The FICD (Trute et al., 2007) is a 20-item measure designed to assess parents’ appraisal of the family consequences of their child having a disability. Item responses range from 1 (not at all) to 4 (substantial degree) on two subscales: FICD Positive (e.g., “Raising a disabled child has made life more meaningful for family members”), and FICD Negative (e.g., “There has been an unwelcome disruption to normal family routines”). FICD Positive and Negative scores were obtained by summing the items in each subscale. In a prior study (N = 103), the Negative and Positive subscales of the FICD significantly predicted maternal perceptions of family functioning (Trute et al., 2007). High internal consistency was reported for mothers (α = .81 Positive; .89 Negative; Trute et al., 2007). An exploratory factor analysis with varimax rotation yielded a two-factor solution with items loading on positive and negative subscales (Trute & Hiebert-Murphy, 2002), and the FICD was correlated concurrently with maternal depression (r = .24), parenting stress (r = .64), and family adjustment (r = .34; Trute & Hiebert-Murphy, 2002). FICD positive (r = −.07) and negative (r = −.10) subscales were not significantly related to social desirability.

scales (Trute & Hiebert-Murphy, 2002), and the FICD was correlated concurrently with maternal depression (r = .24), parenting stress (r = .64), and family adjustment (r = .34; Trute & Hiebert-Murphy, 2002). FICD positive (r = −.07) and negative (r = −.10) subscales were not significantly related to social desirability. Validation Measures Validation measures were selected for (a) coherence with constructs in the process model of stress and coping (Lazarus & Folkman, 1984), (b) sound psychometric properties, (c) suitability for the population, and (d) ability to capture constructs critical to positive adaptation. Respondent burden, social desirability response bias, and ordering of measures were also considered. Measures were ordered to start with general information about the family, then move to more emotionally laden information (e.g., depressive symptoms), and end with demographic information, plus an offer of a gift certificate. Measures could not be counterbalanced because the order was fixed in the CATI delivery format.

Measures were ordered to start with general information about the family, then move to more emotionally laden information (e.g., depressive symptoms), and end with demographic information, plus an offer of a gift certificate. Measures could not be counterbalanced because the order was fixed in the CATI delivery format. Baseline Validation Measures Brief FAM: General Scale The Brief FAM—General Scale (Skinner et al., 1995) is a shorter (14-item) version of the full 50-item, 9 subscale version which provides an overall rating of family functioning. Item (e.g., “We feel loved in our family”) responses range from 0 (strongly agree) to 3 (strongly disagree). Items are summed and translated to T-scores. Lower scores indicate stronger family functioning. Test–retest reliability is .56–.66 over 12 days with good internal consistency (α = .86–.94; Skinner et al., 1995). For this study, Cronbach’s alphas were .88 at baseline and .87 1-year later, and temporal stability over a 4-week interval was r = .71.

ated to T-scores. Lower scores indicate stronger family functioning. Test–retest reliability is .56–.66 over 12 days with good internal consistency (α = .86–.94; Skinner et al., 1995). For this study, Cronbach’s alphas were .88 at baseline and .87 1-year later, and temporal stability over a 4-week interval was r = .71. Personal Well-Being Index The Personal Well-Being Index (PWI; Trivette & Dunst, 1986) is a well-established measure of parental global well-being with 16-items on four subscales: General Emotional; General Physical; Child-Related Emotional; and Child-Related Physical. Each subscale has two positive (e.g., “Feeling that my life is going just great”), and two negative (e.g., “Feeling trapped by my responsibilities”) items rated from 1 (never) to 5 (quite often). Subscale scores are determined by subtracting the negative item points from the positive item points then adding 8. The PWI Total score is the sum of all of the subscale total scores; higher scores indicate higher well-being. The PWI has concurrent validity with the Family Support Scale (Trivette & Dunst, 1986). Test–retest reliability is .56 over 1 month, with strong internal consistency (α = .88). For this study, Cronbach’s alpha for the PWI Total score was .90, and temporal stability over a 4-week interval was r = .82.

higher well-being. The PWI has concurrent validity with the Family Support Scale (Trivette & Dunst, 1986). Test–retest reliability is .56 over 1 month, with strong internal consistency (α = .88). For this study, Cronbach’s alpha for the PWI Total score was .90, and temporal stability over a 4-week interval was r = .82. Positive and Negative Affect Schedule The Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) is a 20-item measure of the frequency of positive and negative emotions over a defined period of time, from right now to in the past year, without affecting internal consistency or factor structure. For this study, in the past week was used. Item (e.g., “excited”, “distressed”) responses range from 1 (rarely or none of the time) to 4 (most or all of the time). Higher scores indicate greater Positive or Negative affect. Test–retest reliabilities over 8 weeks range from .47 to .68. The PANAS Negative is correlated with the Beck Depression Inventory (r = .56−.58; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) and the Hopkins Symptom Checklist (r = .65−.74; Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974; Watson et al., 1988). For this study, Cronbach’s alphas were .88 and .89 for the PANAS Positive and Negative, respectively; temporal stability over a 4-week interval was r = .65 and .80, respectively.

, & Erbaugh, 1961) and the Hopkins Symptom Checklist (r = .65−.74; Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974; Watson et al., 1988). For this study, Cronbach’s alphas were .88 and .89 for the PANAS Positive and Negative, respectively; temporal stability over a 4-week interval was r = .65 and .80, respectively. General Self-Efficacy Scale The General Self-Efficacy Scale (GSE; Schwarzer & Jerusalem, 1995) is a 10-item measure of personal competence. Item (e.g., “I can usually handle whatever comes my way”) responses range from 1 (not at all true) to 4 (exactly true) with higher scores indicating greater perceptions of competence. Internal consistency ranges from .75 to .91 in various cultures (Scholz, Dona, Sud, & Schwarzer, 2002). For this study, Cronbach’s alpha was .82, and temporal stability over a 4-week interval was r = .57. Social Desirability Scale The Social Desirability Scale (SDS) is a 10-item adaptation by Strahan and Gerbasi (1972) of the Marlowe-Crowne (MC) Social Desirability Scale (Crowne & Marlowe, 1960) to capture social desirability response bias. True and false items (e.g., “I have never intensely disliked anyone”) are summed for a total score. The MC 2(10) has internal consistency ranging from .49 to .75. For this study, Cronbach’s alpha was .58, and temporal stability over a 4-week interval was r = .76.

owe, 1960) to capture social desirability response bias. True and false items (e.g., “I have never intensely disliked anyone”) are summed for a total score. The MC 2(10) has internal consistency ranging from .49 to .75. For this study, Cronbach’s alpha was .58, and temporal stability over a 4-week interval was r = .76. One Year Later Validation Measures Center for Epidemiological Studies-Depression Scale The Center for Epidemiological Studies-Depression (CES-D; Radloff, 1977) is 20-item scale designed to measure depressive symptoms in the past week. Item (e.g., “I felt depressed”) responses range from 0 (rarely or none of the time) to 3 (all of the time); higher scores indicate more depressive symptoms. The clinical cut point on the CES-D is 16 (Anthony & Barlow, 2002). Internal consistency is strong (α = .84−.93). Concurrent validity with the Beck Depression Inventory is .86 (Santor, Zuroff, Ramsay, Cervantes, & Palacios, 1995). For this study, Cronbach’s alpha was .86. PSI-SF The PSI-SF (Abidin, 1995) is a 36-item measure of stress related to the parenting role on three subscales: Parental Distress; Parent–Child Dysfunctional Interaction; and Difficult Child. Item (e.g., “My child seems to cry or fuss more than most children”) responses range from 1 (strongly agree) to 5 (strongly disagree); higher scores indicate greater parenting stress. The PSI-SF has concurrent validity with family flexibility and family resources (Abidin, 1995). Internal consistency is strong (α = .80−.91). For this study, Cronbach’s alpha for the PSI-SF Total score was .88.

nses range from 1 (strongly agree) to 5 (strongly disagree); higher scores indicate greater parenting stress. The PSI-SF has concurrent validity with family flexibility and family resources (Abidin, 1995). Internal consistency is strong (α = .80−.91). For this study, Cronbach’s alpha for the PSI-SF Total score was .88. Family Hardiness Index The Family Hardiness Index (FHI; McCubbin, McCubbin, & Thompson, 1987) is a 20-item measure of resistance to stress, and adaptation in families on three subscales: Commitment, Challenge, and Control. Item (e.g., “We work together to solve problems”) responses range from 0 (false) to 3 (true); higher scores indicate greater family hardiness. The FHI has concurrent validity with family flexibility and family time and routines (McCubbin, Thompson, & McCubbin, 1996). Internal consistency is satisfactory (α = .65−.82) and temporal stability is strong (r = .86). For this study, Cronbach’s alpha was .85. Brief FAM: General Scale In addition to being used as a baseline measure, the Brief FAM (Skinner et al., 1995) was used as a 1 year later validation measure. See information in Baseline Validation Measures.

Family Hardiness Index The Family Hardiness Index (FHI; McCubbin, McCubbin, & Thompson, 1987) is a 20-item measure of resistance to stress, and adaptation in families on three subscales: Commitment, Challenge, and Control. Item (e.g., “We work together to solve problems”) responses range from 0 (false) to 3 (true); higher scores indicate greater family hardiness. The FHI has concurrent validity with family flexibility and family time and routines (McCubbin, Thompson, & McCubbin, 1996). Internal consistency is satisfactory (α = .65−.82) and temporal stability is strong (r = .86). For this study, Cronbach’s alpha was .85. Brief FAM: General Scale In addition to being used as a baseline measure, the Brief FAM (Skinner et al., 1995) was used as a 1 year later validation measure. See information in Baseline Validation Measures. Data Analyses There were few missing responses on either the PMI (0.17% missing) or FICD (2.53% missing). Missing values were imputed using regression with the other items on each scale and standard decision rules for each measure. Prior to analyses, data were examined for linearity and normality. Data were not markedly skewed for any measure. Significance was set at p < .05 for all statistical tests. We conducted all analyses in Statistical Package for the Social Sciences (SPSS) version 16.0 software (SPSS, version 16.0; SPSS Inc., Chicago, IL). We calculated descriptive statistics for all measures, including percentile scores for the PMI and FICD. Using maximum likelihood estimation and oblique Promax rotation, as appropriate, we conducted factor analyses on the items for the PMI and FICD to assess the correspondence with the previously identified factor structures from two separate samples of CWD (Trute & Hiebert-Murphy, 2005; Trute et al., 2007). We calculated internal consistency coefficients (Cronbach’s α) and used Pearson’s correlations to calculate temporal stability. We used Pearson’s correlations to assess convergent and discriminant validity (Kazdin, 2003) between the target and validation measures. We used Cohen’s (1969) guidelines to interpret the strength of correlations (i.e., small = .10, medium = .30, and large = .50). Using the PMI and FICD as predictors, we ran separate multiple regression models to predict maternal depressive symptoms, parenting stress, family hardiness, and family functioning over a 1-year interval.

n’s (1969) guidelines to interpret the strength of correlations (i.e., small = .10, medium = .30, and large = .50). Using the PMI and FICD as predictors, we ran separate multiple regression models to predict maternal depressive symptoms, parenting stress, family hardiness, and family functioning over a 1-year interval. Results PMI Factor Analysis, Reliability, and Validity Factor analysis, using maximum likelihood estimation and no rotation, suggests that a single factor solution fits the data, χ2(35) = 191.87, p < .001. Uni-dimensionality was also indicated with an assessment of the intersection of the confidence intervals for the eigenvalues (Reddon, 1997). Factor loadings ranged from .52 (optimistic) to .75 (satisfied). Descriptive statistics, reliabilities, and percentile scores for the PMI are presented in Table II. Internal consistency and temporal stability for the PMI were strong. Descriptive statistics on baseline validation measures are presented in Table III. Based on Cohen’s (1969) guidelines for interpreting the strength of correlations, there was a large correlation, in the expected direction, between scores on the PMI and Brief FAM, r(194) = −.48, p < .01. Correlations between target and baseline validation measures are presented in Table IV. Similarly, there was a large, positive correlation between scores on the PMI and PWI Total, r(193) = .84, p < .001. There were large correlations in the expected direction between scores on the PMI and PANAS Positive, r(195) = .63, p < .001 and Negative subscales, r(193) = −.69, p < .001. In contrast, there was a medium, positive correlation between scores on the PMI and GSE, r(193) = .35, p < .001. Similarly, there was a medium, positive correlation between scores on the PMI and SDS, r(193) = .26, p < .001. Table II. Descriptive Statistics, Reliabilities, and Percentiles for the Parenting Morale Index and Family Impact of Childhood Disability Positive and Negative Subscales

he PMI and GSE, r(193) = .35, p < .001. Similarly, there was a medium, positive correlation between scores on the PMI and SDS, r(193) = .26, p < .001. Table II. Descriptive Statistics, Reliabilities, and Percentiles for the Parenting Morale Index and Family Impact of Childhood Disability Positive and Negative Subscales Full samplea Percentilea Test– retestb Scale Number of items α M SD 80th 90th 98th r PMI 10 .88 30.5 7.1 37 39 44 .88 FICD Positive 10 .85 29.9 5.9 36 37 40 .77 FICD Negative 10 .86 26.5 7.0 33 36 39 .86 aN = 195. bn = 51. Table III. Scores on Measures to Establish Concurrent Validity and Predictive Validity One Year Later Validation measures n M SD Range Baseline Brief FAM 194 10.42 5.99 0–29 PWI General Emotional 193 9.64 3.50 0–16 PWI General Physical 193 7.03 3.42 0–16 PWI Child Emotional 193 9.78 3.56 1–16 PWI Child Physical 193 8.11 3.37 1–16 PWI Total 193 34.56 11.88 6–61 PANAS Positive 195 29.82 5.98 12–40 PANAS Negative 193 19.06 6.49 10–39 General Self-Efficacy 193 31.34 3.82 19–40 One year later CES-D 151 19.34 6.39 10–39 PSI-SF 145 99.20 24.88 45–164 FHI 150 46.13 8.07 24–60 Brief FAM 151 10.87 6.10 0–28 Table IV. Correlations between Mothers’ Scores on Target and Baseline Validation Measures

Validation measures n M SD Range Baseline Brief FAM 194 10.42 5.99 0–29 PWI General Emotional 193 9.64 3.50 0–16 PWI General Physical 193 7.03 3.42 0–16 PWI Child Emotional 193 9.78 3.56 1–16 PWI Child Physical 193 8.11 3.37 1–16 PWI Total 193 34.56 11.88 6–61 PANAS Positive 195 29.82 5.98 12–40 PANAS Negative 193 19.06 6.49 10–39 General Self-Efficacy 193 31.34 3.82 19–40 One year later CES-D 151 19.34 6.39 10–39 PSI-SF 145 99.20 24.88 45–164 FHI 150 46.13 8.07 24–60 Brief FAM 151 10.87 6.10 0–28 Table IV. Correlations between Mothers’ Scores on Target and Baseline Validation Measures Target measures Validation measures PMI FICD positive FICD negative Brief FAM −.48*** −.42*** .11 PWI General Emotional .77*** .15* −.47*** PWI General Physical .72*** .20** −.48*** PWI Child Emotional .68*** .25** −.52*** PWI Child Physical .71*** .21** −.57*** PWI Total .84*** .24** −.59*** PANAS Positive .63*** .31*** −.33*** PANAS Negative −.69*** −.10 .39*** General Self-Efficacy .35*** .20** −.19** *p < .05, **p < .01, ***p < .001.

77*** .15* −.47*** PWI General Physical .72*** .20** −.48*** PWI Child Emotional .68*** .25** −.52*** PWI Child Physical .71*** .21** −.57*** PWI Total .84*** .24** −.59*** PANAS Positive .63*** .31*** −.33*** PANAS Negative −.69*** −.10 .39*** General Self-Efficacy .35*** .20** −.19** *p < .05, **p < .01, ***p < .001. FICD Factor Analysis, Reliability, and Validity Factor analysis, using maximum likelihood estimation and oblique Promax rotation, suggests that a two-factor solution fits the data, χ2(151) = 314.57, p < .001. Evidence in favor of the two-dimensional solution was also obtained with Reddon’s (1997) confidence interval scree test. For the FICD Positive subscale, factor loadings ranged from .48 (item 3, “closer to God”) to .77 (item 20, “life more meaningful”). For the FICD Negative subscale, factor loadings ranged from .40 (item 2, “unwelcome disruptions”) to .78 (item 10, “reduction in time parents could spend with friends”).

terval scree test. For the FICD Positive subscale, factor loadings ranged from .48 (item 3, “closer to God”) to .77 (item 20, “life more meaningful”). For the FICD Negative subscale, factor loadings ranged from .40 (item 2, “unwelcome disruptions”) to .78 (item 10, “reduction in time parents could spend with friends”). Descriptive statistics, reliabilities, and percentile scores are presented in Table II. Internal consistency and temporal stability for the FICD Positive and Negative subscales were strong. There was a medium, negative correlation between scores on the FICD Positive subscale and Brief FAM, r (191) = −.42, p < .001. Contrary to the hypothesis, there was a small, positive correlation between scores on the FICD Negative subscale and Brief FAM, r(194) = .11, p = .12. There was a large, positive correlation between scores on FICD Negative subscale and PWI Total scores, r(193) = −.59, p < .001, and a smaller, positive correlations between scores on the FICD Positive subscale, r (190) = .24, p < .001. FICD Positive and Negative subscales were statistically independent, r(191) = .08, p = .27. FICD Positive and Negative subscales were independent of the SDS score, r(191) = .09, p = .23, and r(193) = −.12, p = .10, respectively.

ositive correlations between scores on the FICD Positive subscale, r (190) = .24, p < .001. FICD Positive and Negative subscales were statistically independent, r(191) = .08, p = .27. FICD Positive and Negative subscales were independent of the SDS score, r(191) = .09, p = .23, and r(193) = −.12, p = .10, respectively. Predictive Validity The PMI and FICD Positive and Negative subscales were significant predictors, and together explained 30% of the variance in maternal depressive symptoms 1-year later (see Table V). The PMI made the greatest contribution to the variance explained in maternal depressive symptoms; the FICD subscales failed to make a statistically significant contribution. Together, the PMI and FICD Positive and Negative subscales explained 36% of the variance in parenting stress. Again, the PMI made the greatest contribution to the variance explained, while the FICD subscales made an additional, statistically significant contribution to the variance explained in parenting stress. A slightly different pattern emerges when using the PMI and FICD subscales to predict family hardiness. Together the PMI and FICD subscales explain 29% of the variance in family hardiness, and the PMI again makes the greatest contribution to the variance explained. However, only the FICD Positive subscale makes a statistically significant contribution to the amount of variance explained in family hardiness. This pattern is similar when using the PMI and FICD subscales to explain variance in family adjustment. Table V. Summary of Multiple Regressions with the PMI and FICD Predicting Depressive Symptoms, Parenting Stress, Family Hardiness, and Family Functioning 1 Year Later

amount of variance explained in family hardiness. This pattern is similar when using the PMI and FICD subscales to explain variance in family adjustment. Table V. Summary of Multiple Regressions with the PMI and FICD Predicting Depressive Symptoms, Parenting Stress, Family Hardiness, and Family Functioning 1 Year Later CES-D (n = 151)a PSI-SF (n = 145)b FHI (n = 150)c Brief FAM (n = 151)d B SE β B SE β B SE β B SE β PMI −.51 .08 −.55** −1.46 .31 −.41** .50 .11 .43** .38 .08 .44** FICD Positive .02 .08 .02 −.84 .31 −.19* .30 .11 .22* .21 .08 .20* FICD Negative .03 .08 .03 .67 .31 .19* -.06 .11 −.05 .10 .08 .12 aAdjusted R2 = .30, F(3, 147) = 22.86, p < .001. bAdjusted R2 = .36, F(3, 141) = 27.99, p < .001. cAdjusted R2 = .29, F(3, 146) = 21.17, p < .001. dAdjusted R2 = .22, F(3, 147) = 15.20, p < .001. *p < .01. **p < .05. ***p < .001.

CES-D (n = 151)a PSI-SF (n = 145)b FHI (n = 150)c Brief FAM (n = 151)d B SE β B SE β B SE β B SE β PMI −.51 .08 −.55** −1.46 .31 −.41** .50 .11 .43** .38 .08 .44** FICD Positive .02 .08 .02 −.84 .31 −.19* .30 .11 .22* .21 .08 .20* FICD Negative .03 .08 .03 .67 .31 .19* -.06 .11 −.05 .10 .08 .12 aAdjusted R2 = .30, F(3, 147) = 22.86, p < .001. bAdjusted R2 = .36, F(3, 141) = 27.99, p < .001. cAdjusted R2 = .29, F(3, 146) = 21.17, p < .001. dAdjusted R2 = .22, F(3, 147) = 15.20, p < .001. *p < .01. **p < .05. ***p < .001. Discussion This study contributes to family assessment in pediatric psychology by documenting the unique information about family adaptation that can be gleaned from two measures of psychological well-being in mothers of CWD. The results of this study provide a first step in the development of a brief, standardized package of measures (PMI and FICD) to complement clinical interviews for the assessment of psychological well-being in mothers of CWD. The results of this study suggest that these brief measures are of acceptable psychometric strength such that practitioners can have confidence in their empirical properties. Both measures showed high internal consistency (Cronbach’s alpha) and temporal stability (test–retest) to suggest measurement reliability. Both showed strong evidence of factorial, discriminant, and predictive validity. There were very few missing values on the PMI and FICD. This suggests that these measures are acceptable to mothers when providing information on their personal and family situation in the context of childhood disability.

ment reliability. Both showed strong evidence of factorial, discriminant, and predictive validity. There were very few missing values on the PMI and FICD. This suggests that these measures are acceptable to mothers when providing information on their personal and family situation in the context of childhood disability. Consistent with the elements in the process model of stress and coping (Lazarus & Folkman, 1984), the PMI appears to be a brief and unique indicator of internal psychological coping resources that each mother draws upon to cope with the daily needs of their CWD. Scores on the PMI were strongly correlated with the PWI Total score suggesting convergent validity of the two measures. Smaller, but still strong correlations between the PMI and PANAS suggest that each is capturing similar, but not identical constructs. That is, the PMI does not appear to be a simple measure of affect. Medium correlations between the PMI and GSE suggest that the PMI is not a measure of self-efficacy. Correlations between the PMI and SDS were <.30, which suggests that social desirability response bias does not confound PMI scores. Over a 1-year interval, the PMI showed a moderate relationship (r > .40) with all criterion measures of parent and family functioning (CES-D, PSI-SF, FHI, and Brief FAM).

of self-efficacy. Correlations between the PMI and SDS were <.30, which suggests that social desirability response bias does not confound PMI scores. Over a 1-year interval, the PMI showed a moderate relationship (r > .40) with all criterion measures of parent and family functioning (CES-D, PSI-SF, FHI, and Brief FAM). Within the process model of stress and coping (Lazarus & Folkman, 1984), the FICD provides unique information about the cognitive appraisal, or the meaning parents make of childhood disability as a factor in their family life and family well-being. Factor analysis helped to establish that the FICD has two orthogonal subscales: positive and negative appraisal of the family impact of childhood disability. Further, the results of correlation analysis suggest that each of these dimensions tap a different element of parenting stress and parental psychological coping resources. Positive appraisal appears to be related to mothers’ view of the ongoing functioning of their family (Brief FAM). This suggests that the measures share common variance (i.e., a focus on the family), but are unique enough that the FICD Positive subscale is measuring something other than family functioning. In contrast, the FICD Negative subscale was unrelated to family functioning suggesting that there was little relationship between mothers’ perceptions of the negative family impact of childhood disability and family functioning. Negative appraisal was found to be moderately related to mothers’ overall expression of affect (PANAS), and a measure of their emotional well-being (PWI).

functioning suggesting that there was little relationship between mothers’ perceptions of the negative family impact of childhood disability and family functioning. Negative appraisal was found to be moderately related to mothers’ overall expression of affect (PANAS), and a measure of their emotional well-being (PWI). After a 1-year interval, the FICD was significantly (albeit weakly) related to measures of parenting stress (PSI-SF), family hardiness (FHI), and family functioning (Brief FAM). Negative appraisal was found to have a moderate relationship with parenting stress (PSI-SF). This suggests that the negative appraisal subscale is a brief indicator of level of parenting stress mothers will experience in the longer term. Small, non-significant correlations between scores on the FICD and SDS suggest that social desirability response bias is not an issue with the FICD. Neither FICD positive nor negative appraisal appeared to be a significant predictor of mothers’ symptoms of depression (CES-D) over the 1-year interval suggesting the FICD does not forecast mothers’ psychological well-being in the longer term.

uggest that social desirability response bias is not an issue with the FICD. Neither FICD positive nor negative appraisal appeared to be a significant predictor of mothers’ symptoms of depression (CES-D) over the 1-year interval suggesting the FICD does not forecast mothers’ psychological well-being in the longer term. It seems that the PMI and FICD are unique, yet complement one another, with the PMI offering an overall indicator of mothers’ parenting morale or psychological coping resources and the FICD serving as an overall indicator of mothers’ attitudes and perceptions of the impacts that a CWD has on the well-being of their family. This joint relationship was confirmed in predictive validity testing using multiple regression analysis. After 1 year, the PMI and FICD jointly explained 30% of the variance in mothers’ symptoms of depression, and 36% of the variance in parenting stress. Similarly, the PMI and FICD jointly explained 22% of the variance in mothers’ assessment of overall family functioning, and 29% of family hardiness after 1 year.

nalysis. After 1 year, the PMI and FICD jointly explained 30% of the variance in mothers’ symptoms of depression, and 36% of the variance in parenting stress. Similarly, the PMI and FICD jointly explained 22% of the variance in mothers’ assessment of overall family functioning, and 29% of family hardiness after 1 year. This research is limited by a sample of mother only respondents. Given the differences in parental responses to CWD (Hastings et al., 2005; Trute, 1995), future studies need to include fathers. Previous studies of the PMI and FICD validated a face-to-face delivery format (Trute & Hiebert-Murphy, 2002, 2005; Trute et al., 2007), and the results of this study validated telephone delivery. Future studies are required to validate internet and mailed delivery formats. The diversity in the sample allows generalization to rural and urban mothers with a range of family incomes. However, the Canadian sample was largely comprised of mothers of European descent and results cannot, therefore, be generalized across cultures. Future studies need to include culturally diverse samples. Future research with culturally and diagnostically diverse subpopulations is needed to explore whether the PMI and FICD are similarly applicable. The wide range of ages for children in this study was a threat to internal validity. However, this age range is representative of the children served by FSCD, and thus strengthens the ecological validity of the study findings. Additionally, future research is required to determine whether the PMI and FICD can be used repeatedly to monitor changes in mothers’ psychological well-being as a result of childhood disability services. Finally, the low Cronbach’s alpha (.58) on the SDS for this study suggests that results related to social desirability response bias need to be treated with caution.

mine whether the PMI and FICD can be used repeatedly to monitor changes in mothers’ psychological well-being as a result of childhood disability services. Finally, the low Cronbach’s alpha (.58) on the SDS for this study suggests that results related to social desirability response bias need to be treated with caution. Maternal cognitive appraisal of the family impacts of childhood disability and parenting morale are not simple assessment issues that can be readily addressed and quickly understood during brief service intake interviews. It is important that professionals do not assume that just because a child has a serious disability that this will inevitably lead to family distress. Many mothers will respond to the challenge of childhood disability with positive coping and resiliency. However, it is important to identify those situations where there is increased risk for maternal and family distress. The results of our study suggest that the FICD and PMI can complement and enrich a service intake interview when the need for resources to support the care of her CWD is being considered. In the early phases of childhood disability services, questions about the allocation of scarce resources to ultimately improve outcomes are at the core of service intake interviews when professionals must determine which mothers have a higher priority. Broad implementation and evaluation of the PMI and FICD as measures to complement clinical interviews at intake to service is required. Future research needs to examine whether the addition of the PMI and FICD to clinical interview results in more effective allocation of psychosocial supports and improved outcomes for mothers of CWD.

ation and evaluation of the PMI and FICD as measures to complement clinical interviews at intake to service is required. Future research needs to examine whether the addition of the PMI and FICD to clinical interview results in more effective allocation of psychosocial supports and improved outcomes for mothers of CWD. Funding Alberta Centre for Child, Family and Community Research (#06040TD and #08SM to KB). Conflicts of interest: None declared. Acknowledgments The authors wish to acknowledge the contribution of Kim Nagan, Project Director, Richelle Mychasiuk, Research Associate, and Rosanna Shih and Donna Fong, Population Research Laboratory, University of Alberta. They also wish to thank the mothers of children with disabilities for taking time from their very busy schedules to participate.

Introduction The past decade has witnessed an increase in the number of studies reporting on the prevalence of traumatic stress among parents of children with cancer (Best, Streisand, Catania, & Kazak, 2001; Jurbergs, Long, Ticona, & Phipps, 2009; Kazak, Boeving, Alderfer, Hwang, & Reilly, 2005; Norberg, Lindblad, & Boman, 2005; Phipps, Long, Hudson, & Rai, 2005; Pöder, Ljungman, & von Essen, 2008). This growing body of research has built on the posttraumatic stress disorder (PTSD) symptomatology as described in the Diagnostic and Statistical Manual of Mental Disorders – Fourth Edition (DSM-IV; American Psychiatric Association, 2000). The criteria for PTSD requires exposure to a traumatic event, after which a response of intense fear, helplessness, or horror follows (Criterion A). According to the DSM-IV, medical stressors such as “learning that one's child has a life-threatening illness” can be a traumatic event potentially leading to PTSD. PTSD comprises 17 posttraumatic stress symptoms (PTSS) pertaining to three factors or symptom clusters: reexperiencing (Criterion B), avoidance/numbing (Criterion C), and hyperarousal (Criterion D). However, an increasing number of studies have failed to confirm the validity of the DSM-IV three-factor solution for a wide variety of populations, suggesting that reexperiencing, avoidance/numbing, and hyperarousal, respectively, may not adequately capture the underlying dimensions of PTSD. In view of this emerging literature, the present study aimed to investigate the construct validity of competing models of the underlying dimensions of PTSD among parents of children with cancer.

xperiencing, avoidance/numbing, and hyperarousal, respectively, may not adequately capture the underlying dimensions of PTSD. In view of this emerging literature, the present study aimed to investigate the construct validity of competing models of the underlying dimensions of PTSD among parents of children with cancer. Research focusing on the levels of PTSS among parents of children with cancer is certainly a matter of dispute, as existing data demonstrate inconsistencies concerning the levels of PTSS among parents of children with cancer when compared to parents of healthy children (Barakat et al., 1997; Brown, Madan-Swain, & Lambert, 2003; Jurbergs et al., 2009; Kazak et al., 2005; Pelcovitz et al., 1996). This underscores the need of employing more sophisticated research strategies (Bruce, 2006; Jurbergs et al., 2009; Pöder et al., 2008) especially as there is a lack of conceptual models to guide clinical practice and empirical research targeting traumatic experiences among parents of pediatric oncology patients. The application of the PTSD symptomatology to this population has been called in question, given the apparent difference between common sources of trauma and medical stressors (Mundy & Baum, 2004). One key difference is that common traumatic stressors in general are past-event oriented, whereas medical stressors not only may refer to past events, such as the specific situation surrounding a diagnosis, but also to future-oriented aspects relating to fears and worries about treatment, recurrence, survival, and so forth.

ey difference is that common traumatic stressors in general are past-event oriented, whereas medical stressors not only may refer to past events, such as the specific situation surrounding a diagnosis, but also to future-oriented aspects relating to fears and worries about treatment, recurrence, survival, and so forth. These nosological issues were subjected to a closer inspection at the National Child Traumatic Stress Network (2003), in which a collaborative effort aimed at elaborating on clinical and empirical knowledge concerning pediatric patients and their next of kin was made. To this end, a conceptual model of pediatric medical traumatic stress (PMTS) was established to bring new dimensions to this line of research (Kazak et al., 2006; Pai & Kazak, 2006). PMTS was defined as “a set of psychological and physiological responses of children and their families to pain, injury, serious illness, medical procedures, and invasive or frightening treatment experiences” (The National Child Traumatic Stress Network, 2003). Like PTSD and acute stress disorder (ASD), PMTS covers key traumatic symptoms such as reexperiencing, avoidance/numbing, and hyperarousal, though PMTS is not proposed as a diagnostic entity. Rather, PMTS is conceptualized as a continuum of symptoms, which not necessarily entails a formal diagnosis of PTSD or ASD. Thus, PMTS is operationalized as symptoms of traumatic stress (yet in a pediatric context) and is therefore assessed with instruments developed for assessing symptoms of traumatic stress (Kazak et al., 2006; Pai & Kazak, 2006).

as a continuum of symptoms, which not necessarily entails a formal diagnosis of PTSD or ASD. Thus, PMTS is operationalized as symptoms of traumatic stress (yet in a pediatric context) and is therefore assessed with instruments developed for assessing symptoms of traumatic stress (Kazak et al., 2006; Pai & Kazak, 2006). The advent of PMTS may contribute to an increased conceptual clarity of psychosocial aspects related to pediatric oncology as it provides a framework from where symptoms of traumatic stress could be understood and at the same time avoiding some of the conceptual problems that the application of pure ASD and PTSD nomenclature entails in the context of medical stressors (as outlined above). One way of further adding to such clarity would be to determine the underlying dimensions of PTSD among parents of children with cancer by examining the factor structure of PTSS. Yet, to the best of our knowledge, the factor structure of PTSS in this group has thus far not been addressed. However, a growing body of evidence from various studies indicates that the predominant PTSD model, as defined in the DSM-IV, is indeed a question at issue. Prior research encompassing both exploratory factor analytic (EFA) and confirmatory factor analysis (CFA) has repeatedly failed to prove empirical support for the DSM-IV three-factor model (Baschnagel, O'Connor, Colder, & Hawk, 2005; DuHamel et al., 2004; Elklit & Shevlin, 2007; King, Leskin, King, & Weathers, 1998; Krause, Kaltman, Goodman, & Dutton, 2007; Marshall, 2004; McWilliams, Cox, & Asmundson, 2005; Palmieri & Fitzgerald, 2005; Palmieri, Weathers, Difede, & King, 2007; Simms, Watson, & Doebbeling, 2002). Instead, there are two competing four-factor models (King et al., 1998; Simms et al., 2002) that by means of CFA have gained the strongest empirical support when evaluated against other proposed models of PTSD. In the first of these four-factor models, King et al. (1998) distinguished the symptoms pertaining to the factor of avoidance/numbing (Criterion C) into two separate factors: effortful avoidance (C1 and C2) and emotional numbing (C3–C7). Thus, the King et al. (1998) model was comprised of the reexperiencing (B1–B5), effortful avoidance (C1 and C2), emotional numbing (C3–C7), and hyperarousal (D1–D5) factors.

taining to the factor of avoidance/numbing (Criterion C) into two separate factors: effortful avoidance (C1 and C2) and emotional numbing (C3–C7). Thus, the King et al. (1998) model was comprised of the reexperiencing (B1–B5), effortful avoidance (C1 and C2), emotional numbing (C3–C7), and hyperarousal (D1–D5) factors. However, Simms et al. (2002) found that a different four-factor model provided the best fit to their data. In conformity with the King et al. (1998) model, Simms et al. (2002) found an intrusion (or reexperiencing) factor (B1–B5) and an avoidance factor (C1 and C2) comprising only two symptoms. However, the Simms et al. (2002) model included a factor of nonspecific, general distress that was termed dysphoria, which comprised symptoms of emotional numbing (C3–C7) and hyperarousal (D1–D3). The remaining two symptoms loaded on a distinctive factor, termed hyperarousal (D4 and D5).

) comprising only two symptoms. However, the Simms et al. (2002) model included a factor of nonspecific, general distress that was termed dysphoria, which comprised symptoms of emotional numbing (C3–C7) and hyperarousal (D1–D3). The remaining two symptoms loaded on a distinctive factor, termed hyperarousal (D4 and D5). In the published CFA studies that support either the King et al. (1998) model (DuHamel et al., 2004; King et al., 1998; Marshall, 2004; McWilliams et al., 2005; Palmieri & Fitzgerald, 2005) or the Simms et al. (2002) model (Baschnagel et al., 2005; Elklit & Shevlin, 2007; Krause et al., 2007; Palmieri et al., 2007; Simms et al., 2002), data have been collected from a variety of populations, e.g., undergraduate students in New York after the September 11, 2001 terrorist attacks (Baschnagel et al., 2005), survivors of bone marrow or stem cell transplantation (DuHamel et al., 2004), low-income minority women exposed to intimate partner violence (Krause et al., 2007), victims of community violence (Marshall, 2004), and sexually harassed women (Palmieri & Fitzgerald, 2005).

2001 terrorist attacks (Baschnagel et al., 2005), survivors of bone marrow or stem cell transplantation (DuHamel et al., 2004), low-income minority women exposed to intimate partner violence (Krause et al., 2007), victims of community violence (Marshall, 2004), and sexually harassed women (Palmieri & Fitzgerald, 2005). Research on the factor structure of PTSS among parents of children with cancer would shed new light on the phenomenology and construct validity of the model of PMTS for this population. A central research objective is to determine whether items designed to measure PTSS function in the same way among parents of children with cancer as they do in other trauma populations, and whether the symptom structures, or patterns of factor loadings, remain stable over time. To date, only two studies have examined the structural invariance of PTSS over time (Baschnagel et al., 2005; Krause et al., 2007), albeit with somewhat disparate data analytic strategies and with dissimilar samples. Based on data from two time points, 1 and 3 months after the September 11, 2001 terrorist attacks, Baschnagel et al. (2005) found that the Simms et al. (2002) model provided the best-fitting factor solution when evaluated against several other proposed models, including the King model. Moreover, Krause et al. (2007) collected data from two samples of low-income minority women exposed to intimate partner violence at two time points: approximately within 3 months after exposure to violence, and then around 1 year thereafter. Krause et al. (2007) also found that the Simms et al. (2002) model represented the best-fitting factor solution across time and setting compared to other examined models such as the King et al. (1998) model.

at two time points: approximately within 3 months after exposure to violence, and then around 1 year thereafter. Krause et al. (2007) also found that the Simms et al. (2002) model represented the best-fitting factor solution across time and setting compared to other examined models such as the King et al. (1998) model. In the present study, we used the PTSD Checklist–Civilian Version (PCL-C; Weathers, Litz, Herman, Huska, & Keane, 1993) to measure PTSS and to compare three models of the underlying dimensions of PTSD among parents of children diagnosed with cancer shortly after diagnosis and 2 and 4 months after diagnosis. The PCL-C consists of 17 items that map directly on the corresponding symptoms in one of the three-symptom clusters of reexperiencing (Criterion B), avoidance/numbing (Criterion C), and hyperarousal (Criterion D) in the DSM-IV. We hypothesized that a four-factor model would provide better fit to the data than the current DSM-IV three-factor conceptualization. Furthermore, based on the findings by Krause et al. (2007) and Simms et al. (2002), we hypothesized that the best-fitting factor solution would evidence stability over time when testing model invariance with data collected 2 and 4 months after diagnosis.

the data than the current DSM-IV three-factor conceptualization. Furthermore, based on the findings by Krause et al. (2007) and Simms et al. (2002), we hypothesized that the best-fitting factor solution would evidence stability over time when testing model invariance with data collected 2 and 4 months after diagnosis. Methods Data were collected in a project with a longitudinal design investigating disease and care-related responses of parents of children with cancer. The design covers seven assessments: 2 weeks after diagnosis (T1), 2 (T2), and 4 (T3) months after diagnosis, 1 week after end of treatment (T4), and 3 (T5), 12 (T6), and 60 (T7) months after end of treatment or the child's death. Participants were included between April 2002 and February 2004. Results from this project have been reported previously and these publications focused on describing proportions of PTSD caseness at T1–T3 (Pöder et al., 2008), perceptions of support and satisfaction with care at T1–T3 (Pöder & von Essen, 2009), perceptions of the child's cancer-related symptoms at T1–T3 (Pöder, Ljungman & von Essen, 2010), and the relationship between avoidance symptoms at T1–T4 and levels of PTSS at T6 (Lindahl Norberg, Pöder, & von Essen, 2011). None of these publications were concerned with the factor structure of PTSS. For the purpose of the present analyses, we used data from T1, T2, and T3.

1–T3 (Pöder, Ljungman & von Essen, 2010), and the relationship between avoidance symptoms at T1–T4 and levels of PTSS at T6 (Lindahl Norberg, Pöder, & von Essen, 2011). None of these publications were concerned with the factor structure of PTSS. For the purpose of the present analyses, we used data from T1, T2, and T3. Participants There were 315 eligible parents during the inclusion period. Two hundred and forty-five parents (128 mothers and 121 fathers) of 137 children treated for cancer at four pediatric oncology centers in Sweden consented to participation representing a 79% response rate. At the time of diagnosis, the mothers’ mean age (SD) was 37 (6.3) and the fathers’ mean age (SD) was 40 (6.8). Regarding educational level, 33% of the parents had completed university education, 51% upper secondary school, and 14% had finished elementary school. The children's mean age (SD) at the time of diagnosis was 8 (5.2) years. The distribution of diagnoses was as follows: Leukemia 40%, Lymphoma 19% Sarcoma 14%, CNS tumor 13%, and other malignant disease 14%. A series of one-way ANOVAs indicated that there were no significant effects of recruitment center on the level of PTSS at any time point, neither in terms of the full scale nor any of the subscales (df = 3, F's ranging between 2.28 and 0.07, p-values ranging between .797 and .079). Out of the 249 parents at T1, 234 provided data at T2 and 203 provided data at T3, respectively.

Participants There were 315 eligible parents during the inclusion period. Two hundred and forty-five parents (128 mothers and 121 fathers) of 137 children treated for cancer at four pediatric oncology centers in Sweden consented to participation representing a 79% response rate. At the time of diagnosis, the mothers’ mean age (SD) was 37 (6.3) and the fathers’ mean age (SD) was 40 (6.8). Regarding educational level, 33% of the parents had completed university education, 51% upper secondary school, and 14% had finished elementary school. The children's mean age (SD) at the time of diagnosis was 8 (5.2) years. The distribution of diagnoses was as follows: Leukemia 40%, Lymphoma 19% Sarcoma 14%, CNS tumor 13%, and other malignant disease 14%. A series of one-way ANOVAs indicated that there were no significant effects of recruitment center on the level of PTSS at any time point, neither in terms of the full scale nor any of the subscales (df = 3, F's ranging between 2.28 and 0.07, p-values ranging between .797 and .079). Out of the 249 parents at T1, 234 provided data at T2 and 203 provided data at T3, respectively. Measures PTSS was assessed with the PCL-C (Weathers et al., 1993), which contains 17 items corresponding to the DSM-IV symptom clusters of reexperiencing (Items 1–5), avoidance/numbing (6–12), and hyperarousal (13–17). The respondents were asked to rate to which extent they had been bothered by each symptom during the previous month. Items were keyed to the child's disease thus providing an indicator of PTSS associated with their child's disease (i.e., PMTS). Ruggiero, Ben, Scotti, and Rabalais (2003) have provided the most thorough investigation on the psychometric properties of the PCL-C. They report that the instrument has adequate internal consistency, test–retest reliability, and that there is evidence for convergent and discriminant validity when compared to other well-established PTSS measures as well as measures of depression and general anxiety. A value of 44 or above on the full scale has been suggested as a clinical cut off suggesting a diagnosis of PTSD (Blanchard, Jones-Alexander, Buckley & Forneris, 1996).

nt. The research assistant then, via telephone, conducted the interview where the PCL-C and other instruments (not reported herein) were administered. Permission to be contacted again was obtained at the end of the interview. The procedure was approved by the ethical review board at each respective faculty of medicine. Statistical Analyses Confirmatory factor analysis (CFA) using Mplus 6.1 (Muthén & Muthén, 1998–2010) were performed as the primary method of analyses. The analytic strategy consisted of subjecting the three theoretical models (DSM-IV, Simms and King outlined in the ‘Introduction’ section) of PTSS factor structure to CFA to determine the best model fit to the current data. This was conducted by performing a longitudinal CFA and testing for measurement invariance across time for each of the three models. In order to control for the dependent nature of the data, i.e., parent dyads nested in children, which can potentially bias standard errors and χ2 estimates, we used the TYPE = COMPLEX and CLUSTER commands in Mplus. We used MLR estimation which is the default estimator in Mplus for this procedure which produce estimates of χ2 and standard errors that are robust to nonindependence and non-normality (Muthén & Muthén, 1998–2010). Measurement invariance was tested in three steps. First, a configural model was tested where all factor loadings and covariances were allowed to be freely estimated. Secondly, metric invariance was tested by constraining factor loadings to be equal across time. Thirdly, phi invariance was investigated by adding constraints on factor covariances to be equal across time. Measurement invariance was investigated with the Satorra–Bentler scaled Δ χ2-test, which is recommended when using MLR estimation (Satorra, 2000), and Δ CFI where convention suggests values equal to or lower than −.01 as nonsignificant (Cheung & Rensvold, 2002). Model test statistics of fit included χ2-tests, and approximate fit indexes used were Steiger–Lind root-mean-square error of approximation (RMSEA; Steiger, 1990) and Bentler comparative fit index (CFI; Bentler, 1990). According to Byrne (2010) RMSEA, values < 0.05 indicate good fit and values ranging between 0.08 and 0.10 moderate fit, while CFI values close to 0.95 indicate good fit and values > 0.90 acceptable fit. For the purpose of comparing fit between models, sample size adjusted Bayesian information criteria (BIC; Raftery, 1995) was used, with lower values indicating better model fit.

fit and values ranging between 0.08 and 0.10 moderate fit, while CFI values close to 0.95 indicate good fit and values > 0.90 acceptable fit. For the purpose of comparing fit between models, sample size adjusted Bayesian information criteria (BIC; Raftery, 1995) was used, with lower values indicating better model fit. Standardized factor loading estimates were used as indices of construct validity and values exceeding .50 were considered to reflect adequate construct validity (Hair, Black, Babin, & Anderson, 2010). Internal consistency for each factor at each assessment was analyzed with Cronbach's α. Finally, descriptive statistics were used to describe the participants in terms of the chosen model. In order to control for the dependent nature of the data, we used linear mixed models with child as random intercept to estimate and test differences in PTSS between mothers and fathers at each assessment point. Repeated measures ANOVA was used to estimate and test change over time among mothers and fathers, respectively. Linear mixed models and repeated measures ANOVA were conducted in IBM SPSS Statistics 19.0.

d as random intercept to estimate and test differences in PTSS between mothers and fathers at each assessment point. Repeated measures ANOVA was used to estimate and test change over time among mothers and fathers, respectively. Linear mixed models and repeated measures ANOVA were conducted in IBM SPSS Statistics 19.0. Results As a first step, we evaluated all three models cross-sectionally with data from each of the three assessments. As is evident from Table I, all models evidenced good to acceptable fit at all three assessments. Inspection of BIC reveals that the Simms model provided best fit at T1 and T2 and that the King model provided best fit at T3. Results from the primary analyses incorporating data from all three assessments in longitudinal CFA are presented in Table II. Comparing baseline configural models, the Simms model had the highest CFI (indicating acceptable fit), equally low RMSEA as the King model (indicating good fit) and the lowest BIC value, indicating that this was the best representation of a longitudinal analyses of the factor structure. When testing for metric invariance (i.e., factor loadings constrained to be equal across time), both the King and the Simms model evidenced nonsignificant Satorra–Bentler scaled Δχ2-test and ΔCFI closer to zero than −.01. In comparison, the DSM-IV model exhibited a significant increase in the Satorra–Bentler scaled Δχ2-test and a ΔCFI closer to −.01. However, when testing for phi invariance all three models had significant Satorra–Bentler scaled Δχ2-tests and ΔCFI closer to −.01 than zero. Thus, both the King and the Simms model evidenced acceptable to good fit, metric invariance, and phi noninvariance. However, when comparing models with BIC, the Simms model evidenced best fit (i.e., lower value) and was therefore chosen for further descriptive analyses. Table I. Fit Statistics for Cross-Sectional Models

hus, both the King and the Simms model evidenced acceptable to good fit, metric invariance, and phi noninvariance. However, when comparing models with BIC, the Simms model evidenced best fit (i.e., lower value) and was therefore chosen for further descriptive analyses. Table I. Fit Statistics for Cross-Sectional Models Model MLR χ2 df CFI RMSEA RMSEA 90% CI BIC T1 (n = 249) DSM-IV 159.09 116 .949 .039 0.022–0.054 12,384.17 King 157.24 113 .948 .040 0.078–0.102 12,388.56 Simms 152.02 113 .954 .038 0.020–0.053 12,383.98 T2 (n = 234) DSM-IV 194.87 116 .927 .055 0.041–0.068 10,674.17 King 188.88 113 .929 .055 0.041–0.068 10,673.22 Simms 183.16 113 .935 .053 0.038–0.066 10,666.66 T3 (n = 203) DSM-IV 235.49 116 .894 .072 0.059–0.086 8,648.31 King 208.25 113 .915 .065 0.051–0.079 8,621.43 Simms 208.05 113 .916 .065 0.051–0.079 8,623.86 Note. DSM-IV, Diagnostic Manual for Mental Disorders—4th edition; MLR, maximum likelihood estimator robust to non-normality and nonindependence; CFI, Bentler comparative fit index; RMSEA, Steiger–Lind root-mean-square error of approximation; CI, confidence interval; BIC, sample size adjusted Bayesian information criteria. Table II. Fit Statistics for Longitudinal Models and Test of Model Invariance

Model MLR χ2 df CFI RMSEA RMSEA 90% CI BIC T1 (n = 249) DSM-IV 159.09 116 .949 .039 0.022–0.054 12,384.17 King 157.24 113 .948 .040 0.078–0.102 12,388.56 Simms 152.02 113 .954 .038 0.020–0.053 12,383.98 T2 (n = 234) DSM-IV 194.87 116 .927 .055 0.041–0.068 10,674.17 King 188.88 113 .929 .055 0.041–0.068 10,673.22 Simms 183.16 113 .935 .053 0.038–0.066 10,666.66 T3 (n = 203) DSM-IV 235.49 116 .894 .072 0.059–0.086 8,648.31 King 208.25 113 .915 .065 0.051–0.079 8,621.43 Simms 208.05 113 .916 .065 0.051–0.079 8,623.86 Note. DSM-IV, Diagnostic Manual for Mental Disorders—4th edition; MLR, maximum likelihood estimator robust to non-normality and nonindependence; CFI, Bentler comparative fit index; RMSEA, Steiger–Lind root-mean-square error of approximation; CI, confidence interval; BIC, sample size adjusted Bayesian information criteria. Table II. Fit Statistics for Longitudinal Models and Test of Model Invariance Model MLR χ2 df S-B Δ χ2 CFI ΔCFI RMSEA RMSEA 90% CI BIC DSM-IV Config. 1,635.35 1,153 .898 .042 0.037–0.046 30,384.90 Metric 1,694.98 1,180 58.83*** .892 −.006 .043 0.038–0.047 30,387.07 Phi 1,704.72 1,184 69.37*** .890 −.008 .043 0.038–0.047 30,387.68 King Config. 1,616.83 1,146 .901 .041 0.037–0.046 30,375.97 Metric 1,652.14 1,172 35.19 .899 −.002 .041 0.037–0.046 30,353.14 Phi 1,677.71 1,178 61.98** .895 −.006 .042 0.037–0.047 30,367.47 Simms Config. 1,604.62 1,146 .903 .041 0.036–0.046 30,365.13 Metric 1,632.67 1,172 28.05 .903 .000 .041 0.036–0.045 30,334.92 Phi 1,661.01 1,178 56.78** .898 −.005 .041 0.037–0.046 30,351.07 Note. All models contain data from all three assessments. DSM-IV, Diagnostic Manual for Mental Disorders—4th edition; MLR, maximum likelihood estimator robust to non-normality and nonindependence; S–B, Satorra–Bentler; CFI, Bentler comparative fit index; RMSEA, Steiger–Lind root-mean-square error of approximation; CI, confidence interval; BIC, sample size adjusted Bayesian information criteria.

Manual for Mental Disorders—4th edition; MLR, maximum likelihood estimator robust to non-normality and nonindependence; S–B, Satorra–Bentler; CFI, Bentler comparative fit index; RMSEA, Steiger–Lind root-mean-square error of approximation; CI, confidence interval; BIC, sample size adjusted Bayesian information criteria. **p < .01. ***p < .001. Table III displays factor loadings for each item and internal consistency for each factor according to the Simms model. Item factor loadings were acceptable in general with values exceeding .50 with the exception of Item 3 (reliving) at T1, Item 6 (avoiding thoughts) at T2, Item 8 (trouble remembering) T1–T3, and Item 12 (future cut short) T1–T3. Internal consistency was acceptable with the exception of the avoidance factor at T1 and T2. Table III. Internal Consistency of Factors and Item Factor Loadings across Assessments

tion of Item 3 (reliving) at T1, Item 6 (avoiding thoughts) at T2, Item 8 (trouble remembering) T1–T3, and Item 12 (future cut short) T1–T3. Internal consistency was acceptable with the exception of the avoidance factor at T1 and T2. Table III. Internal Consistency of Factors and Item Factor Loadings across Assessments T1 T2 T3 PCL—C factors and items α λ α λ α λ Reexperiencing .68 .79 .82 1. Disturbing memories .62 .75 .76 2. Disturbing dreams .52 .50 .63 3. Suddenly reliving .46 .56 .57 4. Upset when reminded .59 .73 .78 5. Physical reactions when reminded .62 .72 .72 Avoidance .47 .49 .73 6. Avoiding thoughts .53 .49 .72 7. Avoiding activities .56 .63 .82 Dysphoria .75 .82 .83 8. Trouble remembering .39 .50 .50 9. Loss of interest .52 .60 .62 10. Feeling distant .58 .65 .65 11. Emotionally numb .54 .62 .66 12. Future cut short .40 .46 .46 13. Trouble sleeping .55 .61 .61 14. Irritable/angry outburst .60 .66 .68 15. Difficulty concentrating .68 .73 .75 Hyperarousal .70 .79 .77 16. Being “super alert” .83 .86 .82 17. Jumpy/easily startled .67 .78 .78 Note. PCL-C, PTSD Checklist—Civilian; T1, 2 weeks after diagnosis; T2, 2 months after diagnosis; T3, 4 months after diagnosis; α, Cronbach's alpha; λ, standardized factor loading.

15. Difficulty concentrating .68 .73 .75 Hyperarousal .70 .79 .77 16. Being “super alert” .83 .86 .82 17. Jumpy/easily startled .67 .78 .78 Note. PCL-C, PTSD Checklist—Civilian; T1, 2 weeks after diagnosis; T2, 2 months after diagnosis; T3, 4 months after diagnosis; α, Cronbach's alpha; λ, standardized factor loading. Table IV presents descriptive statistics for the PCL-C and its subscales according to the Simms model. As is evident, there were significant differences between mothers and fathers in the full scale and all subscales except for the avoidance factor at all assessment points. For mothers, there was a significant main effect of time in the full scale and all subscales except for the avoidance factor. (Full scale: F = 15.57, df = 1.86, p < .001; reexperiencing: F = 5.90, df = 1.63, p < .01; avoidance: F = 2.63, df = 2, p = .08; dysphoria: F = 14.92, df 1.88, p < .001; hyperarousal: F = 15.23, df = 2, p < .001). For fathers, there was a similar pattern with a main effect of time in the full scale and all subscales. (Full scale: F = 24.26, df = 2, p < .001; reexperiencing: F = 23.38, df = 2, p < .001: avoidance: F = 3.31, df = 2, p < .05; dysphoria: F = 8.94, df = 1.87, p < .001; hyperarousal: F = 22.95, df = 1.79, p < .001). At T1, 43% of mothers and 21% of fathers scored above the suggested clinical cutoff (i.e., 44). The corresponding proportions at T2 were 33% for mothers and 19% for fathers, and 28% for mothers and 7% for fathers at T3. Thus, both mothers and fathers evidenced declining levels of PTSS across the assessments; however, a considerable number of parents, mothers especially, scored above the clinical cutoff. Table IV. Descriptive Statistics of PTSS among Participants across Assessments Based on the Simms et al. (2002) Model

fathers at T3. Thus, both mothers and fathers evidenced declining levels of PTSS across the assessments; however, a considerable number of parents, mothers especially, scored above the clinical cutoff. Table IV. Descriptive Statistics of PTSS among Participants across Assessments Based on the Simms et al. (2002) Model T1 T2 T3 PCL-C Mothers M (SD) Fathers M (SD) Estimate Mothers M (SD) Fathers M (SD) Estimate Mothers M (SD) Fathers M (SD) Estimate Full scale 42.06 (12.72) 35.26 (9.59) −6.81*** 39.65 (12.67) 32.82 (10.54) −6.98*** 36.29 (12.76) 29.92 (8.80) −6.68*** Reexperiencing 12.17 (4.50) 10.25 (3.44) −1.91*** 11.68 (4.51) 9.31 (3.51) −2.38*** 10.81 (4.64) 8.14 (2.91) −2.72*** Avoidance 3.79 (1.98) 3.35 (1.69) −0.44 3.49 (1.75) 3.03 (1.39) −0.44* 3.79 (1.98) 3.35 (1.69) −0.44 Dysphoria 20.46 (6.43) 16.90 (4.99) −3.55*** 19.30 (6.21) 16.66 (5.87) −2.71*** 17.65 (6.12) 15.38 (5.10) −2.37*** Hyperarousal 5.56 (2.39) 4.76 (2.00) −0.81** 5.18 (2.23) 3.87 (1.86) −1.34*** 4.52 (2.11) 3.67 (1.52) −0.88*** Note. PCL-C, PTSD Checklist—Civilian; T1, 2 weeks after diagnosis; T2, 2 months after diagnosis; T3, 4 months after diagnosis. *p < .05. **p < .01. ***p < .001.

T1 T2 T3 PCL-C Mothers M (SD) Fathers M (SD) Estimate Mothers M (SD) Fathers M (SD) Estimate Mothers M (SD) Fathers M (SD) Estimate Full scale 42.06 (12.72) 35.26 (9.59) −6.81*** 39.65 (12.67) 32.82 (10.54) −6.98*** 36.29 (12.76) 29.92 (8.80) −6.68*** Reexperiencing 12.17 (4.50) 10.25 (3.44) −1.91*** 11.68 (4.51) 9.31 (3.51) −2.38*** 10.81 (4.64) 8.14 (2.91) −2.72*** Avoidance 3.79 (1.98) 3.35 (1.69) −0.44 3.49 (1.75) 3.03 (1.39) −0.44* 3.79 (1.98) 3.35 (1.69) −0.44 Dysphoria 20.46 (6.43) 16.90 (4.99) −3.55*** 19.30 (6.21) 16.66 (5.87) −2.71*** 17.65 (6.12) 15.38 (5.10) −2.37*** Hyperarousal 5.56 (2.39) 4.76 (2.00) −0.81** 5.18 (2.23) 3.87 (1.86) −1.34*** 4.52 (2.11) 3.67 (1.52) −0.88*** Note. PCL-C, PTSD Checklist—Civilian; T1, 2 weeks after diagnosis; T2, 2 months after diagnosis; T3, 4 months after diagnosis. *p < .05. **p < .01. ***p < .001. Discussion In the current study, we investigated the factor structure of PTSS among parents of children with cancer. We used a longitudinal model-fitting approach based on CFA and tested three models of the underlying dimensions of PTSD and assessed model stability over time. In line with our hypothesis, a four-factor model provided the best fit to the data. Considering a confirmatory factor model including all three assessments, the Simms model evidenced better fit than the DSM-IV model and somewhat better fit than the King model. We therefore decided to choose the Simms model as the best fitting model. This model, comprising the factors reexperiencing, avoidance, dysphoria, and hyperarousal, provided acceptable fit when analyzing data collected 2 weeks, and 2 and 4 months after diagnosis. There was evidence for configural and metric invariance over time, indicating that the basic factor loading pattern and size of loading were equivalent over time, which is in line with previous investigations in other populations (Baschnagel et al., 2005; Elklit & Shevlin, 2007; Krause et al., 2007; Palmieri et al., 2007; Simms et al., 2002). However, it should be noted that we did not find evidence for phi invariance in the longitudinal model, which indicates that factor covariance's seemed to vary across time.

estigations in other populations (Baschnagel et al., 2005; Elklit & Shevlin, 2007; Krause et al., 2007; Palmieri et al., 2007; Simms et al., 2002). However, it should be noted that we did not find evidence for phi invariance in the longitudinal model, which indicates that factor covariance's seemed to vary across time. To our knowledge, this is the first report on the factor structure of PTSS among parents of children with cancer. A particular strength of the current investigation is its longitudinal design allowing for the test of model invariance over time, and the extension of this type of investigation to a new population and language context provides cross-cultural validation of previous findings. However, a notable limitation of the current study is the relatively small sample size. A larger sample size would have enabled a comparison of model fit between mothers and fathers, and future research should investigate model invariance across gender. Another limitation is that measures of general anxiety and depression were not administered, which would have enabled further validation of the construct of PTSD and its underlying factor structure in this context. Future research should include such measures in longitudinal designs of PTSS in parents of children with cancer.

limitation is that measures of general anxiety and depression were not administered, which would have enabled further validation of the construct of PTSD and its underlying factor structure in this context. Future research should include such measures in longitudinal designs of PTSS in parents of children with cancer. In the present study, we did not find support for the conceptualization of the underlying dimensions of PTSD according to the DSM-IV, which proposes three intercorrelated factors reexperiencing (5 items), avoidance/numbing (7), and hyperarousal (5). Instead, when considering all three assessment points, we found best support for an intercorrelated four-factor model proposed by Simms et al. (2002) comprising reexperiencing (5 items), avoidance (2), dysphoria (8), and hyperarousal (2). Compared to the DSM-IV conceptualization, the reexperiencing factor is identical but the two explicit avoidance items are distinguished in a separate factor. Furthermore, five items from the DSM-IV avoidance/numbing factor and three items from the DSM-IV hyperarousal factor are collapsed into a separated factor, which Simms et al. (2002) termed dysphoria. Finally, in the Simms et al. (2002) model, only two items are designated to the factor labeled hyperarousal. The reason for the term dysphoria was that only this factor was highly correlated with measures of depression and general distress, such as generalized anxiety and panic symptoms (Simms et al., 2002). However, it should be noted that the Simms model and the King model provided almost equally good fit to the data and both evidenced metric invariance across time. These findings are also consistent with a recent meta-analytic investigation of the structure of PTSS, aggregating 50 data sets with different samples, which found best support for the Simms et al. (2002) and King et al. (1998) models, with evidence for slightly better fit for the Simms model (Yufik & Simms, 2010).

me. These findings are also consistent with a recent meta-analytic investigation of the structure of PTSS, aggregating 50 data sets with different samples, which found best support for the Simms et al. (2002) and King et al. (1998) models, with evidence for slightly better fit for the Simms model (Yufik & Simms, 2010). According to the results of the present study, the internal consistency of the factors in the Simms model was acceptable with the exception for the avoidance factor, which evidenced poor internal consistency at T1 and T2. The avoidance factor in the Simms model only consist of two items and since internal consistency is strongly linked to the number of items in a given scale (Streiner & Norman, 2008), the current results may have been due to too few items mapping on to this construct. This indicates that more items targeting the phenomenon of avoidance in relation to ones child's serious illness needs to be generated if reliable assessment of this construct is to be ensured. Furthermore, at all assessments items 8 (trouble remembering aspects of trauma) and 12 (sense of future cut short) evidenced poor factor loadings. These items have also produced the poorest factor loadings in previous factor analytic investigations in other populations (e.g., Baschnagel et al., 2005; King et al., 1998; Palmieri & Fitzgerald, 2005). This may of course indicate problems with the current conceptualization of PTSD/PTSS and is also in part consistent with our clinical and research experience using the PCL-C in interviews with the population under investigation, as the item assessing a sense of future cut short often is misunderstood by respondents. The poor factor loadings of trouble remembering aspects of the designated trauma may be especially problematic under the current circumstances since the child's disease actually was ongoing and not a discrete past-oriented event.

, as the item assessing a sense of future cut short often is misunderstood by respondents. The poor factor loadings of trouble remembering aspects of the designated trauma may be especially problematic under the current circumstances since the child's disease actually was ongoing and not a discrete past-oriented event. The current findings with the avoidance factor evidencing poor psychometric properties and several items showing poor factor loadings may be indicative of a more inherent problem of applying measures designed to capture the construct of PTSD to the population of parents of children with serious illness. As outlined by Kazak et al. (2006), the construct of PMTS can be measured with instruments assessing traumatic stress and according to this view the PCL-C could be considered a good option as it maps directly onto the items forming PTSD in the DSM-IV. However, it may be the case that these items do not fully capture the phenomenology of traumatic stress reactions of parents of children with cancer. Future research is needed to determine whether there is a need for a new operationalization of PMTS to better assess this construct among parents of children with cancer.

However, it may be the case that these items do not fully capture the phenomenology of traumatic stress reactions of parents of children with cancer. Future research is needed to determine whether there is a need for a new operationalization of PMTS to better assess this construct among parents of children with cancer. Both mothers and fathers evidenced declining levels of PTSS during their child's treatment, which is in line with previous longitudinal investigations (e.g., Dolgin et al., 2007; Steele, Long, Reddy, Luhr & Phipps, 2003). A considerable number of individuals, especially mothers, scored above the suggested cutoff. This suggest that tailored intervention based on individual distress levels among parents of children with cancer might be warranted.

tudinal investigations (e.g., Dolgin et al., 2007; Steele, Long, Reddy, Luhr & Phipps, 2003). A considerable number of individuals, especially mothers, scored above the suggested cutoff. This suggest that tailored intervention based on individual distress levels among parents of children with cancer might be warranted. Evidence-based assessment is an integral part of research and practice in pediatric psychology (Kazak et al., 2007) and construct validation is an important aspect of measurement development and their use in clinical practice (Holmbeck & Devine, 2009; Streiner & Norman, 2008). Establishing valid factor models and measurement invariance is an important part of measurement development and practical use as it allows for cross-group comparisons of parameters such as means and regression coefficients. Furthermore, establishing a valid model of the underlying dimensions of PTSD among parents of children with cancer could allow for the investigation of how symptom clusters (i.e., factors) of PTSS are related to each other over time, which in turn could enhance interventions aiming to alleviate PTSS in this population. We see the present analysis as a first step in determining the best fitting model of PTSS in parents of children undergoing cancer treatment, and our results tentatively suggest using a four-factor model in favor of the DSM-IV three-factor model. However, constructive replication of the current results is needed before firm conclusions can be drawn regarding which model researchers and clinicians should use when assessing PTSS in this population.

treatment, and our results tentatively suggest using a four-factor model in favor of the DSM-IV three-factor model. However, constructive replication of the current results is needed before firm conclusions can be drawn regarding which model researchers and clinicians should use when assessing PTSS in this population. Funding The Swedish Research Council grant K2008-70X-20836-01-3 and The Swedish Childhood Cancer Foundation grants PROJ01/005, PROJ02/004, PROJ05/030. Conflicts of Interest: None declared. Acknowledgment The authors thank Susanne Lorenz and Ulrika Pöder for their work with data collection, and Filip Arnberg for thoughtful comments on an earlier draft of this article. *Data presented in this study have in part been published previously in Pöder et al. (2008), Lindahl Norberg et al. (2011), and Pöder et al. (2010). Data not presented in this study but from the same sample have also been published previously in Pöder and von Essen (2009).

Introduction Caring for children with special health care needs (SHCN) may take a toll on parental health, divert attention from typical aspects of family functioning, and can also influence possibilities for participation in paid employment (DeRigne, 2012; Hauge et al., 2013; Reichman, Corman, & Noonan, 2008). Children with SHCN often have a substantial need for professional medical care and are typically at risk of having chronic physical, developmental, behavioral, or emotional conditions that require health-related services of a type or amount beyond that required by typically developing children of similar age (McPherson et al., 1998; Perrin, 2002). Caring for a child with SHCN can be an enormous responsibility and can far exceed the demands of the typical caregiver role (Raina et al., 2004). This caregiver role is typically also not chosen or planned, and preparation for and adjustment to the role will most often need to occur once it has already been acquired.

. Caring for a child with SHCN can be an enormous responsibility and can far exceed the demands of the typical caregiver role (Raina et al., 2004). This caregiver role is typically also not chosen or planned, and preparation for and adjustment to the role will most often need to occur once it has already been acquired. The care demands associated with raising a child with SHCN are typically both highly intensive and long-lasting, and tend to fall more heavily on mothers as compared with fathers (Crowe & Michael, 2011; Curran, Sharples, White, & Knapp, 2001; Tadema & Vlaskamp, 2010). High levels of child-related stress, constant or returning worries about the child’s condition, and sorrow arising from the “loss” of an expected healthy child may also compromise the caregiver’s health. In terms of mental health, mothers of children with SHCN have been found to display elevated levels of depressive symptoms as compared with both fathers and mothers raising children without SHCN, with symptom levels often remaining high over time (Brehaut et al., 2009; Kuhlthau, Kahn, Hill, Gnanasekaran, & Ettner, 2010; Nes et al., 2014; Olsson & Hwang, 2001, 2006; Resch, Elliott, & Benz, 2012). Such mental health impairments often arise from the chronic strains involved in the caregiver role, as well as from emotional reactions evoked and sustained by the child’s condition (Raina et al., 2004). As such, mothers caring for children with the most severe conditions are commonly also those most affected (Churchill, Villareale, Monaghan, Sharp, & Kieckhefer, 2010).

chronic strains involved in the caregiver role, as well as from emotional reactions evoked and sustained by the child’s condition (Raina et al., 2004). As such, mothers caring for children with the most severe conditions are commonly also those most affected (Churchill, Villareale, Monaghan, Sharp, & Kieckhefer, 2010). The additional care demands associated with raising a child with SHCN may influence parents in other vital areas as well. In terms of work participation following the birth of a child with SHCN, participation levels among mothers are commonly more affected than those of fathers (Parish & Cloud, 2006). Early parenthood and the responsibility of caring for young children in general may require adjustments to parental work participation, and women especially tend to reduce work hours or temporarily stop working when their children are young (Bø, Kitterød, Køber, Nerland, & Skoglund, 2008). Implementation of comprehensive family and equality policies, such as the Norwegian Kindergarten Act, has enabled women especially to continue participation in paid work and to pursue their occupational careers during the early years of motherhood (Norwegian Ministry of Education and Research, 2005). With close to 90% of children attending kindergarten (Statistics Norway, 2011), women with young children have steadily increased their employment levels since the 1990s to a current rate of about 73%, a rate increasing further as their children age (Bø et al., 2008). Moreover, with an employment rate exceeding 80% for women aged ≥25 years, the overall employment levels among Norwegian women of childbearing age is high and higher than in most The Organisation for Economic Co-operation and Development (OECD) countries (Statistics Norway, 2011). As unemployment is low in Norway, well below the OECD average (OECD, 2011), most nonemployed women are out of paid work due to other reasons than unemployment or lack of available child care.

h and higher than in most The Organisation for Economic Co-operation and Development (OECD) countries (Statistics Norway, 2011). As unemployment is low in Norway, well below the OECD average (OECD, 2011), most nonemployed women are out of paid work due to other reasons than unemployment or lack of available child care. Despite work participation being high in Norway, maternal employment opportunities appear rather restricted when caring for children with SHCN (Hauge et al., 2013). Many mothers of children with SHCN report having missed days from work, having cut work hours, or having left employment altogether, due to their children’s additional health care needs (DeRigne & Porterfield, 2010; Hedov, Wikblad, & Anneren, 2006; Porterfield, 2002). Apart from its obvious financial aspects, employment provides additional benefits such as social inclusion and appreciation by others, and may also reduce feelings of isolation and peripherality (Shearn & Todd, 2000). However, the inability to properly meet employment demands while providing optimal care for their children may necessitate shorter or longer employment adjustments for many mothers. The additional care demands associated with raising a child with SHCN may thus prevent mothers at risk of mental health problems from using the likely beneficial respite effects of employment (Morris, 2012).

iding optimal care for their children may necessitate shorter or longer employment adjustments for many mothers. The additional care demands associated with raising a child with SHCN may thus prevent mothers at risk of mental health problems from using the likely beneficial respite effects of employment (Morris, 2012). Work impairment due to mental health problems in general is a considerable and increasing public health problem in many Western countries. After musculoskeletal disorders, psychiatric disorders (PD) are now the most common diagnostic group reported by physicians on sick leave certificates (Hensing & Wahlström, 2004). Recovery and return to work after a sick leave due to PD generally takes longer than absence due to other conditions, and many long-term absentees do not recover and end up on permanent disability pensions (Bratberg, Gjesdal, & Mæland, 2009; Henderson, Glozier, & Holland Elliott, 2005). Apart from studies showing elevated levels of depressive symptoms and an increased risk of reducing or leaving employment altogether, little is known about how the additional care demands associated with raising a child with SHCN may impact on shorter or longer work absences due to PD among mothers who remain in the work force. Population-based research on employment-related consequences of caring for children with SHCN is scant, and longitudinal research based on data other than parental report is needed (Reichman et al., 2008). In addition, most studies on mental health problems in caregivers of children with SHCN rely on small samples, are often recruited in clinical settings, and may thus lack generalizability to the wider population (Resch et al., 2012).

l research based on data other than parental report is needed (Reichman et al., 2008). In addition, most studies on mental health problems in caregivers of children with SHCN rely on small samples, are often recruited in clinical settings, and may thus lack generalizability to the wider population (Resch et al., 2012). To address some of the abovementioned shortcomings, this study aimed to explore associations between caring for a child with SHCN and sick leave due to PD during the early years of motherhood. Self-report data from a large Norwegian population-based birth cohort was applied and linked with national registry-based data on physician-certified sick leave and relevant background factors associated with both sick leave and maternal employment status more generally. Based on the previous literature, we expected the risk of sick leave due to PD in general and due to depression more specifically to be higher among mothers of children with SHCN as compared with mothers of typically developing children during early motherhood. We further hypothesized that the mothers’ risk of sick leave due to PD would increase with the severity of the child’s care needs.

to PD in general and due to depression more specifically to be higher among mothers of children with SHCN as compared with mothers of typically developing children during early motherhood. We further hypothesized that the mothers’ risk of sick leave due to PD would increase with the severity of the child’s care needs. Methods Study Sample The study population included participants in the population-based Norwegian Mother and Child Cohort Study (MoBa), conducted by the Norwegian Institute of Public Health (Magnus et al., 2006). Pregnant women were recruited at their first routine ultrasound examination at weeks 17–18 of gestation between 1999 and 2008 (response rate 40.6%), and the cohort includes approximately 90,000 unique observations of expectant mothers, as participants with about 107,000 pregnancies in total among them (Nilsen et al., 2009). The MoBa cohort is linked to the Medical Birth Registry of Norway (Irgens, 2000), which contains the national identification number for all participants in the study, allowing linkage with the Central Population Register, benefit registries from the Norwegian Labour and Welfare Administration, and the education and income registries of Statistics Norway. This linkage provided longitudinal data with annual updates for both mothers and their children throughout 2010. For the present study, only mothers with children aged ≥4 years by the end of 2010 were considered for inclusion and a total of 66,211 cases were found eligible. Among eligible cases, we excluded cases where the mother had emigrated or where either the mother or the child had died by the time of follow-up. Norwegian sick leave regulations are complex and dependent on employment status, income level, as well as receipt of other social benefits (Norwegian Labour and Welfare Administration, 2015). Therefore, as sickness benefit is granted to compensate for a temporary loss of income from employment while on sick leave, we also excluded cases for which the mothers were not considered to be at risk of sick leave 1 year after childbirth, that is, participants not active in the work force at the time of childbirth. This latter group included mothers with an income from employment below the limit entitling them to sickness benefit, in addition to mothers granted disability pension before the start of follow-up, leaving a sample of 58,532 mothers and children for the analyses. The study was approved by the Regional Committee for Medical Research Ethics in south-eastern Norway.

ome from employment below the limit entitling them to sickness benefit, in addition to mothers granted disability pension before the start of follow-up, leaving a sample of 58,532 mothers and children for the analyses. The study was approved by the Regional Committee for Medical Research Ethics in south-eastern Norway. Outcomes and Study Variables The study outcome was physician-certified sick leave due to PD. In Norway, sickness benefit is granted to compensate for loss of income for employed members of the National Insurance Scheme who are temporarily occupationally disabled due to an illness or injury. The benefit is connected to employment status, and economic compensation equivalent to the individual’s employment income is given from the first day of a sick leave period to all employees with an income exceeding the limit entitling them to sickness benefit (i.e., approximately 4,600 € in 2010). Employers are obliged to compensate for the first 16 days of a sick leave period, and a certificate from a physician evaluating whether there are significant medical reasons for an absence from work is required after 3 days of absence. All periods of sick leave exceeding 16 days for up to 1 year are compensated for and recorded, including the main diagnosis, by the Norwegian Labour and Welfare Administration, with diagnostic information according to the International Classification of Primary Care (ICPC-2). Based on information on sickness benefit obtained from the registry, three measures reflecting sick leave due to PD during follow-up were constructed. First, an indicator for any sick leave exceeding 16 days with the ICPC codes P01–P99 was constructed (i.e., any psychiatric disorder). Second, an indicator for a long-term leave (i.e., continuous absence exceeding 8 weeks) with the ICPC codes P01–P99 was constructed, and third, an indicator reflecting a long-term leave due to depression (i.e., ICPC P03 or P76) was constructed. Any sick leave for conditions other than the ICPC codes P01–P99 was ignored.

econd, an indicator for a long-term leave (i.e., continuous absence exceeding 8 weeks) with the ICPC codes P01–P99 was constructed, and third, an indicator reflecting a long-term leave due to depression (i.e., ICPC P03 or P76) was constructed. Any sick leave for conditions other than the ICPC codes P01–P99 was ignored. The main exposure was early childhood SHCN, assessed as receipt of attendance benefit by the age of 3 years. Attendance benefit is a universally accessible benefit provided by the Norwegian Labour and Welfare Administration to compensate for domestic care-related expenses. The benefit may be granted to children with a medically documented need for special care and supervision due to illness, injury, or congenital disabilities. To be eligible for the benefit, the care has to be provided in a private care setting and is granted for persons who are not able to cope without supervision or who need help in performing activities of daily living. Attendance benefit is granted solely based on the health care needs of the recipient and is not dependent on the financial situation of the recipient or the family. The benefit is granted to children who have care needs well exceeding those common to otherwise healthy children of comparable age, and about 2–4% of children <18 years of age receive attendance benefit. The most common diagnoses among recipients include endocrine and neurological diseases, asthma, congenital malformations, and mental conditions, most of which are conditions that have an early onset and last over a prolonged period of time (Bjerkedal, Kristensen, Skjeret, & Brevik, 2006; Sletvold & Rendedal, 2004). Based on the degree to which the condition impairs the child’s physical or psychological functional ability, and how demanding the care arrangement is for the parents, higher rate benefit may be granted to children whose need for care and supervision is considerably greater than that covered by ordinary attendance benefit. Ordinary benefit at rate 1 reflects mild care needs, while rate 2 reflects moderate care needs, and rates 3 and 4 reflect severe care needs. Moderate care needs entitle the recipient to a benefit twice as large as that for mild care needs, whereas benefit entitlement for severe care needs is four to six times that for mild care needs. Congenital malformations and neurological and respiratory diseases are common among recipients of higher rate attendance benefit.

ate care needs entitle the recipient to a benefit twice as large as that for mild care needs, whereas benefit entitlement for severe care needs is four to six times that for mild care needs. Congenital malformations and neurological and respiratory diseases are common among recipients of higher rate attendance benefit. Due to relatively low numbers of children receiving benefits for the most severe care needs, rates 2–4 were merged for the analyses.

ate care needs entitle the recipient to a benefit twice as large as that for mild care needs, whereas benefit entitlement for severe care needs is four to six times that for mild care needs. Congenital malformations and neurological and respiratory diseases are common among recipients of higher rate attendance benefit. Due to relatively low numbers of children receiving benefits for the most severe care needs, rates 2–4 were merged for the analyses. Information on factors commonly associated with both sick leave and with maternal employment status more generally were included and adjusted for in the analyses (Allebeck & Mastekaasa, 2004; Hensing & Wahlström, 2004). The Medical Birth Registry of Norway provided data on the mothers’ age and marital status at the time of childbirth, while the Central Population Register provided data on the mothers’ number of children <6 years of age by the end of the year of childbirth. Data on educational attainment was obtained from the National Education Database of Statistics Norway, and the mothers’ highest level of attainment at the time of childbirth was categorized as below high school graduate, high school graduate, lower college or university level, and higher college or university level, including postgraduate levels. At weeks 17–18 of gestation, the expectant mothers were asked to complete a five-item version of the Hopkins Symptom Checklist (SCL-5), reflecting susceptibility to anxiety and depression. The SCL-5 has been shown to perform similarly to the long version and is suitable for detecting psychological problems in a nonpsychiatric setting (Tambs & Moum, 1993). The mothers were asked to indicate on a 4-point scale if, during the past 2 weeks, they had been bothered: (1) not at all, (2) a little, (3) quite a bit, or (4) very much by problems such as “Feeling blue” and “Worrying too much about things.” Cronbach’s α was .79 in the current sample and an average item score >2.0 was used as a clinical cutoff for psychological distress according to convention (Strand, Dalgard, Tambs, & Rognerud, 2003).

(2) a little, (3) quite a bit, or (4) very much by problems such as “Feeling blue” and “Worrying too much about things.” Cronbach’s α was .79 in the current sample and an average item score >2.0 was used as a clinical cutoff for psychological distress according to convention (Strand, Dalgard, Tambs, & Rognerud, 2003). Statistical Analysis To take into account variation in the year and month of childbirth among participants in the MoBa cohort and to ensure equal and comparable time at risk for all participants in the study, all periods of sick leave due to PD were rescaled into its corresponding month following childbirth. In Norway, parental leave is connected to employment and a parent is entitled to parental benefit and leave from work during the child’s first year of life if he or she has been gainfully employed with a pensionable income for at least 6 of the 10 months before the benefit period (Nordic Council of Ministers, 2011). The start of follow-up was thus set to the month the child turned 1 year of age for all participants, and in cases of no sick leave due to PD, follow-up lasted until the month the child turned 4 years of age. For cases with a sick leave due to PD, censoring occurred (i.e., participants left the risk pool) at the respective month of any first sick leave due to PD, a long-term leave due to PD, or a long-term leave due to depression. In addition, mothers being granted disability pension, and who therefore left the work force during the follow-up period, were no longer considered to be at risk of sick leave and were censored at the respective month of receipt of disability pension (equalling 0.6% of the sample). Analyses were performed using Stata/SE version 12.1 (StataCorp, 2011).

d disability pension, and who therefore left the work force during the follow-up period, were no longer considered to be at risk of sick leave and were censored at the respective month of receipt of disability pension (equalling 0.6% of the sample). Analyses were performed using Stata/SE version 12.1 (StataCorp, 2011). Descriptive analyses were performed to assess demographic differences between mothers of children with and without SHCN on the covariate factors included in all models estimated (i.e., maternal age, educational attainment, marital status, number of preschoolers, and psychological distress). Hazard ratios (HR) with 95% confidence intervals (CI) to reflect participants’ risk of a sick leave due to PD were computed using the Cox proportional hazard model, adjusted for the covariate factors listed above. The rate of sick leave among mothers of children without SHCN was used as the reference for which to compare the rates of sick leave among mothers of children with mild and moderate/severe care needs, respectively. The HR is an outcome measure in time-to-event analysis, and an HR of 1 indicates that event rates are equal across respective groups, while for instance an HR of 2 indicates that twice as many in the group in question experience the event as in the reference group. If the range of the corresponding 95% CI does not contain the value 1, there is a significant difference in risk at the p < .05 level between the reference group and the comparison group in question.

nstance an HR of 2 indicates that twice as many in the group in question experience the event as in the reference group. If the range of the corresponding 95% CI does not contain the value 1, there is a significant difference in risk at the p < .05 level between the reference group and the comparison group in question. Results Assessed as receipt of attendance benefit by 3 years of age, a total of 1.7% of the women in this population-based sample were mothers of children with medically documented SHCN, of whom approximately half were mothers of children with mild care needs, while the other half were mothers of children with moderate and severe care needs. Close to 50% of the children were granted attendance benefit already by 1 year of age, and close to 80% were granted the benefit by 2 years of age. Receipt of attendance benefit at an early age was most common among children with moderate and severe care needs. Descriptive analyses were performed to examine demographic differences between mothers of children with and without SHCN on the covariate adjustment factors, all assessed before the start of follow-up. No significant differences were found for either maternal age or marital status at the time of childbirth, while a significantly larger proportion of mothers of children without SHCN had completed education at a college or university level (68.7%) and were having their first child (60.7%) as compared with mothers of children with SHCN (62.3% [χ2 18.8; p < .01] and 56.7% [χ2 6.4; p < .01], respectively). In addition, a somewhat larger proportion of mothers of children with SHCN reported psychological distress during early pregnancy (15.1%) as did mothers of children without SHCN (10.2% [χ2 25.4; p < .01]).

compared with mothers of children with SHCN (62.3% [χ2 18.8; p < .01] and 56.7% [χ2 6.4; p < .01], respectively). In addition, a somewhat larger proportion of mothers of children with SHCN reported psychological distress during early pregnancy (15.1%) as did mothers of children without SHCN (10.2% [χ2 25.4; p < .01]). A considerable proportion of the mothers in this population-based sample had at least one sick leave due to PD during 1–4 years after childbirth. Moreover, a consistent association between the severity of the child’s health care needs and the mother’s risk of being absent from work due to PD was evident. Whereas close to 11% of mothers of children without any documented SHCN had a sick leave due to PD during the follow-up period, the corresponding proportions among mothers of children with mild and moderate/severe care needs were about 16% and 21%, respectively (Table I). This trend was even more evident for long-term sick leaves due to PD in general and due to depression more specifically. Mothers of children with mild and moderate/severe care needs had approximately twice the amount of physician-certified sick leave lasting for ≥8 weeks as compared with mothers of children without SHCN for both long-term measures. Table I. Percentages and Adjusted Hazard Ratios for Maternal Sick Leave Due to Psychiatric Disorders 1–4 Years Following Childbirth

e/severe care needs had approximately twice the amount of physician-certified sick leave lasting for ≥8 weeks as compared with mothers of children without SHCN for both long-term measures. Table I. Percentages and Adjusted Hazard Ratios for Maternal Sick Leave Due to Psychiatric Disorders 1–4 Years Following Childbirth Number of observations Short-term sick leave (<8 weeks) Long-term sick leave (≥8 weeks) Any psychiatric disorder (ICPC P01–P99) Any psychiatric disorder (ICPC P01–P99) Depression (ICPC P03 or P76) % sick leave Hazard ratio 95% confidence interval % sick leave Hazard ratio 95% confidence interval % sick leave Hazard ratio 95% confidence interval Total 58,532 10.9 6.6 3.2 Child SHCN None 57,354 10.7 1.00 Reference 6.5 1.00 Reference 3.1 1.00 Reference Mild 495 16.2 1.47 [1.18–1.83] 11.5 1.71 [1.31–2.22] 6.5 1.93 [1.36–2.74] Moderate/severe 503 21.1 1.99 [1.64–2.41] 13.5 2.05 [1.62–2.61] 6.8 2.05 [1.46–2.88] Maternal age ≤24 years 5,719 11.4 1.00 Reference 6.3 1.00 Reference 3.4 1.00 Reference 25–29 years 19,799 10.9 1.07 [0.97–1.17] 6.4 1.16 [1.03–1.31] 3.1 1.12 [0.95–1.33] 30–34 years 23,047 10.5 1.04 [0.95–1.14] 6.5 1.21 [1.07–1.37] 3.0 1.12 [0.94–1.32] ≥35 years 9,967 11.3 1.12 [1.01–1.24] 7.3 1.35 [1.19–1.54] 3.5 1.30 [1.08–1.56] Educational attainment <High school graduate 4,252 13.1 1.42 [1.27–1.59] 8.1 1.44 [1.25–1.66] 4.1 1.62 [1.32–1.99] High school graduate 14,108 11.9 1.33 [1.22–1.45] 7.3 1.39 [1.24–1.55] 3.9 1.64 [1.39–1.93] Lower college/university 31,391 10.7 1.22 [1.13–1.32] 6.4 1.22 [1.11–1.35] 3.0 1.30 [1.12–1.51] Higher college/university 8,781 8.8 1.00 Reference 5.4 1.00 Reference 2.3 1.00 Reference Marital status Married/cohabiting 56,762 10.8 1.00 Reference 6.5 1.00 Reference 3.1 1.00 Reference Single 1,770 13.9 1.13 [0.99–1.29] 9.3 1.24 [1.05–1.45] 5.0 1.30 [1.04–1.61] Number of preschoolers One 35,463 10.6 1.00 Reference 6.4 1.00 Reference 3.1 1.00 Reference Two 20,940 11.2 1.08 [1.02–1.13] 6.8 1.05 [0.99–1.13] 3.2 1.04 [0.94–1.14] Three or more 2,129 11.9 1.16 [1.02–1.32] 7.4 1.16 [0.99–1.37] 3.5 1.17 [0.92–1.48] Psychological distress No 52,492 9.8 1.00 Reference 5.9 1.00 Reference 2.7 1.00 Reference Yes 6,040 19.7 2.08 [1.95–2.22] 12.8 2.23 [2.06–2.41] 7.1 2.55 [2.28–2.84] Note. All estimates adjusted for other variables in respective models.

Three or more 2,129 11.9 1.16 [1.02–1.32] 7.4 1.16 [0.99–1.37] 3.5 1.17 [0.92–1.48] Psychological distress No 52,492 9.8 1.00 Reference 5.9 1.00 Reference 2.7 1.00 Reference Yes 6,040 19.7 2.08 [1.95–2.22] 12.8 2.23 [2.06–2.41] 7.1 2.55 [2.28–2.84] Note. All estimates adjusted for other variables in respective models. ICPC = international classification of primary care; SHCN = special health care needs.

Three or more 2,129 11.9 1.16 [1.02–1.32] 7.4 1.16 [0.99–1.37] 3.5 1.17 [0.92–1.48] Psychological distress No 52,492 9.8 1.00 Reference 5.9 1.00 Reference 2.7 1.00 Reference Yes 6,040 19.7 2.08 [1.95–2.22] 12.8 2.23 [2.06–2.41] 7.1 2.55 [2.28–2.84] Note. All estimates adjusted for other variables in respective models. ICPC = international classification of primary care; SHCN = special health care needs. After adjustment for factors commonly associated with sick leave and maternal employment status more generally, the mothers’ risks of being absent from work due to PD were strong and consistent for both short-term and long-term absences. Apart from the effects of educational attainment and self-reported psychological distress, which were both associated with a substantial increased risk of sick leave, the mothers’ age, marital status, and number of preschoolers had only limited effects on their risk of being absent from work due to PD. In addition to an increased risk of sick leave due to any PD (HR: 1.47; 95% CI: 1.18–1.83; HR: 1.99; 95% CI: 1.64–2.41, respectively), mothers of children with mild (HR: 1.71; 95% CI: 1.31–2.22) and moderate/severe care needs (HR: 2.05; 95% CI: 1.62–2.61) both had a substantial risk of being long-term absent from work due to any PD in general and due to depression more specifically (HR: 1.93; 95% CI: 1.36–2.74; HR: 2.05; 95% CI: 1.46–2.88, respectively) compared with mothers of children without SHCN. Thus, the risk of sick leave due to PD following the birth of a child with SHCN was substantial and demonstrates that children’s SHCN constitute an important prospective factor for mental health problems in maternal caregivers during the early years of motherhood.

tively) compared with mothers of children without SHCN. Thus, the risk of sick leave due to PD following the birth of a child with SHCN was substantial and demonstrates that children’s SHCN constitute an important prospective factor for mental health problems in maternal caregivers during the early years of motherhood. Discussion The findings of this population-based study show that mental health impairments are common among mothers of children with SHCN. Assessed as being granted sickness benefit due to PD during early motherhood, their mental health was significantly poorer than that of mothers of healthy children of similar age for all outcomes examined. Additional childhood care needs were related to an increased risk of both short-term and long-term sick leave due to PD, evident for mothers of children with mild and moderate/severe care needs alike. The risk of sick leave was strong also after adjustment for important factors such as self-reported susceptibility to anxiety and depression, in this study assessed before any knowledge of children’s SHCN. Self-reported susceptibility to psychological distress has previously been shown to be strongly related to long-term sick leave due to PD in both men and women (Foss et al., 2010). The finding that the excess risk of being absent from work due to PD remained strong in the adjusted models indicates that the care demands and child-related stress experienced by many mothers of children with SHCN may have a profound impact on their mental health. Our findings thus concur with findings from previous studies on caregiver health following the birth of a child with SHCN (Brehaut et al., 2009; Olsson & Hwang, 2006; Resch et al., 2012). As children with additional needs often require care and assistance over an extended period of time in which otherwise healthy children become gradually more independent, their mothers may therefore be at prolonged risk of mental health problems. The often long-lasting care responsibilities associated with raising children with SHCN have been associated with depressive symptoms in caregivers well beyond the early years of motherhood (Rosenthal, Learned, Liu, & Weitzman, 2013).

ependent, their mothers may therefore be at prolonged risk of mental health problems. The often long-lasting care responsibilities associated with raising children with SHCN have been associated with depressive symptoms in caregivers well beyond the early years of motherhood (Rosenthal, Learned, Liu, & Weitzman, 2013). Because recovery from psychiatric conditions generally takes longer than recovery from other conditions (Henderson et al., 2005), many caregivers may thus be prevented from participating in regular employment and from using the possible respite effects of employment for a prolonged period of time (Gordon, Cuskelly, & Rosenman, 2008; Morris, 2012).

tric conditions generally takes longer than recovery from other conditions (Henderson et al., 2005), many caregivers may thus be prevented from participating in regular employment and from using the possible respite effects of employment for a prolonged period of time (Gordon, Cuskelly, & Rosenman, 2008; Morris, 2012). To our knowledge, this is the first study to investigate associations between children’s SHCN and mothers’ risk of sick leave due to PD. Application of a large population-based sample with longitudinal register-based data constitutes a major strength, and ensures complete follow-up of all eligible participants who were considered to be at risk of sick leave 1 year after childbirth (i.e., all gainfully employed with an income entitling them to sickness benefit). Such register-based studies are a powerful alternative to traditional longitudinal approaches, which often suffer from a large loss to follow-up and of systematic attrition (Wadsworth et al., 2003). Moreover, the data obtained from national registers provide valid objective measurements of factors previously assessed mostly through parental self-report, evident for both sick leave and for children’s SHCN. As such, the use of data on sick leave with physician-certified diagnostic information in accordance with the established ICPC-2 coding system constitutes a considerable strength. Although sick leave with PD and its corresponding ICPC-2 diagnoses reflect psychiatric health problems resulting in lowered work capacity only among those employed, work participation among Norwegian women is high during early motherhood (Bø et al., 2008), including that of mothers of young children with SHCN (Hauge et al., 2013). Still, it may be that the extent of sick leave due to PD is underestimated due to the stigma associated with PD in general. However, as diagnoses are confidential and as knowledge of PD has increased while the stigma associated with these conditions has decreased, physician-certified sick leave is likely a more valid measurement than traditional self-reports, which may be prone to both selection and reporting bias (Bratberg et al., 2009; Stansfeld et al., 1995).

are confidential and as knowledge of PD has increased while the stigma associated with these conditions has decreased, physician-certified sick leave is likely a more valid measurement than traditional self-reports, which may be prone to both selection and reporting bias (Bratberg et al., 2009; Stansfeld et al., 1995). Moreover, while parents know best the specific care needs of their children and parental report has been shown to be fairly reliable for severe conditions, parental report of children’s SHCN may be subject to both response and recall bias that can invalidate findings (Brehaut et al., 2009). Although several studies have investigated employment-related consequences of specific conditions such as asthma and autism, variation in severity among individual cases may be great and need not reflect the actual care burden for the parents (van Dyck, Kogan, McPherson, Weissman, & Newacheck, 2004). Applying a medically documented assessment of the extent of the child’s health care needs, relative to healthy children of similar age, may therefore better reflect the additional parental care burden associated with raising a child with SHCN. Attendance benefit is universally accessible and no children with documented needs, cared for within a private care setting, are left out of the benefit. Although attendance benefit is granted to compensate for domestic expenses related to the child’s additional care needs, not even the maximum amount granted for the most severe conditions will compensate for a loss of regular income. As such, it is not likely that attendance benefit alone is an incentive for mothers to reduce work hours or leave paid work altogether. Rather, it is likely that mothers who cannot remain in regular employment due to children’s SHCN will gradually experience a more constrained economic situation.

oss of regular income. As such, it is not likely that attendance benefit alone is an incentive for mothers to reduce work hours or leave paid work altogether. Rather, it is likely that mothers who cannot remain in regular employment due to children’s SHCN will gradually experience a more constrained economic situation. Some limitations of the present study also need to be acknowledged. Although not uncommon for large epidemiological studies (Hartge, 2006), the response rate in the MoBa cohort is lower than optimal. Self-selection to the study may thus result in deviations from the larger population from which the women were sampled. Comparisons of cohort participants with all women giving birth in Norway during the same period identified several deviations in prevalence estimates, notably the underrepresentation of women <25 years of age (Nilsen et al., 2009). However, as young women are less likely to be stably employed in the labor market (OECD, 2011), they are also less likely to have an income from employment exceeding the limit entitling them to sickness benefit and thus of being eligible for the present study. As we adjusted for age and as deviations between the MoBa cohort and the general population mostly reflect differences in prevalence estimates rather than exposure-outcome associations (Nilsen et al., 2009), selection bias is not likely to have threatened the validity of associations reported herein. In addition, as consent to participate in the study was obtained during early pregnancy before any knowledge of children’s SHCN, and as the data obtained from national registries ensures complete follow-up of all participants considered eligible for sickness benefit, bias due to possible systematic attrition is also reduced. Moreover, the study of mental health impairments following the birth of a child with SHCN based on the mother’s use of sick leave is restricted to those in employment only. Consequently, we are not able to observe the mental health in mothers not participating in the labor market. Although many mothers withdraw from the labor market following the birth of a child with SHCN, most do not, and temporary withdrawal from the labor market during early motherhood is common among mothers of children without SHCN alike (Bø et al., 2008; Hauge et al., 2013). Thus, our findings on mental health impairments following childbirth are not likely to be systematically biased based on the mothers’ withdrawal from the labor market.

y withdrawal from the labor market during early motherhood is common among mothers of children without SHCN alike (Bø et al., 2008; Hauge et al., 2013). Thus, our findings on mental health impairments following childbirth are not likely to be systematically biased based on the mothers’ withdrawal from the labor market. Another potential limitation that deserves some attention is that our exposure measure of children’s SHCN yields receipt of attendance benefit by age 3 years only. As such, it is possible that some children are not identified as having SHCN by the end of the follow-up period, while in reality having care needs well exceeding those of otherwise healthy children. However, as receipt of attendance benefit is based on a medically documented need for special care and supervision, the special needs of children that may affect their mothers’ risk of sick leave due to PD will likely precede being granted the benefit. As approximately 80% of the children receiving attendance benefit in this study were granted the benefit already by 2 years of age, children’s SHCN is likely to influence mothers’ mental health, necessitating prolonged work absences already from an early age. Moreover, although the study included a sizeable sample of Norwegian mothers, too few of their children were nonetheless identified as having moderate and severe care needs for them to be analyzed as distinct categories. As the severity of children’s SHCN has previously been shown to be associated with maternal employment-related outcomes (Hauge et al., 2013), it is possible that its effect on mothers’ use of sick leave is somewhat attenuated for the most severe conditions. However, as mothers of children with mild and moderate/severe care needs had an increased risk of sick leave due to PD alike, their risk was nevertheless different from that of mothers of children without SHCN during the early years of motherhood.

use of sick leave is somewhat attenuated for the most severe conditions. However, as mothers of children with mild and moderate/severe care needs had an increased risk of sick leave due to PD alike, their risk was nevertheless different from that of mothers of children without SHCN during the early years of motherhood. Notwithstanding the limitations outlined above, the present study provided reliable evidence that caring for children with SHCN can indeed take a toll on the mental health of many caregivers. The consistent finding that mothers of children with both mild and more severe health care needs are more often long-term absent from work due to PD in general and due to depression more specifically, indicates that their often heavier-than-average caregiving burdens can severely impact their own health and may possibly impact on the overall welfare of the entire family in the long run (Reichman et al., 2008). Prolonged work absences due to their own or their children’s health impairments can lead to feelings of isolation, erosion of qualifications, and lowered self-esteem for this group of mothers, making return to work increasingly more difficult and increasing the mothers’ risk of ending up on permanent disability pensions (Bratberg et al., 2009; Shearn & Todd, 2000). Inclusion of employees faced by such challenges in more adapted and flexible work arrangements might hinder some mothers of children with SHCN from dropping out of employment altogether. Furthermore, mental health impairments need not only affect the mothers themselves. Maternal depression has been shown to be related also to child adjustment problems across a wide age range (Downey & Coyne, 1990; Goodman et al., 2011). Children with SHCN are likely to be particularly vulnerable to adverse effects of their mothers’ mental health impairments, with maternal depression constituting an additional source of disadvantage faced by this group of children who are already at risk. The psychological, behavioral, and economic implications outlined underscore the importance of early identification and intervention. Services and support for mothers that address their emotional reactions and challenges are clearly needed. Affected parents and appropriate health care services therefore need information relating to particular challenges and risks faced by families raising children with SHCN.

of early identification and intervention. Services and support for mothers that address their emotional reactions and challenges are clearly needed. Affected parents and appropriate health care services therefore need information relating to particular challenges and risks faced by families raising children with SHCN. Reducing parenting stress and alleviating their caregiving burdens can therefore serve as important targets for future prevention and intervention (Cousino & Hazen, 2013). Acknowledgments We are grateful to all families in Norway who take part in this ongoing study. Funding This project was supported by a grant from the Research Council of Norway (ES-464464). The Norwegian Mother and Child Cohort Study is supported by grants from the Norwegian Ministry of Health and the Ministry of Education and Research, NIH/NIEHS (contract no. N01-ES-75558), NIH/NINDS (grant nos UO1 NS 047537-01 and UO1 NS 047537-06A1), and the Research Council of Norway/FUGE (grant no. 151918/S10). Conflicts of interest: None declared.

According to the Centers for Disease Control (2013), 13–20% of youth nationwide have a psychiatric diagnosis in any given year, and rates of individual diagnoses are rising. For youth in need of behavioral health (BH) services, the first point of contact is generally the primary care provider (Ford, Steinberg, Pidano, Honigfeld, & Meyers, 2006; Williams, Klinepeter, Palmes, Pulley, & Foy, 2004). However, tasked with surveying, screening, preventing, and treating a broad range of issues, primary care providers often fall short of adequately identifying (Kathol, deGruy & Rollman, 2014) and addressing BH problems. A recent study by Valleley and colleagues (2015) showed that even with routine BH screening, primary care providers responded to BH issues in only 21.4–50% of identified patients. Contributing to the divide between the mental health needs of youth and the BH services they require are mental health stigma (Cauce et al., 2002) and limited mental health access, particularly in underserved areas (Cummings, Wen, & Druss, 2013).

providers responded to BH issues in only 21.4–50% of identified patients. Contributing to the divide between the mental health needs of youth and the BH services they require are mental health stigma (Cauce et al., 2002) and limited mental health access, particularly in underserved areas (Cummings, Wen, & Druss, 2013). Fortunately, health care in the United States is in a state of reform that affords psychologists the opportunity to embed within primary care and collaboratively address BH issues. The Patient Protection and Affordable Care Act (ACA; Public Law No: 111-148, March 23, 2010) mandates mental health coverage at a level similar to that provided for medical care. Additionally, health care delivery models such as the patient-centered medical home model and accountable care organizations encourage comprehensive, cross disciplinary approaches to primary care that include BH providers (Korda and Eldridge, 2011). The ACA also actively promotes value-based health care that prioritizes affordable and accessible yet high-quality care, and financially incentivizes accordingly. While the fee-for-service era persists, active advocacy for health care reform that supports sustainable BH services within primary care is underway (Berwick, 2016; Kathol et al., 2014). With such reform, psychologists will be supported in providing revenue and non-revenue-generating services that improve both patient, provider, and clinic-wide functioning (Stancin & Perrin, 2014).

th care reform that supports sustainable BH services within primary care is underway (Berwick, 2016; Kathol et al., 2014). With such reform, psychologists will be supported in providing revenue and non-revenue-generating services that improve both patient, provider, and clinic-wide functioning (Stancin & Perrin, 2014). For BH providers collaborating to address the BH needs of primary care patients (i.e., integrated primary care; IPC), a number of delivery models or approaches exist. Blount (2003) described IPC approaches based on the level of integration between providers and treatment records, level of coordination in care, and degree of co-location between BH and primary care services. Heath, Wise, Romero, and Reynolds (2013) similarly outlined a continuum-based perspective to differentiate IPC delivery models. Along their continuum, IPC approaches ranged from simple coordination with improved communication (e.g., limited communication; separate records and facilities; little perceived value for one another’s roles), to full integration (e.g., fully integrated record; shared offices and exam rooms; routine communication at system and individual levels; blended roles with full appreciation for one another’s contributions).

(e.g., limited communication; separate records and facilities; little perceived value for one another’s roles), to full integration (e.g., fully integrated record; shared offices and exam rooms; routine communication at system and individual levels; blended roles with full appreciation for one another’s contributions). Multiple additional IPC taxonomies exist. For example, Gatchel and Oordt (2003) described five IPC models including co-located clinics, primary care provider, staff adviser, stepped-care approach, and primary care BH. Collins, Hewson, Munger, and Wade (2010) described eight IPC models, three of which overlap those of Gatchel and Oordt. More recently, Margolis, Pollard, and Niemiec (2013), along with Vogel, Malcore, Illes, and Kirkpatrick (2014), used the terms vertical and horizontal integration to differentiate additional aspects of IPC, including the scope of referral (targeted vs. broad), selection of intervention (prespecified vs. clinician selected), and outcomes of interest (targeted outcomes vs. overall symptom reduction and functional improvement).

4), used the terms vertical and horizontal integration to differentiate additional aspects of IPC, including the scope of referral (targeted vs. broad), selection of intervention (prespecified vs. clinician selected), and outcomes of interest (targeted outcomes vs. overall symptom reduction and functional improvement). Closer inspection of IPC delivery models reveals that each has strengths and limitations. For example, while Gatchel and Oordt’s (2003) colocated model allows for traditional mental health services and related interventions within primary care, the time required for services greatly reduces BH accessibility. On the other end of the continuum, the staff adviser model maximizes BH accessibility by providing brief consultation to primary care providers, but prevents the use of manualized treatments designed for the traditional mental health arena. Additionally, any IPC approach could be seen as poorly suited to a given practice on account of its incompatibility with their interests and priorities. Thus, as many conclude (e.g., Collins et al., 2010; Gatchel & Oordt, 2003; Robinson & Reiter, 2007), there is no gold standard, one-size-fits-all approach to IPC.

ena. Additionally, any IPC approach could be seen as poorly suited to a given practice on account of its incompatibility with their interests and priorities. Thus, as many conclude (e.g., Collins et al., 2010; Gatchel & Oordt, 2003; Robinson & Reiter, 2007), there is no gold standard, one-size-fits-all approach to IPC. Rather, selection of the “best” IPC approach rests on careful consideration of ones’ unique context, including provider preferences, skills, and comfort with integration; buy-in of relevant stakeholders; and practical considerations such as staffing, space, and availability of electronic medical records. In the interest of finding the delivery model with the best fit, it is common for IPC practices to incorporate aspects of multiple models into a singular, hybridized approach (Collins et al., 2010; Talen, Valeras, & Cesare, 2013; Wallace et al., 2015).

as staffing, space, and availability of electronic medical records. In the interest of finding the delivery model with the best fit, it is common for IPC practices to incorporate aspects of multiple models into a singular, hybridized approach (Collins et al., 2010; Talen, Valeras, & Cesare, 2013; Wallace et al., 2015). Within pediatric populations, psychology-led, integrated BH services have produced significant clinical improvements for children. For example, Berkovits, O’Brien, Carter, and Eyberg (2010) compared two approaches to parent–child interaction therapy that were modified for the primary care setting. After intervention and at 6-month follow-up, both treatment groups demonstrated significant decreases in problematic parenting practices and child behavioral problems. In another study, Lavigne and colleagues (2008) compared three interventions delivered to preschoolers within primary care; all three interventions resulted in significant reductions in oppositional behaviors and externalizing difficulties that were maintained at 12-month follow-up. Using brief treatment approaches within a co-located pediatric psychology clinic, Sobel, Roberts, Rayfield, Barnard, and Rapoff (2001) found youth demonstrated significant improvements in target behaviors after intervention. Together, outcome-based studies suggest a variety of IPC delivery models and BH interventions of varying length and modality can benefit patients.

d pediatric psychology clinic, Sobel, Roberts, Rayfield, Barnard, and Rapoff (2001) found youth demonstrated significant improvements in target behaviors after intervention. Together, outcome-based studies suggest a variety of IPC delivery models and BH interventions of varying length and modality can benefit patients. Data supporting the value-added benefits of pediatric IPC are also accumulating. Studies show IPC improves primary care provider utilization of BH services (Brawer, Martielli, Pye, Manwaring, & Tierney, 2010), reduces mental health stigma (Brawer et al., 2010), and increases both patient satisfaction (Finney, Riley, & Cataldo, 1991; Lavigne et al., 2008) and primary care provider satisfaction (Blount et al., 2007). BH services within primary care have also generated small, albeit meaningful, medical cost offsets (Katon, 1995), a finding lending support to the role of BH in reducing the exorbitant cost of health care for families in the United States (Cohen & Kirzinger, 2014).

mary care provider satisfaction (Blount et al., 2007). BH services within primary care have also generated small, albeit meaningful, medical cost offsets (Katon, 1995), a finding lending support to the role of BH in reducing the exorbitant cost of health care for families in the United States (Cohen & Kirzinger, 2014). Amidst mounting evidence in favor of IPC, a considerable gap in the literature exists with regard to BH provider productivity (i.e., percent of BH provider time spent in direct patient care) and billing practices. Despite multiple publications describing productivity and billing practices as important future directions for study (e.g., Bruns, Kessler, & VanDorsten, 2014; Rozensky & Janicke, 2012; Tynan & Woods, 2013), an extensive literature review revealed no practice-based data on the percent of time BH providers spend in patient care; data on billing practices associated with BH services within primary care were also nonexistent. Like all models of service delivery, BH services within primary care should not be evaluated solely on the basis of economic viability, but must include such factors to demonstrate sustainability (Goodheart, 2010). As we approach an era of unparalleled health care reform, psychology is not likely to maintain its footing within primary care in the absence of practice-based evidence to guide its future.

valuated solely on the basis of economic viability, but must include such factors to demonstrate sustainability (Goodheart, 2010). As we approach an era of unparalleled health care reform, psychology is not likely to maintain its footing within primary care in the absence of practice-based evidence to guide its future. The purpose of the current investigation was to provide descriptive information on BH productivity and billing practices obtained within a large, urban, pediatric primary care clinic. Specifically, this study explored trends and overall BH productivity, billing codes used, and total BH charges overall and by insurance type (e.g., Medicaid HMO plans, Blue Cross/Blue Care Network) over 2.5 years within the primary care setting. To provide context for results, data on nonattendance rates and BH encounter types (i.e., initial vs. follow-up visits) were also collected. In the absence of prior publications, the aforementioned factors were explored based on the authors’ clinical experience, which suggested clinic-wide attendance difficulties and a preponderance of initial visits may impact BH productivity and billing outcomes.

s (i.e., initial vs. follow-up visits) were also collected. In the absence of prior publications, the aforementioned factors were explored based on the authors’ clinical experience, which suggested clinic-wide attendance difficulties and a preponderance of initial visits may impact BH productivity and billing outcomes. Methods Integrated Primary Care Setting The pediatric primary care clinic is located in an underserved, urban setting, where 62.4% of those <18 years old live below the federal poverty level (United States Census Bureau, 2013). The clinic is a large, pediatric-residency-affiliated patient-centered medical home (i.e., 4,705 patients served from September 1, 2012 to February 28, 2015) that operates 4.5 days per week. Primary care providers include two to three attending physicians, four to six medical residents, one physician’s assistant, and one to three medical students. Primary care providers engage in developmental screening routinely with patients aged 0–5 years, and use standardized screening tools to evaluate for social/emotional difficulties and autism spectrum disorder as concerns arise.

four to six medical residents, one physician’s assistant, and one to three medical students. Primary care providers engage in developmental screening routinely with patients aged 0–5 years, and use standardized screening tools to evaluate for social/emotional difficulties and autism spectrum disorder as concerns arise. The BH service within the primary care clinic is run by two licensed pediatric psychologists and a postdoctoral psychology fellow. Costs associated with the BH service are subsidized by the affiliated hospital entity on account of the service’s central role in delivering coordinated, cross disciplinary care. Time allocated for the BH service is 1–3 half days per week (i.e., 195 min per half day). During each half day of BH services, one BH provider is on site and three BH encounters (i.e., initial or follow-up visits) are scheduled. BH encounters include both spontaneous (i.e., immediately after a patient’s primary care visit) and scheduled (i.e., BH appointments made no more than 4 weeks in advance on a first-available basis) visits. Consequently, BH encounters on any given half day include varying combinations of initial, follow-up, scheduled, and spontaneous visits. When on site, but not engaged in direct patient care, BH providers participate in on-the-fly consultation with primary care providers, educate learners, develop and acquire BH handouts for general clinic use, generate flyers promoting BH services, conduct quality improvement projects, assist in case management activities, and conduct scholarly activities specific to IPC.

BH providers participate in on-the-fly consultation with primary care providers, educate learners, develop and acquire BH handouts for general clinic use, generate flyers promoting BH services, conduct quality improvement projects, assist in case management activities, and conduct scholarly activities specific to IPC. The primary care and BH service lines are highly integrated with unified records, shared exam rooms, and a communal provider workspace. BH requests come from primary care physicians and patient self-referrals for a myriad of developmental, behavioral, and social/emotional issues. Interventions are selected by the BH provider and designed to be problem-specific and brief (i.e., fewer sessions; shorter session duration) relative to traditional mental health settings. Initial visits generally incorporate assessment, feedback (i.e., including identified psychiatric diagnoses when applicable), recommendations, and intervention. Follow-up visits are scheduled at 3–4 week intervals, remain problem focused, and generally incorporate intervention and recommendations. Patients or families in need of a higher level of support owing to problem acuity or complexity are referred to traditional mental health settings. Specific interventions commonly used include diagnostic interviews, social/emotional screening, psychoeducation, cognitive and behavioral interventions, skills training, and motivational interviewing. Outcomes of interest generally consist of symptom reduction, problem resolution, and functional improvements. After BH encounters, feedback is routinely communicated to the primary care provider within the medical record, and whenever possible, in person.

l interventions, skills training, and motivational interviewing. Outcomes of interest generally consist of symptom reduction, problem resolution, and functional improvements. After BH encounters, feedback is routinely communicated to the primary care provider within the medical record, and whenever possible, in person. Additional BH professionals (e.g., social workers, family navigators, care coordinators) are not on site. When the BH service is unavailable, primary care providers stratify the severity (i.e., mild, moderate, severe) and scope (e.g., narrow, broad) of the problem(s) identified. Based on their assessment, primary care providers then deliver general guidance and education that includes resources compiled by the BH service, provide warm handoffs for a future BH scheduled visit, and/or refer patients to community-based providers (e.g., community mental health, outpatient psychiatry and psychotherapy, school resources) using a referral guide created by the BH service.

uidance and education that includes resources compiled by the BH service, provide warm handoffs for a future BH scheduled visit, and/or refer patients to community-based providers (e.g., community mental health, outpatient psychiatry and psychotherapy, school resources) using a referral guide created by the BH service. Participants On approval by the institution’s review board, electronic medical records were reviewed for all BH and primary care encounters between September 1, 2012 and February 28, 2015. Demographic characteristics of primary care and BH patients can be found in Table I. BH patients (N  =  204) ranged in age from <1 year old to 17 years old (M  = 8.62 years; SD  =  4.14). A majority of BH patients were male (62.25%, N  =  127), and 75% (N  = 153) of patients were African American. Relative to the primary care service, the BH service had a generally similar racial composition (χ2  =  3.09, df  =  2, p  =  .2133), young patients (0–2 years) were underrepresented (χ2  =  45.98, df  =  3, p  <  .0001), and patients were more likely to be male (χ2 = 8.46, df  =  1, p  < .01). Table I. Demographic Characteristics of Patients

care service, the BH service had a generally similar racial composition (χ2  =  3.09, df  =  2, p  =  .2133), young patients (0–2 years) were underrepresented (χ2  =  45.98, df  =  3, p  <  .0001), and patients were more likely to be male (χ2 = 8.46, df  =  1, p  < .01). Table I. Demographic Characteristics of Patients Variables BH service (N = 204) PC service (N = 4705) n % n % ÷2 p Age (years) 45.98 <.0001 0–2 9 4.41 1,023 21.74 3–5 45 22.06 780 16.58 6–11 97 47.55 1,436 30.52 12–17 53 25.98 1,165 24.76 18+ 0 – 301 6.40 Gender 8.46 <.001. Male 127 62.25 2,428 51.60 Female 77 37.75 2,277 48.40 Race 3.09 .2133 African American 153 75.00 3,673 78.07 Caucasian (Non-Hispanic) 46 22.55 857 18.21 Other 5 2.45 175 3.72 Note. Behavioral health service age is the patient age at the time of first BH service encounter. Primary care service age is the patient age on the last day of the data collection period (February 28, 2015). Other race  =  Asian, Biracial, Hispanic, Multiracial, and Unknown; BH  =  behavioral health; PC  =  primary care.

e. Behavioral health service age is the patient age at the time of first BH service encounter. Primary care service age is the patient age on the last day of the data collection period (February 28, 2015). Other race  =  Asian, Biracial, Hispanic, Multiracial, and Unknown; BH  =  behavioral health; PC  =  primary care. Definition of Productivity BH productivity was broadly defined as the percent of BH provider time spent within the primary care clinic in direct patient care. Total half days, each consisting of 195 min, were used to calculate total BH provider time in the primary care clinic. Time spent in direct patient care was calculated based on the billing codes (current procedural terminology codes; CPT codes) assigned to each patient encounter at the time of service. Time spent in patient care for each encounter was estimated as follows: ■ Psychiatric diagnostic evaluations and interviews: 60 min were assigned, based on the clinical experience of the providers. ■ Psychotherapy sessions with duration ranges: The number representing the middle of the range (e.g., 25 min for a 20–30 min CPT code range) was assigned. ■ Psychotherapy sessions with a specified duration: The duration specified as part of the CPT code (e.g., 50 min) was assigned. ■ Health and behavior codes: 15 min were assigned per health and behavior unit specified in billing documentation.

■ Psychotherapy sessions with duration ranges: The number representing the middle of the range (e.g., 25 min for a 20–30 min CPT code range) was assigned. ■ Psychotherapy sessions with a specified duration: The duration specified as part of the CPT code (e.g., 50 min) was assigned. ■ Health and behavior codes: 15 min were assigned per health and behavior unit specified in billing documentation. ■ No charge codes: For the few BH encounters for which a billable CPT code did not apply, initial visits were assigned the mean duration of all billable diagnostic evaluations and initial health and behavior visits; follow-up visits were assigned the mean duration of all billable psychotherapy and health and behavior follow-up visits. Patterns of BH productivity were explored by calendar-based months, quarters, and years. Quarters consisted of consecutive, 3-month intervals beginning at the start of the investigation (September 2012). Calendar years included 2012, 2013, 2014, and 2015. BH productivity was also explored by season and included autumn (September through November), winter (December through February), spring (March through May), and summer (June through August).

onth intervals beginning at the start of the investigation (September 2012). Calendar years included 2012, 2013, 2014, and 2015. BH productivity was also explored by season and included autumn (September through November), winter (December through February), spring (March through May), and summer (June through August). Contextual factors explored included nonattendance rates and BH encounter types. For the primary care service, nonattendance rates were defined as the percent of unattended appointments (cancelled and no show appointments combined) relative to the total number of scheduled appointments. For the BH service, nonattendance rates were defined as the percent of unattended appointments (cancelled and no show appointments combined) relative to the total number of scheduled and spontaneous BH encounters. At the outset of each BH visit, BH providers selected the encounter type and used a corresponding documentation template (i.e., “Behavioral Health Intake” or “Follow-Up Behavioral Health Consultation”). BH encounter types were selected based on the following prespecified criteria: Initial visits included patients new to the BH service and patients previously seen by the BH service with a new presenting problem or who had not been seen in the past 6 months; follow-up visits included BH encounters addressing a previously identified BH problem within 6 months of the last BH encounter.

ied criteria: Initial visits included patients new to the BH service and patients previously seen by the BH service with a new presenting problem or who had not been seen in the past 6 months; follow-up visits included BH encounters addressing a previously identified BH problem within 6 months of the last BH encounter. Definition of Billing Practices Billing practices included three main components: billing codes used, total BH charges, and total BH charges by insurance type. Unfortunately, data on reimbursement for BH services were incomplete owing to extended delays in collection, and thus unavailable for analysis. Billing codes were defined as the CPT codes submitted for BH services provided in the primary care setting from September 1, 2012 through February 28, 2015. All billing codes were assigned at the time of service (i.e., immediately after BH encounter completion) and selected for their congruence with the respective CPT code definitions (e.g., psychotherapy, psychiatric diagnostic evaluation, health and behavior initial assessment). BH charges were defined as the total dollar amount of charges for BH services provided overall and according to insurance type (e.g., Commercial plans, Medicaid alone, Medicaid HMO plans). BH charges were calculated based on a fee schedule the billing department applied to all BH encounters from September 1, 2012 through February 28, 2015.

ed as the total dollar amount of charges for BH services provided overall and according to insurance type (e.g., Commercial plans, Medicaid alone, Medicaid HMO plans). BH charges were calculated based on a fee schedule the billing department applied to all BH encounters from September 1, 2012 through February 28, 2015. Procedure Data Collection Data were retrospectively collected as part of a broader investigation of BH services within the pediatric primary care setting. For the current investigation, two categories of data were analyzed: BH productivity and billing practices. Productivity data included the following: minutes of BH provider time spent within the primary care setting, minutes BH providers spent in direct patient care, dates of BH encounters, BH encounter types (initial vs. follow-up visit), and attendance rates for both BH and primary care services. The minutes of BH provider time spent in primary care were collected via retrospective review of the BH schedule within the electronic medical record and confirmed by the BH provider network schedule. A retrospective chart review was also used to determine the dates of completed BH encounters within the primary care setting, the billing codes assigned, and the number of initial and follow-up BH encounters. A report generated by the primary care clinic’s electronic medical record provided data regarding nonattendance rates for both the BH and primary care services.

ermine the dates of completed BH encounters within the primary care setting, the billing codes assigned, and the number of initial and follow-up BH encounters. A report generated by the primary care clinic’s electronic medical record provided data regarding nonattendance rates for both the BH and primary care services. Billing practices data included the CPT codes submitted for each BH encounter, and charges for BH services within primary care from September 1, 2012 through February 28, 2015. A report generated by the primary care clinic’s electronic medical record, a financial report generated by the billing department, and a retrospective chart review provided data regarding the billing codes used. The financial report generated by the billing department also included total BH charges by billing code and by insurance type (e.g., Medicaid HMO plans, Commercial plans). Process to Ensure Reliability of Data Collection Chart reviews were completed by dyads of affiliated and nonaffiliated researchers with 100% agreement for data collection and entry. Three documentation sources for service data and billing codes were cross-referenced with 100% accuracy. Financial and electronic medical record reports compiled by unaffiliated specialists were reviewed with a researcher and confirmed 100% accuracy of the dates recorded and operational definitions of variables captured.

ree documentation sources for service data and billing codes were cross-referenced with 100% accuracy. Financial and electronic medical record reports compiled by unaffiliated specialists were reviewed with a researcher and confirmed 100% accuracy of the dates recorded and operational definitions of variables captured. Statistical Data Analysis Methods Productivity estimates were calculated by dividing the amount of time BH providers spent in direct patient care by the total time BH providers were in the primary care setting and multiplying the result by 100. Descriptive statistics were used to examine BH encounter types and BH productivity by day, month, quarter, and calendar year. Productivity was also inferentially explored for differences between seasons and between calendar years using a one-way analysis of variance (ANOVA). Chi-square analysis was used to compare attendance rates between the BH and primary care services. Billing (i.e., CPT) codes used by BH providers and BH charges (i.e., total dollar amount of charges for BH services overall and by insurance type) were explored descriptively. To provide an estimate of charges a BH provider could expect to submit per hour of time spent in the primary care setting, an average hourly rate of BH charges was also calculated. This was done by dividing the total BH charges by the total hours BH providers were in the primary care setting.

re explored descriptively. To provide an estimate of charges a BH provider could expect to submit per hour of time spent in the primary care setting, an average hourly rate of BH charges was also calculated. This was done by dividing the total BH charges by the total hours BH providers were in the primary care setting. Results Productivity During the 2.5 years under exploration, BH providers engaged in 149 separate days of service (some half days), and were present in the primary care clinic for 646.75 hr. Two hundred four patients were seen by the BH service (4.34% of the 4,705 patients served by the primary care clinic). BH providers completed 244 encounters, which translated into an estimated 228.17 hr of direct patient care. Initial visits (i.e., psychiatric diagnostic evaluations; health and behavior initial assessments) had an average estimated duration of 59.05 min (SD  =  5.84). Follow-up visits (i.e., psychotherapy sessions; health and behavior reevaluations or intervention) had an average estimated duration of 35.86 min (SD =  12.64). One follow-up visit was erroneously billed as a diagnostic interview. Consequently, the visit was treated as missing data and assigned the mean value of all follow-up visits (i.e., 35.86 min).

rapy sessions; health and behavior reevaluations or intervention) had an average estimated duration of 35.86 min (SD =  12.64). One follow-up visit was erroneously billed as a diagnostic interview. Consequently, the visit was treated as missing data and assigned the mean value of all follow-up visits (i.e., 35.86 min). Overall BH productivity, defined as the percent of BH provider time spent within the primary care clinic in direct patient care, was 35.28%. BH productivity was characterized by high degrees of variability at daily (M  =  34.02%, SD  =  27.43%) and monthly (M  =  36.51%, SD  =  10.21%) intervals. As shown in Figure 1, increased stability in BH productivity was demonstrated when the interval under exploration was lengthened from months to quarters of a year (M  =  36.42%, SD  =  6.46%). While data on daily productivity were analyzed, they were not included in Figure 1 owing to the complexity of representing daily data graphically. One-way ANOVAs revealed no statistically significant difference in daily BH productivity means between seasons (F(3,143)  =  1.896, df  =  3, p  =  .133), or calendar years (F(3, 143)  =  1.978, df  =  3, p  =  .120). A descriptive comparison of BH productivity means by month (e.g., combining BH productivity means from all February months occurring between September 1, 2012 and February 28, 2015) revealed moderate differences by month. The combination of all July months were the least productive months (M =  20.00%, SD = 24.90%) and the combination of all January months were the most productive months (M  =  46.00%; SD  = 31.30%). Figure 1. Behavioral health service productivity by month and quarter from September 1, 2012 through February 28, 2015. Mean productivity was highly variable by month, and less variable by quarter.

.00%, SD = 24.90%) and the combination of all January months were the most productive months (M  =  46.00%; SD  = 31.30%). Figure 1. Behavioral health service productivity by month and quarter from September 1, 2012 through February 28, 2015. Mean productivity was highly variable by month, and less variable by quarter. Rates of nonattendance (no-shows and cancelled visits combined) were 59.54% (N  =  359) for the BH service and 42.31% (N  =  11,554) for the primary care service. While the nonattendance rates for both services were high, the BH rate of nonattendance was significantly higher than the primary care rate (χ2  =  71.59, df  =  1, p < .0001). With regard to rates of initial and follow-up BH encounters, 87.30% (N  =  213) of all 244 completed BH encounters were initial visits. The remaining 12.70% (N  =  31) BH encounters were follow-up visits.

Rates of nonattendance (no-shows and cancelled visits combined) were 59.54% (N  =  359) for the BH service and 42.31% (N  =  11,554) for the primary care service. While the nonattendance rates for both services were high, the BH rate of nonattendance was significantly higher than the primary care rate (χ2  =  71.59, df  =  1, p < .0001). With regard to rates of initial and follow-up BH encounters, 87.30% (N  =  213) of all 244 completed BH encounters were initial visits. The remaining 12.70% (N  =  31) BH encounters were follow-up visits. Billing Practices As shown in Table II, psychiatric diagnostic evaluation billing codes were most commonly used to charge for BH services provided (N  =  200) from September 1, 2012 through February 28, 2015. When examining financial data, a total of $52,050.00 was charged for BH services provided within the primary care clinic. BH charges by insurance type were as follows: $46,890.00 (90.09% of charges) was submitted to Medicaid HMO plans; $1,405.00 (2.70% of charges) was submitted for self-pay; $1,345.00 (2.58% of charges) was submitted to Medicaid alone; $1,325.00 (2.55% of charges) was submitted to Blue Cross/Blue Care Network plans; and $1,085.00 (2.08% of charges) was submitted to Commercial plans. Dividing the total charges for BH services ($52,050.00) by the total hours BH providers were in the primary care setting (646.75 hr) from September 1, 2012 through February 28, 2015, BH providers were found to bill an average rate of $80.48 per hour of time on site. Table II. Billing Codes for Behavioral Health Services Within Pediatric Primary Care (September 2012–February 2015)

total hours BH providers were in the primary care setting (646.75 hr) from September 1, 2012 through February 28, 2015, BH providers were found to bill an average rate of $80.48 per hour of time on site. Table II. Billing Codes for Behavioral Health Services Within Pediatric Primary Care (September 2012–February 2015) Encounter type Current procedural terminology codes n Initial No Charge 7 Psychiatric Diagnostic Evaluation [90791] 174 Psychiatric Diagnostic Interview [90801] 25 Health & Behavior Assessment, Initial [96150] 7 Follow-up No Charge 1 Psychotherapy Patient and/or Family 30 Minutes [90832] 6 Psychotherapy Patient and/or Family 45 Minutes [90834] 11 Psychotherapy Patient and/or Family 60 Minutes [90837] 4 Psychotherapy (20–30) [90804] 3 Psychotherapy (45–50) [90806] 2 Health & Behavior Assessment, Re-Assessment [96151] 1 Health & Behavior Intervention [96152] 1 Family Psychotherapy, No patient [90846] 1 Psychiatric Diagnostic Interview [90801]a 1 aFollow-up intervention erroneously coded as a Psychiatric Diagnostic Interview.

therapy (20–30) [90804] 3 Psychotherapy (45–50) [90806] 2 Health & Behavior Assessment, Re-Assessment [96151] 1 Health & Behavior Intervention [96152] 1 Family Psychotherapy, No patient [90846] 1 Psychiatric Diagnostic Interview [90801]a 1 aFollow-up intervention erroneously coded as a Psychiatric Diagnostic Interview. Discussion Despite widespread IPC interest and an unprecedented opportunity for psychology to establish itself within the patient-centered medical home, evidence of sustainability remains insufficient. To that end, this study retrospectively explored BH productivity and billing data within a large primary practice located in an urban, underserved community. In a fee-for-service framework, findings reveal fairly significant threats to IPC sustainability, namely suboptimal productivity, unpredictably variable productivity over time, and low billing rates. Suboptimal Overall Productivity In a purely economic sense, BH providers were underproductive. Explanations for inadequate time in direct patient care are broad, and include barriers to getting children with BH issues connected to BH services, high rates of nonattendance for scheduled visits, and limited continuity in BH caseload.

verall Productivity In a purely economic sense, BH providers were underproductive. Explanations for inadequate time in direct patient care are broad, and include barriers to getting children with BH issues connected to BH services, high rates of nonattendance for scheduled visits, and limited continuity in BH caseload. Although a goal of the IPC delivery approach under exploration was to increase BH service access, only 4.34% of primary care patients were seen by BH providers. Given that >25% of pediatric primary care patients present with BH risks (Blucker et al., 2014), the integrated BH service was underused. As suggested by Valleley, Romer, Kupzyk, Evans, & Allen (2015), one potential explanation may be that even with high integration, primary care providers fall short of identifying BH issues and responding to them with referral to BH providers once detected. Other possible explanations may be inherent to the IPC approach used. That is, aspects of the IPC approach such as absence of clear referral pathways (e.g., routine referral of all children with insomnia), lack of routine social/emotional screening at well-child visits, and part-time BH availability that results in an overreliance on scheduled visits may also contribute to reduced BH volume.

is, aspects of the IPC approach such as absence of clear referral pathways (e.g., routine referral of all children with insomnia), lack of routine social/emotional screening at well-child visits, and part-time BH availability that results in an overreliance on scheduled visits may also contribute to reduced BH volume. Compounding the issue of BH underutilization was the poor attendance rates for scheduled BH encounters. Barriers to BH service attendance span multiple issues. One factor illuminated by the current study was the poor attendance rate of the primary care service (57.69%). Similar primary care attendance rates have been documented elsewhere (e.g., George & Rubin, 2003), and suggest barriers to BH attendance within primary care are not unique to BH services. Studies identify several factors associated with nonattendance that are disproportionately represented within underserved areas, including low income, high familial stress, racial minority status, and single–parent households (Gopalan et al., 2010; Kalb et al., 2012).

within primary care are not unique to BH services. Studies identify several factors associated with nonattendance that are disproportionately represented within underserved areas, including low income, high familial stress, racial minority status, and single–parent households (Gopalan et al., 2010; Kalb et al., 2012). While some barriers to BH attendance are likely shared with primary care providers, the nonattendance rate of the BH service was significantly higher than that of the primary care service and comparable with those of nonintegrated, noncolocated, traditional child mental health services in urban settings (48–62%; McKay & Bannon, 2004). These findings suggest barriers unique to BH services also exist. Previous research commonly cites mental health stigma (Cauce et al., 2002) as a barrier to BH treatment. Additionally, the authors speculate that high nonattendance rates may be explained in part by delays between the BH referral and BH visit, an insufficient frequency of spontaneous visits, and by approaches to BH referral that inconsistently include aspects essential to facilitating follow-through (e.g., clarity and relevance of the referral reason to the patient and family; provision of brief education on the BH provider’s role; reassurance of BH provider expertise and trustworthiness; Robinson & Reiter, 2007).

pproaches to BH referral that inconsistently include aspects essential to facilitating follow-through (e.g., clarity and relevance of the referral reason to the patient and family; provision of brief education on the BH provider’s role; reassurance of BH provider expertise and trustworthiness; Robinson & Reiter, 2007). Another identified factor likely contributing to low BH productivity was how few BH patients returned for follow-up visits. While the modal number of attended BH sessions in traditional mental health settings is one (Connolly-Gibbons et al., 2011), IPC was developed, in part, to ameliorate barriers contributing to this outcome (United States Public Health Service Office of the Surgeon General, 2001). Further study is needed to identify factors contributing to poor BH follow-up. The authors suspect a combination of patient (e.g., multiple psychiatric comorbidities; pharmacologic management indicated), provider or service (e.g., limited appointment times available for scheduled visits; over-referral to outpatient services), and social (e.g., transportation problems, competing needs prioritized to BH issues) factors are implicated.

patient (e.g., multiple psychiatric comorbidities; pharmacologic management indicated), provider or service (e.g., limited appointment times available for scheduled visits; over-referral to outpatient services), and social (e.g., transportation problems, competing needs prioritized to BH issues) factors are implicated. Variable, Unpredictable Productivity Fluctuations in productivity that could not be ascribed to a reliable pattern were noted day-to-day and month-to-month. Given this result, BH providers within primary care should be prepared for a fair degree of unpredictability in daily routines. To optimize nonrevenue generating time, BH providers are encouraged to focus efforts toward value-added contributions and benefits to the primary care setting at large. Examples include brief consultation with primary care providers, education on a variety of BH topics for health professionals, development of routine referral pathways to increase future BH productivity, development or acquisition of BH handouts, quality improvement projects, case-management activities, and scholarly activities specific to IPC. Empirically, additional studies of longitudinal BH productivity and identification of value-added benefits associated with IPC are needed.

ys to increase future BH productivity, development or acquisition of BH handouts, quality improvement projects, case-management activities, and scholarly activities specific to IPC. Empirically, additional studies of longitudinal BH productivity and identification of value-added benefits associated with IPC are needed. Low Billing Rates If executed within a fee-for-service setting where the value-added benefits of IPC are not taken into account, BH charges would likely prove inadequate to support long-term financial sustainability. Billing rates in the range evidenced are particularly concerning given the significant barriers to BH reimbursement, including generally low reimbursement rates for mental health services (e.g., Michigan Department of Health and Human Services, 2016), and rates that vary based on insurance type (Margolis, Pollard, & Niemiec, 2013), BH date of service (i.e., same vs. separate day primary care visit; Robinson & Reiter, 2007), billing code used, and the geographic location of practice (e.g., Department of Health and Mental Hygiene, 2016; Missouri Department of Social Services, 2016). Clinically, results suggest strategies to increase BH productivity will be essential to IPC sustainability in fee-for-service models. Outside fee-for-service models, IPC providers should strive to increase their value-added benefit, as previously discussed. Empirically, extending research into the cost and reimbursement of BH services delivered within various IPC models is critical. Securing grant funding will help existing IPC practices fund research initiatives to explore IPC and publish outcomes, and aid those interested in initiating IPC in securing the start-up costs.

mpirically, extending research into the cost and reimbursement of BH services delivered within various IPC models is critical. Securing grant funding will help existing IPC practices fund research initiatives to explore IPC and publish outcomes, and aid those interested in initiating IPC in securing the start-up costs. Limitations The current investigation makes significant progress toward filling a void in IPC literature, but is not without limitations. By relying on a retrospective chart review, the accuracy, consistency, and breadth of data available were wholly dependent on the providers’ documentation and the electronic medical record’s capabilities. To that end, BH productivity results were estimates based on billing data, rather than precise documentation of time spent with patients. While productivity results were also affected by nonbillable BH encounters where an average encounter duration was applied in the absence of precise data, their infrequent occurrence renders the impact minimal. Factors of interest that could not be explored owing to lack of available data included reasons for nonattended visits, comparisons of spontaneous and scheduled visits, BH referral reasons, and reimbursement rates. Finally, owing to the focus on direct patient care activities and billing practices, highly valuable yet non-revenue-generating BH services within the primary care setting under investigation were not captured. Thus, the economically less desirable outcomes of the current investigation fail to account for the value-added services of integrated BH providers within primary care.

es and billing practices, highly valuable yet non-revenue-generating BH services within the primary care setting under investigation were not captured. Thus, the economically less desirable outcomes of the current investigation fail to account for the value-added services of integrated BH providers within primary care. Statistically, the number of BH encounters captured within this study were insufficient to complete inferential trend analyses. Nevertheless, the results obtained were meaningful, and a first step toward revealing practice-based data where none previously existed. In addition, when comparing daily productivity between calendar years, the differences in group size limited the sensitivity of the statistical analysis. Methodologically, three of the five researchers functioned as BH providers within the IPC practice. While IPC experience aided in developing hypotheses for exploration in the absence of preexisting evidence, it potentially introduced bias into study design and result interpretation. Bias was limited by involvement of multidisciplinary, multisite researchers and reviewers from study conception through completion.

ice. While IPC experience aided in developing hypotheses for exploration in the absence of preexisting evidence, it potentially introduced bias into study design and result interpretation. Bias was limited by involvement of multidisciplinary, multisite researchers and reviewers from study conception through completion. As noted by Collins and colleagues (2010), the diversity among IPC delivery models is vast. Consequently, study generalizability is limited to populations, primary care clinics, and models of IPC similar to those included in this investigation. That stated, the setting under exploration is also a study strength, given its likely resemblance to many other IPC practices (Cameron & Mauksch, 2002) and the proposed benefits of IPC to patients particularly within urban, underserved, low socioeconomic areas (Blount, 2003; Sanchez, Chapa, Ybarra, & Martinez, 2012). Finally, the exploratory nature of this investigation renders explanations for results preliminary.

to many other IPC practices (Cameron & Mauksch, 2002) and the proposed benefits of IPC to patients particularly within urban, underserved, low socioeconomic areas (Blount, 2003; Sanchez, Chapa, Ybarra, & Martinez, 2012). Finally, the exploratory nature of this investigation renders explanations for results preliminary. Future Directions As an exploratory study, the present investigation requires replication across a wide range of IPC approaches, primary care settings, and communities. Through replication, the scope, mitigating, and ameliorating factors associated with this potential threat to IPC financial sustainability can be clarified. Investigations are greatly needed that identify which IPC delivery models, billing practices, primary care characteristics, and BH services are associated with enhanced fiscal outcomes. Physicians have demonstrated support for the patient-centered medical home model by amassing data (Ward-Zimmerman & Cannata, 2012). Pediatric psychology too must establish empirical support for the value-added contributions of IPC services. Alongside IPC reimbursement analysis, investigations of medical cost offsets (e.g., reduced emergency department and urgent care visits; increased physician productivity), and ongoing exploration of value-added IPC benefits (e.g., increased provider and patient satisfaction; reduced physician burnout; improved primary care and BH attendance; improved patient outcomes) are needed. As previously alluded to, informing the future of psychologists within primary care will prove challenging without practice-based data to guide the way.

benefits (e.g., increased provider and patient satisfaction; reduced physician burnout; improved primary care and BH attendance; improved patient outcomes) are needed. As previously alluded to, informing the future of psychologists within primary care will prove challenging without practice-based data to guide the way. Clinically, increased interdisciplinary collaboration is needed to solve the practical, sociocultural, systemic, and economic obstacles continuing to impede access to BH services. Approaches to clinical practice that optimize patient and physician engagement, improve detection of BH concerns, and maximize provider productivity are also imperative. In addition, strategies to improve interdisciplinary synergy would be beneficial (e.g., interdisciplinary education, expanded BH service hours, routine screening practices, strategic primary care and BH scheduling, vertical service lines). Providers should also develop strategies to increase the likelihood of BH follow-up beyond the first visit.

improve interdisciplinary synergy would be beneficial (e.g., interdisciplinary education, expanded BH service hours, routine screening practices, strategic primary care and BH scheduling, vertical service lines). Providers should also develop strategies to increase the likelihood of BH follow-up beyond the first visit. Beyond service-sustaining implications, findings clearly reflect persistent barriers to BH service engagement. To maximize patient benefit and maintain BH accessibility within primary care, clinical innovation and research is needed with regard to best practices for single-session and brief interventions. While continuing to address barriers to service, BH providers have a responsibility to ensure the efficacy of the care provided in the present climate. The results of this investigation should compel providers to explore what change can be effected within one to three visits, if that is all that will likely occur. Finally, without health care reform toward increased coordination, integration, and outcome-based incentives, BH services within primary care remain vulnerable (Vogel et al., 2014). While many support such reform, change will not happen without psychologists garnering evidence of IPC benefits and actively advocating at local, regional, state, and national levels. Through advocacy for reform, a climate conducive to IPC sustainability can be promoted.

ary care remain vulnerable (Vogel et al., 2014). While many support such reform, change will not happen without psychologists garnering evidence of IPC benefits and actively advocating at local, regional, state, and national levels. Through advocacy for reform, a climate conducive to IPC sustainability can be promoted. Conclusion The current investigation reveals both challenges and opportunities faced by psychologists within an increasingly collaborative, cross-disciplinary health care climate. This study is the first to provide practice-based information on BH productivity and billing practices within a pediatric primary care setting. Study outcomes underscore the need for additional investigations of IPC sustainability both from an economic lens, but also more broadly. Results also suggest BH providers within primary care must strive for favorable economic, patient, and clinic-wide outcomes. For IPC to sustain, clinical and empiric efforts to improve BH productivity, demonstrate the value-added contributions of BH services within primary care, and successfully advocate for BH-supporting health care reform are essential. Findings should not deter further IPC growth, but rather serve as the catalyst for continued exploration and innovation. Acknowledgments The authors express appreciation to Jenny LaChance for her helpful suggestions during manuscript preparation. Funding This work was supported by a grant from the Hurley Children’s Hospital Pediatric Research & Education Fund. Conflicts of interest: None declared.

Introduction Nocturnal enuresis (bedwetting) is a common childhood disorder, and the risk is higher in boys than girls (Fergusson, Horwood, & Shannon, 1986). The etiology of bedwetting is believed to be multifactorial, involving genetic (von Gontard, Schaumburg, Hollmann, Eiberg, & Rittig, 2001), neurobiological (Jarvelin, 1989), and psychological risk factors (Joinson, Sullivan, von Gontard, & Heron, 2015). It has also been suggested that environmental factors including stressors in early childhood could play an important role in the etiology of bedwetting (Douglas, 1973). Stressors that have been examined in earlier studies include acute life events (e.g., death of parent), chronic stressors (e.g., family financial problems), normative events (e.g., birth of a sibling), and non-normative (unpredictable) stressors (e.g., parent’s serious injury). Exposure to early stressful events has been found to be associated with an increased risk of bedwetting in a small case control study of 7-year-olds (Jarvelin, Moilanen, Vikeväinen-Tervonen, & Huttunen, 1990), a small longitudinal study of children followed up to age 8 years (Kaffman & Elizur, 1977) and a cross sectional study of 6–16-year-olds (Kalo & Bella, 1996). Only one early prospective cohort study has examined the effects of stressful events on risk for subsequent bedwetting (Douglas, 1973). The study found that children who were exposed to “disturbing” events (e.g., family breakdown, moving house, accidents, separation from mother) in the first 4 years of life (especially at age 3–4 years) had an increased prevalence of bedwetting up to age 15 years. Children exposed to four or more disturbing events had around double the risk of experiencing bedwetting than those not exposed to such events (Douglas, 1973).

use, accidents, separation from mother) in the first 4 years of life (especially at age 3–4 years) had an increased prevalence of bedwetting up to age 15 years. Children exposed to four or more disturbing events had around double the risk of experiencing bedwetting than those not exposed to such events (Douglas, 1973). A limitation of earlier studies is that they examined only the presence or absence of bedwetting and did not take into account the heterogeneity in development of continence during childhood. There is now evidence for distinct patterns (longitudinal phenotypes) of development of nighttime bladder control during childhood that are characterized by normative development, delayed attainment, persistent bedwetting, or relapses (Croudace, Jarvelin, Wadsworth, & Jones, 2003; Joinson et al., 2009; Sullivan, Joinson, & Heron, 2015). No studies have examined whether these different patterns of incontinence in childhood are differentially associated with early stressful events. Another limitation of earlier studies is the lack of adjustment for a range of potential confounders that might explain the association between stressful events and bedwetting.

studies have examined whether these different patterns of incontinence in childhood are differentially associated with early stressful events. Another limitation of earlier studies is the lack of adjustment for a range of potential confounders that might explain the association between stressful events and bedwetting. We use data from a large UK cohort to examine whether stressful events in early childhood are associated with bedwetting at school age. Different patterns (latent classes) of typical and atypical development of nighttime bladder control have been previously identified using longitudinal data derived from parental reports of frequency of bedwetting at ages 4–9 years in almost 11,000 children from the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort (Sullivan et al, 2015). In this article, we use these latent classes to examine whether exposure to stressful events in the first 4 years of life is associated with an increased risk of bedwetting at 4–9 years.

4–9 years in almost 11,000 children from the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort (Sullivan et al, 2015). In this article, we use these latent classes to examine whether exposure to stressful events in the first 4 years of life is associated with an increased risk of bedwetting at 4–9 years. Methods Participants The sample comprised participants from the ALSPAC. Detailed information about ALSPAC is available on the study Web site (http://www.bristol.ac.uk/alspac), which includes a fully searchable dictionary of available data (http://www.bris.ac.uk/alspac/researchers/data-access/data-dictionary). Pregnant women resident in the former Avon Health Authority in south-west England, having an estimated date of delivery between April 1, 1991 and December 31, 1992 were invited to take part, resulting in a cohort of 14,541 pregnancies and 13,973 singletons/twins (7,217 boys and 6,756 girls) alive at 12 months (Boyd et al., 2013). Ethical approval for the study was obtained from the ALSPAC Law and Ethics committee and local research ethics committees. Written informed consent was obtained after the procedure(s) had been fully explained.

1 pregnancies and 13,973 singletons/twins (7,217 boys and 6,756 girls) alive at 12 months (Boyd et al., 2013). Ethical approval for the study was obtained from the ALSPAC Law and Ethics committee and local research ethics committees. Written informed consent was obtained after the procedure(s) had been fully explained. Exposure: Stressful Events in Early Childhood Stressful events were measured using a maternally reported questionnaire comprising 42 life events (see Supplementary Appendix) that was derived for ALSPAC using previous inventories as a basis for item selection (Barnett, Hanna, & Parker, 1983; Brown and Harris, 1978). Mothers completed the questionnaire at three time points when the study children were approximately 1 year 9 months, 2 years 9 months, and 3 years 11 months. At the first time point mothers were asked, “Have any of these (life events) occurred since the baby was 8 months old?”; at the second time point, “Have any of these occurred since the study child was 18 months old?”; and at the third time point, “Have any of these occurred since the study child was 2½ years old?”. This allowed us to examine exposure to stressful life events in three time periods: 8 months to approximately 1 year 9 months; 18 months to approximately 2 years 9 months; and 2½ years to approximately 3 years 11 months.

and at the third time point, “Have any of these occurred since the study child was 2½ years old?”. This allowed us to examine exposure to stressful life events in three time periods: 8 months to approximately 1 year 9 months; 18 months to approximately 2 years 9 months; and 2½ years to approximately 3 years 11 months. Outcome: Latent Classes of Bedwetting At ages 41/2, 51/2, 61/2, 71/2, and 91/2 years (hereafter referred to as 4–9 years), parents were asked, “How often usually does your child wet the bed?” and were given the options “never”; “less than once a week”; “about once a week”; “2–5 times a week”; “nearly every night”; “more than once a night.” Latent classes of bedwetting were previously derived by collapsing the bedwetting frequency data at each time point into three-level ordinal variables indicating no current bedwetting, infrequent bedwetting (less than once a week or about once a week) and frequent bedwetting (two to five times a week, nearly every night, or more than once a night). The latter category corresponds to the frequency of bedwetting required for a DSM-V diagnosis of nocturnal enuresis. Full details of the derivation of the latent classes are in (Sullivan et al., 2015).

about once a week) and frequent bedwetting (two to five times a week, nearly every night, or more than once a night). The latter category corresponds to the frequency of bedwetting required for a DSM-V diagnosis of nocturnal enuresis. Full details of the derivation of the latent classes are in (Sullivan et al., 2015). The latent classes describe typical and atypical development of nighttime bladder control: “normative development” (71.5% of the sample)—low probability of bedwetting at any time point; “infrequent delayed” (14.3%)—delayed attainment of nighttime bladder control and decreasing probability of infrequent bedwetting from 4 to 9 years; “infrequent persistent” (8.6%)—relatively high probability of infrequent bedwetting from 4 to 9 years; “frequent delayed” (2.4%)—high probability of frequent bedwetting at age 4 years, which decreased and became more infrequent at 6–9 years; “frequent persistent” (3.2%)—relatively high probability of bedwetting at least twice a week from 4 to 9 years.

latively high probability of infrequent bedwetting from 4 to 9 years; “frequent delayed” (2.4%)—high probability of frequent bedwetting at age 4 years, which decreased and became more infrequent at 6–9 years; “frequent persistent” (3.2%)—relatively high probability of bedwetting at least twice a week from 4 to 9 years. Potential Confounders We adjusted for potential confounders that are reported in the empirical literature to be risk factors for stressful life events and bedwetting. These included a range of socio-demographic measures and maternal depression. The socio-demographic measures were derived from responses to a questionnaire, completed by mothers during the antenatal period, containing items on socioeconomic position and adversity. Binary variables were generated from these questions and each item was scored as 1 if an adversity was present and 0 if not. The items included social class based on the lower of the mother or partner's occupational social class using the 1991 British Office of Population and Census Statistics classification and dichotomized into non-manual (professional, managerial, or skilled professions) and manual (partly or unskilled occupations); early parenthood (<19 years vs. 19 years), housing adequacy (yes/no—comprising crowding, periods of homelessness, living conditions, major defects/infestation), maternal education (defined as none vs. high school qualifications or greater), major financial difficulties (yes/no), family size (less than three children vs. more than or equal to three children), and the presence of a social network (yes/no—comprising emotional support, practical/financial support). We adjusted for maternal depression using the Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden, & Sagovsky, 1987), which was completed by mothers when their study child was 8 months old. The EPDS was dichotomized at the standard cutoff (score > 12) used to indicate probable depressive disorder (Evans et al., 2001).

adjusted for maternal depression using the Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden, & Sagovsky, 1987), which was completed by mothers when their study child was 8 months old. The EPDS was dichotomized at the standard cutoff (score > 12) used to indicate probable depressive disorder (Evans et al., 2001). We also adjusted for gender and a maternal rating of the child’s developmental level at 18 months, which was assessed using a questionnaire developed by ALSPAC including items from the Denver Developmental Screening Test (Frankenburg, Dodds, Archer, Shapiro, & Bresnick, 1992). The questionnaire assesses four domains of development (fine motor, gross motor, communication, and social skills) and scores on each domain were adjusted for age in weeks, standardized (using a linear regression model and extracting the residuals), and reversed where appropriate so that high values on all scores reflected a lower level of development. We used a total development score derived from the sum of the scores on each domain.

scores on each domain were adjusted for age in weeks, standardized (using a linear regression model and extracting the residuals), and reversed where appropriate so that high values on all scores reflected a lower level of development. We used a total development score derived from the sum of the scores on each domain. Statistical Modeling We generated a total life events score for each time period by adding all events (presence of life event = 1, absence = 0) and then standardized the score to allow us to interpret the change in odds of membership to the bedwetting latent classes per standard deviation (SD) increase in life events score. We estimated the association between stressful events and membership of the bedwetting latent classes using a series of univariable multinomial logistic regression models and employing the normative latent class as the reference category. Models were then adjusted for the confounders described above. Parameter estimates were obtained using the “Modal ML” three-step method (Vermunt, 2010) implemented in Mplus. This has been shown to produce less-biased estimates than traditional three-step methods such as probability weighting, while avoiding the problem of covariates impacting on the measurement model itself (Asparouhov & Muthen, 2013). Bias-adjusted estimates were obtained using the Mplus “auxiliary (r3step)” command.

lus. This has been shown to produce less-biased estimates than traditional three-step methods such as probability weighting, while avoiding the problem of covariates impacting on the measurement model itself (Asparouhov & Muthen, 2013). Bias-adjusted estimates were obtained using the Mplus “auxiliary (r3step)” command. Results Bedwetting data were available for 10,810 children on at least one measurement occasion. Of these, 8,761 had data from at least three time points and 5,849 had complete data. The proportion of children with bedwetting decreased over time and proportions did not change markedly when the sample was restricted to participants with more data (Sullivan et al., 2015). For the analyses presented here, we focused on the sample with bedwetting data available from at least three time points (n = 8,761). Conclusions were consistent for the other two samples described earlier (available on request). While the sample with complete bedwetting data had the lowest rates of socioeconomic disadvantage, there was little variation in the other confounders and the risk factors across samples (Table I). Table I. Means and Proportions of Risk Factors and Confounders in the Three Samples Considered for the Analysis

request). While the sample with complete bedwetting data had the lowest rates of socioeconomic disadvantage, there was little variation in the other confounders and the risk factors across samples (Table I). Table I. Means and Proportions of Risk Factors and Confounders in the Three Samples Considered for the Analysis Risk factors and confounders Complete data (n = 5,489) Partially missing data (n = 8,761) At least one bedwetting measure (n = 10,810) Mean (SD) for stressful events score in each time period 8 months to ∼1 year 9 months 4.40 (2.81) 4.45 (2.87) 4.48 (2.93) 18 months to ∼2 years 9 months 4.87 (2.97) 4.92 (3.03) 4.96 (3.10) 2½ years to ∼3 years 11 months 4.56 (3.10) 4.61 (3.16) 4.66 (3.22) Confounders Gender (male) 2,959 (50.6%) 4,508 (51.46%) 5,580 (51.6%) Maturational level (mean total score at 18 months) 37.72 (5.5) 37.73 (5.6) 37.74 (5.7) Socioeconomic variables Manual social class 705 (12.8%) 1,238 (15.4%) 1,661 (17.3%) Early parenthood <20 years 213 (3.64%) 419 (4.8%) 648 (6.0%) Housing inadequacy 408 (7.1%) 748 (8.6%) 982 (9.3%) Low maternal education 580 (10.0%) 995 (11.6%) 1,342 (12.9%) Financial difficulties 803 (13.9%) 1,297 (15.1%) 1,653 (16.0%) Family size ≥ 3 251 (4.3%) 424 (4.9%) 600 (5.8%) Poor social network 716 (12.3%) 1,129 (13.0%) 1,433 (13.6%) Maternal depression (mean EPDS score at 8 months) 3.18 (2.89) 3.27 (2.95) 3.31 (2.99) Note. Actual number with available data on each risk factor varies in each of the samples shown here. Manual social class includes manual and part/unskilled.

4.9%) 600 (5.8%) Poor social network 716 (12.3%) 1,129 (13.0%) 1,433 (13.6%) Maternal depression (mean EPDS score at 8 months) 3.18 (2.89) 3.27 (2.95) 3.31 (2.99) Note. Actual number with available data on each risk factor varies in each of the samples shown here. Manual social class includes manual and part/unskilled. Association Between Stressful Events and Bedwetting Latent Classes Table II presents the results of the analysis examining the associations between stressful events experienced in three time periods and latent class membership. Odds ratios were derived in relation to the normative latent class of nighttime bladder control, which was used as the reference category in this analysis. The results show the increase in odds of membership to each latent class per 1 SD increase in the stressful events score. There was evidence that stressful events occurring in the first two time periods were associated with membership of the infrequent delayed, infrequent persistent, and frequent persistent classes, but not the frequent delayed class. For instance, in the unadjusted model a 1 SD increase in the stressful events score for the first time period was associated with a 29% (13–47%) increase in the odds of belonging to the frequent persistent class compared with the normative class. Adjustment for confounders led to a small attenuation of these effects, but there was still evidence that stressful events were associated with increased odds of membership to these classes. Odds ratios were generally highest for the frequent persistent class in the adjusted models, for example, a 1 SD increase in the stressful events score for the first time period was associated with a 27% (10–47%) increase in the odds of belonging to the frequent persistent class compared with the normative class in the adjusted model. Stressful events at the latest time period were only associated with increased odds of membership to the frequent persistent class. For example, a 1 SD increase in the stressful events score was associated with a 30% (11–52%) increase in the odds of belonging to the frequent persistent class compared with the normative class. Table II. Odds Ratios and 95% Confidence Intervals for the Association Between Stressful Events and Latent Class Membership

For example, a 1 SD increase in the stressful events score was associated with a 30% (11–52%) increase in the odds of belonging to the frequent persistent class compared with the normative class. Table II. Odds Ratios and 95% Confidence Intervals for the Association Between Stressful Events and Latent Class Membership Stressful events N Class Infrequent delayed Infrequent persistent Frequent delayed Frequent persistent Stressful events 8 months to ∼1 year 9 months Unadjusted 8,193 1.14 (1.04, 1.26) p = .008 1.21 (1.10, 1.32) p < .001 0.99 (0.81, 1.20) p = .911 1.29 (1.13, 1.47) p < .001 Adjusted 1: Gender, developmental level 7,827 1.15 (1.04, 1.28) p = .006 1.21 (1.10, 1.33) p < .001 0.99 (0.81, 1.20) p = .923 1.30 (1.14, 1.49) p < .001 Adjusted 2: Socio-demographic factors 7,266 1.14 (1.02, 1.28) p = .017 1.18 (1.07, 1.31) p = .001 0.98 (0.79, 1.21) p = .815 1.31 (1.14, 1.50) p < .001 Adjusted 3: Maternal depression 7,025 1.12 (1.00, 1.25) p = .042 1.16 (1.04, 1.30) p = .007 0.96 (0.77, 1.18) p = .689 1.27 (1.10, 1.47) p = .001 18 months to ∼ 2 years 9 months Unadjusted 7,635 1.17 (1.07, 1.28) p = .001 1.18 (1.08, 1.30) p < .001 0.97 (0.79, 1.18) p = .732 1.19 (1.02, 1.38) p = .032 Adjusted 1: Gender, developmental level 7,266 1.17 (1.06, 1.29) p = .002 1.21 (1.09, 1.33) p < .001 0.97 (0.79, 1.19) p = .784 1.20 (1.03, 1.40) p = .019 Adjusted 2: Socio-demographic factors 6,735 1.18 (1.06, 1.31) p = .002 1.21 (1.09, 1.34) p < .001 1.00 (0.80, 1.25) p = .998 1.22 (1.04, 1.43) p = .012 Adjusted 3: Maternal depression 6,519 1.17 (1.05, 1.30) p = .004 1.20 (1.08, 1.33) p = .001 0.98 (0.79, 1.21) p = .819 1.21 (1.03, 1.42) p = .024 2½ years to ∼ 3 years 11 months Unadjusted 8,196 1.08 (0.98, 1.18) p = .111 1.10 (1.00, 1.21) p = .049 0.84 (0.69, 1.03) p = .089 1.25 (1.10, 1.43) p = .001 Adjusted 1: Gender, developmental level 7,749 1.09 (0.99, 1.20) p = .067 1.12 (1.01, 1.24) p = .028 0.88 (0.72, 1.08) p = .223 1.29 (0.99, 1.67) p < .001 Adjusted 2: Socio-demographic factors 7,138 1.12 (1.01, 1.24) p = .033 1.10 (0.99, 1.22) p = .069 0.87 (0.70, 1.08) p = .193 1.33 (1.14, 1.54) p < .001 Adjusted 3: Maternal depression 6,900 1.10 (0.99, 1.22) p = .083 1.08 (0.97, 1.20) p = .169 0.85 (0.69, 1.06) p = .153 1.30 (1.11, 1.52) p = .001 Adjusted 1: Gender, developmental level at 18 months (total development score derived from the sum of the scores on each domain).

0, 1.08) p = .193 1.33 (1.14, 1.54) p < .001 Adjusted 3: Maternal depression 6,900 1.10 (0.99, 1.22) p = .083 1.08 (0.97, 1.20) p = .169 0.85 (0.69, 1.06) p = .153 1.30 (1.11, 1.52) p = .001 Adjusted 1: Gender, developmental level at 18 months (total development score derived from the sum of the scores on each domain). Adjusted 2: Social class (manual vs. nonmanual), early parenthood (<19 years vs. ≥ 19 years), housing adequacy (yes/no), maternal education (none vs. high school qualifications or greater), major financial difficulties (yes/no), family size (less than three children vs. More than or equal to three children), presence of a social network (yes/no). Adjusted 3: Maternal depression—Edinburgh Postnatal Depression Scale score at 8 months.

Adjusted 2: Social class (manual vs. nonmanual), early parenthood (<19 years vs. ≥ 19 years), housing adequacy (yes/no), maternal education (none vs. high school qualifications or greater), major financial difficulties (yes/no), family size (less than three children vs. More than or equal to three children), presence of a social network (yes/no). Adjusted 3: Maternal depression—Edinburgh Postnatal Depression Scale score at 8 months. Discussion Most children are expected to attain nighttime bladder control by 4–6 years (Fergusson et al., 1986). We find evidence that the risk of experiencing problems attaining bladder control at 4–9 years is greater if the child is exposed to stressful events in early childhood. These results are consistent with previous studies reporting a link between early stressful events and bedwetting (Douglas, 1973; Jarvelin et al., 1990; Kaffman & Elizur, 1977; Kalo & Bella, 1996). This, however, is the first prospective study to examine whether early stress is associated with distinct patterns of atypical development of nighttime bladder control. Our findings were mostly in agreement with a dose response relationship, in which increasing levels of exposure to early stress were associated with increasing severity (frequency and persistence) of bedwetting. We did not find evidence for distinct etiologies for the latent classes. It is possible that distinct risk factors would emerge if we further refined our bedwetting classes by incorporating additional symptom information such as concurrent daytime wetting and indicators of bladder dysfunction.

rsistence) of bedwetting. We did not find evidence for distinct etiologies for the latent classes. It is possible that distinct risk factors would emerge if we further refined our bedwetting classes by incorporating additional symptom information such as concurrent daytime wetting and indicators of bladder dysfunction. Strengths and Limitations The study is based on a large contemporary cohort and takes advantage of repeated measures of frequency of bedwetting and the availability of a range of confounders. We have extended previous work by using developmental trajectories of frequency of bedwetting throughout childhood as our outcomes rather than simply the presence or absence of bedwetting at a particular age.

takes advantage of repeated measures of frequency of bedwetting and the availability of a range of confounders. We have extended previous work by using developmental trajectories of frequency of bedwetting throughout childhood as our outcomes rather than simply the presence or absence of bedwetting at a particular age. We did not examine the effect of specific life events, nor did we distinguish between life events that are acute (e.g., death of parent) versus more chronic in nature (e.g., financial problems) or between normative stressors (e.g., marriage, pregnancy) and non-normative (unpredictable) stressors (e.g. accidents). Some studies have found evidence that single events, especially parental divorce or separation, are particularly important risk factors for enuresis (Jarvelin et al., 1990; Kaffman & Elizur, 1977). We additionally examined single stressful events (e.g., divorce and separation), but we did not find individual associations with bedwetting (available on request). It is difficult to isolate the effect of single stressful events because they are often interrelated (e.g., parental divorce/separation may be related to financial problems). Consistent with the early prospective study (Douglas, 1973), we find that it is the total burden of exposure to stressful events during early childhood, rather than any single event, that is important in determining risk for subsequent bedwetting.

e.g., parental divorce/separation may be related to financial problems). Consistent with the early prospective study (Douglas, 1973), we find that it is the total burden of exposure to stressful events during early childhood, rather than any single event, that is important in determining risk for subsequent bedwetting. In agreement with earlier studies examining stressful life events, we generated a total life events score for each time period by adding all events (Araya et al., 2009). This differs to the approach taken by Jarvelin et al. (1990), who weighted the life events according to the parents’ perceived “seriousness” of the event. There is evidence that appraisals of the negative impact of life events are systematically elevated in individuals with depression (Espejo, Hammen, & Brennan, 2012). A mother’s own subjective appraisal of the impact of a life event may not reflect the child's own experience of the impact.

ed “seriousness” of the event. There is evidence that appraisals of the negative impact of life events are systematically elevated in individuals with depression (Espejo, Hammen, & Brennan, 2012). A mother’s own subjective appraisal of the impact of a life event may not reflect the child's own experience of the impact. In this article, we focused on exposure to stressful life events in early childhood and we did not examine the effects of more proximal stressors on risk for bedwetting. There is evidence that proximal exposure to stressful life events increases the risk of secondary enuresis (relapse in bedwetting after a period of at least 6 months’ dryness; Fergusson, Horwood, & Shannon, 1990; Jarvelin et al., 1990). There is also a possibility that the association between stress and bedwetting is bidirectional. The occurrence of bedwetting after the age at which most children would be expected to be dry at night can often lead to increased levels of stress in the family. Intolerance and punishment are not uncommon among parents of children with incontinence (Butler & McKenna, 2002) and such reactions could increase stress in the child, resulting in further episodes of bedwetting.

t children would be expected to be dry at night can often lead to increased levels of stress in the family. Intolerance and punishment are not uncommon among parents of children with incontinence (Butler & McKenna, 2002) and such reactions could increase stress in the child, resulting in further episodes of bedwetting. Potential Mechanisms Explaining the Link Between Stressful Events and Bedwetting Bedwetting is conceptualized as a “biobehavioral” problem (Houts, 1991) with biological, behavioral, and psychosocial factors contributing to the etiology. The age of 2–4 years is believed to be a sensitive period for learning bladder control (Douglas, 1973; Jarvelin et al., 1990; MacKeith, 1968). It is possible that parents experiencing high levels of stress may have less time and sensitivity to cope with the demands of toilet training (Jarvelin et al., 1990), and this could adversely affect their child’s transition to continence. There is evidence that delayed or inadequate toilet training is associated with an increased risk of bladder dysfunction (e.g., urgency, urge incontinence, emptying difficulties, bladder instability, and/or dyscoordinated micturition; Bakker & Wyndaele, 2000; Hellstrom, 2000; Hodges, Richards, Gorbachinsky, & Krane, 2014). Bladder dysfunction is not only a contributing factor for daytime wetting, but it is also associated with bedwetting (Franco, von Gontard, & DeGennaro, 2013; Nevéus et al., 2000). There is also evidence that the stress hormone cortisol suppresses the release of antidiuretic hormone (Bähr, Franzen, Oelkers, Pfeiffer, and Diederich, 2006), a lack of which leads to polyuria (increased volume of urine in the bladder; Aikawa, Kasahara, Uchiyama, 1998).

ontard, & DeGennaro, 2013; Nevéus et al., 2000). There is also evidence that the stress hormone cortisol suppresses the release of antidiuretic hormone (Bähr, Franzen, Oelkers, Pfeiffer, and Diederich, 2006), a lack of which leads to polyuria (increased volume of urine in the bladder; Aikawa, Kasahara, Uchiyama, 1998). It is notable that higher levels of stressful events at the latest time point (up to around age 4 years) were associated with the frequent persistent class, but not the infrequent bedwetting classes. There is evidence that during the period of transition to bladder control, the risk of developing continence problems may be greater if the transition period is prolonged and if the child is older (Hellstrom, 2000). Later initiation of toilet training could prolong the child’s exposure to potential stressors and this could interfere with the process of learning bladder control. We examined whether toilet training was initiated later (after 24 months) among children in the frequent persistent class compared with the infrequent bedwetting classes. The proportions with later initiation of toilet training were 42.2% in the frequent persistent class compared with 36.8% in the infrequent delayed class and 41.2% in the infrequent persistent class. To properly examine this, we would need to further refine the late initiation group to identify those with toilet training initiated after 2½ years, after 3 years, and beyond, but the data did not permit this.

istent class compared with 36.8% in the infrequent delayed class and 41.2% in the infrequent persistent class. To properly examine this, we would need to further refine the late initiation group to identify those with toilet training initiated after 2½ years, after 3 years, and beyond, but the data did not permit this. There was a relative lack of evidence for associations between the risk factors and the frequent delayed class. This might be owing to this being the smallest class, leading to larger standard errors and hence, a lack of precision in our estimates of the effect of the risk factors on membership to this class. Alternatively, it is possible that this pattern of wetting has a slightly different etiology. Three central nervous system (CNS) factors are believed to contribute to the development of enuresis: nocturnal polyuria (production of abnormally large amounts of urine at night), a lack of arousal, and a lack of inhibition of the emptying reflex of the bladder during sleep (von Gontard & Neveus, 2006). Overall, bedwetting has a spontaneous remission rate of approximately 15% per year (Forsythe & Redmond, 1974). In the delayed classes, this rate is lower, that is, children take a longer time to achieve continence. It could be speculated that this is a less severe condition (characterized by a maturational delay) in contrast to the persistent classes (characterized by a maturational disorder). With increasing maturational development, children in the frequent delayed class would be expected to progressively attain improved bladder stability, increased CNS recognition of bladder fullness, and the ability to suppress bladder contractions (Watanabe & Azuma, 1989). Children are more easy to arouse with increasing maturity (Busby, Mercier, & Pivik, 1994), and lack of arousal is one of the possible pathophysiological aspects of nocturnal enuresis (Koff, 1996; Wolfish, Pivik, & Busby, 1997). From a clinical perspective, this developmental delay would be expected to resolve over time and these children would eventually attain nighttime continence, albeit at a later age than their typically developing peers.

ossible pathophysiological aspects of nocturnal enuresis (Koff, 1996; Wolfish, Pivik, & Busby, 1997). From a clinical perspective, this developmental delay would be expected to resolve over time and these children would eventually attain nighttime continence, albeit at a later age than their typically developing peers. It is possible that the effects of stressors could be mediated through comorbid behavioral and emotional problems. Although some earlier studies (many based on small samples) find no increased rates of psychological problems in children who wet the bed, there is now a wide and growing literature providing evidence that children with bedwetting have higher levels of both internalizing (e.g., depression and anxiety) and externalizing symptoms (e.g., attention/activity and conduct problems) (Joinson, Heron, Emond, & Butler, 2007). Stressors not only increase the risk for relapses in bedwetting (see von Gontard, Baeyens, Van Hoecke, Warzak, & Bachmann, 2011a for a review), but also the risk of psychological disorders in children (Harland et al., 2002). These, in turn, could be responsible for the persistence of enuresis. A review paper cites evidence linking psychological distress to bladder dysfunction (Cortes, Sahai, Pontari, & Kelleher, 2012). The authors suggest that this link may be owing to alterations in neurotransmitters having direct or indirect effects on bladder function.

could be responsible for the persistence of enuresis. A review paper cites evidence linking psychological distress to bladder dysfunction (Cortes, Sahai, Pontari, & Kelleher, 2012). The authors suggest that this link may be owing to alterations in neurotransmitters having direct or indirect effects on bladder function. Gene–environment interactions also need to be considered. Previous analyses of the ALSPAC cohort have demonstrated that the risk for bedwetting is significantly increased if parents were also affected by enuresis (von Gontard, Heron, & Joinson, 2011b). With a heritability of 0.7, bedwetting is a highly genetically determined disorder (von Gontard et al., 2001). This genetic risk is present in all types of enuresis and can be modulated and activated by stressors (von Gontard et al., 2001).

parents were also affected by enuresis (von Gontard, Heron, & Joinson, 2011b). With a heritability of 0.7, bedwetting is a highly genetically determined disorder (von Gontard et al., 2001). This genetic risk is present in all types of enuresis and can be modulated and activated by stressors (von Gontard et al., 2001). Conclusions There is well-established evidence that early exposure to stress interferes with brain development and is associated with a range of adverse developmental and health outcomes (Shonkoff & Garner, 2012). Our findings add to the evidence that continence is a developmental outcome that is associated with exposure to high levels of stress in the family. Family stress has also been found to be an important prognostic factor in the treatment of enuresis among children referred to a community-based enuresis clinic (Devlin & O’Cathain, 1990). Health practitioners have a role to play in identifying families experiencing high levels of stress and directing them to appropriate sources of support. There is a need to educate parents about the long-term consequences of stress and the potential benefits of preventing or reducing sources of stress in early childhood (Garner & Shonkoff, 2012). Families at high risk of stress should be offered anticipatory guidance, especially for sensitive developmental transition periods such as toilet training, because they may find the demands even more challenging. Provision of support to families under stress could help parents navigate sensitive transitions in child development and promote healthy development during the early years.

idance, especially for sensitive developmental transition periods such as toilet training, because they may find the demands even more challenging. Provision of support to families under stress could help parents navigate sensitive transitions in child development and promote healthy development during the early years. Supplementary Data Supplementary data can be found at: http://www.jpepsy.oxfordjournals.org/. Supplementary Data Acknowledgments This study is based on the ALSPAC. We are extremely grateful to all the families who took part in this study, the midwives for their help in recruiting them, and the whole ALSPAC team, which includes interviewers, computer and laboratory technicians, clerical workers, research scientists, volunteers, managers, receptionists, and nurses. The UK Medical Research Council, the Wellcome Trust (grant reference: 102215/2/13/2) and the University of Bristol provide core support for ALSPAC. Funding This research was specifically funded by a grant from the Medical Research Council (Increasing understanding of risk factors and outcomes associated with continence problems in children and adolescents. MRC reference: MR/L007231/1). Conflicts of interest: None declared.