Correlates of Subjective Caregiver Strain in Caregivers of Youth Evaluated in a Pediatric Psychiatric Emergency Room

Abstract

Background:

Although caregiver burden is relevant to the outcome for psychiatrically ill youth, most studies have focused on caregiver burden in the community or research settings. Therefore, we aimed at evaluating the subjective caregiver strain (SCS) at the time of presentation of youth to a pediatric psychiatric emergency room (PPER), assessing potential correlates to provide leads for improvements in formal support systems.

Methods:

In this retrospective cohort study, the internalized, externalized, and total SCS were assessed in caregivers of youth <18 years of age consecutively evaluated at a PPER during a 1 year period using the Caregiver Strain Questionnaire. Sociodemographic and a broad range of clinical data were collected during the PPER visit using a 12-page semistructured institutional evaluation form. The Appropriateness of Pediatric Psychiatric Emergency Room Contact scale, incorporating acuity, severity and harm potential, was used to rate appropriateness of the visit.

Results:

In caregivers of 444 youth, the internalized SCS was significantly higher than the externalized SCS (p < 0.001). Multivariable analyses indicated that higher total and externalized SCS were associated with disruptive behavior or substance abuse/dependent disorder diagnosis, presenting complaint of aggression, and discharge plan to the police. Higher total and internalized SCS were associated with lower child functioning, whereas total and internalized SCS were lower in adopted children. In addition, higher externalized SCS was associated with investigator-rated inappropriateness of the emergency visit, presenting complaint of defiance, and a lack of prior psychiatric ER visits.

Conclusions:

High levels of CS in PPER highlight the necessity to adhere to existing guidelines regarding the inclusion of caregivers' perceptions into comprehensive psychiatric assessments. The particularly high strain in caregivers of children with externalizing disorders and in families with low-functioning youth may need to prompt PPER staff to provide efficient information on appropriate treatment options for these children and on support facilities for the parents.

Introduction

The interaction between childhood psychopathology and family burden is a topic of considerable constant interest in child and adolescent psychiatry (Johnston and Mash 2001; Bussing et al. 2003a,b; Anderson 2008; Gerkensmeyer et al. 2011). A number of studies have shown that child pathology has a greater impact on family burden than demographic variables or parental characteristics, and that the burden is predictive of mental health service utilization. Angold et al. (1998) reported that the number of children with psychiatric disorders who were evaluated by a mental health specialist increased by fivefold with increased parentally perceived family burden, and, at every level of the child's psychopathology, the presence of parental perceived burden was a powerful predictor of specialty mental health services use. Caregiver strain (CS), which can be divided into objective (personal time and work interruptions) and subjective (emotional reactions) components has also been associated with use of nonspecialty services including emergency room (ER) visits (Chavira et al. 2009), and with early referrals to mental health services (Sayal 2004).

Studies in attention-deficit/hyperactivity disorder (ADHD) especially with comorbid disruptive behavior disorders, with oppositional defiant disorder (ODD) or conduct disorder (CD) symptoms show particularly high parental burden (Evans et al. 2009; Oruche et al. 2015). Similarly, in autism spectrum disorders, child problem behaviors were more strongly associated with parental burden than were reductions in daily living skills (Estes et al. 2009). In a study of parents of children and adolescents without psychiatric disorders, exposure to enacted child disruptive behavior disorders significantly increased negative affect and alcohol consumption (Pelham et al. 1997), consistent with a line of research showing a strong influence of pediatric externalized aggression on caregiver strain. By contrast, pediatric inwardly directed aggression, in particular suicidal behavior, resulted in high levels of internalized CS (Barksdale et al. 2009). Additional studies have also demonstrated high family burden as a result of other childhood psychiatric disorders, including obsessive compulsive disorder (OCD) and posttraumatic stress disorder (Kalra et al. 2008), pervasive developmental disorder (Mugno et al. 2007), autism spectrum disorders (Estes et al. 2009), Tourette's disorder (Cooper et al. 2003), bipolar disorder (Woods 2000; Perlick et al. 2001, 2007), and depression (Wu et al. 2001). However, comparative studies across a representative array of diagnostic categories are scarce (Abbeduto et al. 2004; Cadman et al. 2012).

In addition to the adverse effect of the children's psychiatric diagnosis on caregiver burden, demographic (Horwitz and Reinhard 1995; Harrison and Sofronoff 2002; Garland et al. 2005) and medical factors (Chavira et al. 2009) have been shown to interact with the development of CS, which in turn interacts with mental health service use. In particular, minority families with high-risk youth and high CS experience significant difficulty in obtaining access to mental health services, especially proper outpatient care (Garland et al. 2005). By contrast, among families whose child was already involved in psychiatric care, Caucasian parents reported substantially greater CS than parents of minorities, even after controlling for income, age, and sex (Horwitz and Reinhard 1995).

A dynamic worsening of family interactions and relationships as a function of increasing CS and child psychopathology has repeatedly been demonstrated (Barkley et al. 1985; Ferro et al. 2000; Kalra et al. 2008; Kouros and Garber 2010). Importantly, this interaction appears to be bidirectional, with a worsening of both the child's and the parents' psychopathology over time (Weissman et al. 2006; Kouros and Garber 2010). Nevertheless, this dynamic may be moderated by parents' coping strategies (Higgins et al. 2005; Duchovic et al. 2009), which in turn may be supported through mental health services (Kazdin and Whitley 2003).

Similarly, treatment of the child is expected to modulate CS. Many studies assessed childhood psychopathology and caregiver burden at the time of treatment initiation in mental care services. An obvious, yet important, question is whether symptom relief during longitudinal treatment would result in a decrease in CS. Blader et al. (2006) reported that parents of children whose externalizing behaviors improved over 12 months reported higher stress in the beginning of treatment, but also endorsed a significant reduction in stress levels at study end-point. By contrast, parents who reported low stress at baseline had children who displayed less improvement, despite similar baseline levels of externalizing behaviors. Similarly, another 1 year study of children and adolescents with ADHD and comorbid oppositional symptoms reported that children whose ODD/CD symptoms decreased had parents who reported a significant reduction in CS (Evans et al. 2009). Taken together, these findings suggest that parental concern and distress can be both dysfunctional and functional for the child's outcome. Unless the family is paralyzed by or disintegrates as a result of the child's problems, or the child's problems are fueled by parental discord and strain, more acutely distressed and concerned parents might be more motivated to participate in treatment and aid a child in improving symptomatology.

Earlier studies of caregiver burden have been conducted within the community or research settings; that is, with families engaged or starting treatment (Bussing et al. 2003a,b; Blader 2006; Chavira et al. 2009). By contrast, we obtained measures of subjective CS (SCS) at the time of presentation to a pediatric psychiatric emergency room (PPER), as this represents a unique help-seeking situation for a mixed cohort of families with and without prior treatment. Understanding potential correlates of CS will hopefully provide leads for improvements in formal support systems. Based on the available literature, we hypothesized that higher SCS would not only be associated with the child's psychopathology but also with the parental demographic profile and psychiatric history.

To test these hypotheses, we performed multivariable predictor analyses of CS scores (total, internalized, and externalized) with both child- and parent- related characteristics. As a secondary aim, we also compared internalized and externalized caregiver scores.

Methods

Design and procedures

Data used in this study were collected as part of a retrospective chart review study of a consecutive cohort of pediatric psychiatric emergency room service (PPERS) visits by patients <18 years old, consecutively evaluated at the Long Island Jewish Medical Center (LIJMC) pediatric ER during a 12 month period, between January 1, 2002 and December 31, 2002. Repeat visits for a psychiatric evaluation by the same patient were considered as separate events, each representing a unique PPERS utilization pattern, whereas nonpsychiatric emergency department (ED) visits were not captured in the study. Also included were visits of patients directly admitted to the medical floors because of the severity of their psychiatric presentation (i.e., suicide attempt or severe self-mutilation). Evaluations of these patients were performed by the pediatric psychiatric consultation liaison service. These patients were specifically included to avoid excluding the most severe PPERS visits in the sample. Seventeen PPERS visits were excluded because patients were >18 years old. This study was approved by the LIJMC Institutional Review Board. Additional details are described elsewhere (Soto et al. 2009).

Setting

LIJMC is a tertiary care metropolitan hospital located in Queens, New York. It is the largest facility and the only academic teaching hospital in its catchment area that covers Queens and parts of Brooklyn and Long Island, New York.

In 2002, there were 21,749 child and adolescent visits, of which 4.9% were seen by the PPERS. Children and adolescents with psychiatric problems are initially evaluated by the pediatric medical ER staff and then referred to the PPERS for further psychiatric evaluation. During regular business hours, patients evaluated by the PPERS were seen by child and adolescent psychiatry residents supervised by child and adolescent psychiatry attending physicians. After hours, patients were evaluated by general psychiatry residents supervised by child and adolescent fellows with additional psychiatric attending backup coverage as needed. Patients evaluated by the pediatric psychiatric consultation liaison service were always seen by child and adolescent psychiatry fellows with supervision from psychiatry attending physicians.

Data sources and collection

The PPERS evaluation package consisted of a one page ER “face sheet” (a registration document with demographic and insurance information), the triage form, and a 12 page semistructured institutional child and adolescent clinical evaluation intake form. This comprehensive institutional intake form was used for all patients, whether or not they were discharged from the ER. The primary data source for the 18 consultations of patients directly admitted to the medical floors status postintoxication/serious self-harm was a two page consultation form. Both the institutional intake and consultation form include in-depth data on age; sex; race; gender; insurance status; arrival mode; presenting time and date; presenting problem; precipitant; referral mode; prior outpatient consultation; current and past psychiatric history and treatment, including history of abuse and neglect; family psychiatric history; clinical Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV) diagnoses (excluding rule-out diagnoses) (American Psychiatric Association 1994); current, highest, and lowest Global Assessment of Functioning (GAF) score in the past year; and disposition/recommendation. The qualitative field “identifiable precipitant” from the clinical evaluation form was collapsed as a categorical variable: “Family conflict,” “peer conflict,” “school conflict,” “trauma,” “treatment non-adherence,” “other,” and “no identifiable precipitant.” When multiple precipitants were identified, the most clinically significant/severe one was used in the analyses.

We restricted the analyses to the subsample of patients (n = 444, 41.4%), for whom caregivers completed the 10 page social and developmental history questionnaire (including additional data on demographics, past and current psychiatric history and treatment, family psychiatric history, and recent family stressors) as well as the Caregiver Strain Questionnaire (CSQ) (Brannan et al. 1997) at the time of the PPER visit. (However, patients included in the analyses were compared against the rest of the sample in order to assess for generalizability of the results, at least based on patient characteristics).

The CSQ is a 21 item self-report instrument designed to assess parents' perceptions of the objective and subjective burden of caring for a child with emotional and behavioral problems. The CSQ measures three different but related dimensions of CS. The Objective Strain Scale (11 items) assesses concrete adverse effects attributed to the child's disorder, such as interruption of personal time and work and disruption of family routine. The Subjective Caregiver Strain scale (10 items) measures parental emotional reactions resulting from the child's difficulties. Factor analyses (Brannan et al. 1997) identified two SCS subscales: the subjective internalized subscale (six items) and the subjective externalized subscale (four items). The subjective internalized subscale assesses inwardly directed negative feelings, such as worry, guilt, sadness and fatigue; the subjective externalized subscale measures outwardly directed negative feelings, such as anger, resentment, and embarrassment about the child's problems, respectively. Responses are scored on a five point Likert scale ranging from 1 (not at all a problem) to 5 (very much a problem), and the global score as well as the total subscale scores are calculated as the mean of the items. The subjective subscale's internal consistency and convergent and predictive validity have been demonstrated previously (Brannan et al. 1997). In the current study, only the subjective subscale was used.

Further, each PPERS visit was rated by the investigators for 1) severity, 2) acuity, and 3) harm potential on a seven point scale modeled after the Clinical Global Impressions (CGI) scale (Soto et al. 2009). Finally, patient visits were also rated for appropriateness using the Appropriateness of Pediatric Psychiatric Emergency Room Contact (APPERC) scale (Soto et al. 2009) that takes into account the severity, acuity, and harm potential ratings described previously.

Data analysis

Data were analyzed with analysis of variance (ANOVA) for continuous variables and χ² test or Fisher's exact tests for categorical variables. Given the large number of assessed variables, univariable analyses were only conducted with the aim of identifying significant variables that would be entered into the explanatory multivariable analyses. Three separate multivariable backward elimination linear regression analyses were conducted to identify variables that independently contributed to significant levels of total, externalized, and internalized strain in caregivers of children and adolescents assessed in the psychiatric ER. Only variables that had a univariable p value of <0.05 (Table 2) were entered into the initial models. For categorical variables with more than two different values, we first conducted an ANOVA analysis and only entered individual variables when the omnibus ANOVA test results as well as the post-hoc variable level test results were significant. Because of the number of analyses and comparisons, in the Results and Discussion sections we focused only on significant variables in univariable analyses that were also independent correlates of CS in the multivariable models.

Data were analyzed with JMP 5.0.1, 1989–2003 (SAS Institute, Inc, Cary, NC). All tests were two sided, with α set at 0.05.

Results

Sample characteristics

Demographic, illness, treatment, visit, and disposition characteristics are listed in Table 1. Of 1062 youth evaluated between January 1, 2002 and December 31, 2002; caregivers of 444 (41.4%) children and adolescents (mean age = 13.6 ± 3.1 years, 51.8% male, 53.7% Caucasian) completed the internalized, externalized, and total SCS self-assessment.

Table 1.

Demographic, Illness, Treatment, and Visit Characteristics of Youth Presenting to a Pediatric Psychiatric Emergency Room Service

Baseline characteristics
Male gender (n, %) (444)	230 (51.8)
Age (years ± SD) (444)	13.6 ± 3.1
Ethnicity (419^a)
Caucasian (n, %)	225 (53.7)
African American (n, %)	122 (29.1)
Hispanic (n, %)	48 (11.5)
Asian (n, %)	19 (4.5)
Multiethnic (n, %)	5 (1.2)
Living in foster care (n, %) (433^a)	26 (6.0)
Adopted (n, %) (430^a)	34 (7.9)
Insurance status (362^a)
Private/Managed care (n, %)	260 (71.8)
Medicaid (n, %)	70 (19.3)
None (n, %)	32 (8.8)
Psychiatric family history (n, %) (408^a)	275 (67.4)
Anxiety disorders (n, %)	91 (22.0)
Mood disorders (n, %)	160 (38.7)
Schizophrenia (n, %)	43 (10.4)
ADHD (n, %)	66 (15.9)
Disruptive behavior disorders (n, %)	49 (11.8)
Substance abuse (n, %)	140 (33.6)
Special education (n, %) (415^*)	134 (32.3)
Current or past sexual abuse (n, %) (431^a)	25 (5.8)
Current or past physical abuse (n, %) (429^a)	42 (9.8)
Current or past ACS involvement (n, %) (443^a)	31 (7.0)
Primary psychiatric diagnosis (444)
Depression (n, %)	119 (26.8)
ADHD (n, %)	79 (17.8)
Disruptive behavior disorders (n, %)	63 (14.2)
Mood disorder NOS (n, %)	42 (9.5)
Adjustment disorder (n, %)	46 (10.4)
Bipolar Disorder (n, %)	31 (7.0)
Anxiety disorder (n, %)	20 (4.5)
Schizophrenia-spectrum disorder (n, %)	13 (2.9)
Other disorder (n, %)	10 (2.3)
Pervasive developmental disorders (n, %)	13 (2.9)
Lifetime substance abuse/Dependence disorder (n, %)	8 (1.8)
Number of psychiatric diagnoses (mean ± SD) (444)	1.8 ± 0.9
Current Global Assessment of Functioning (GAF) Scale score (mean ± SD) (424^a)	50.0 ± 12.5
Treatment characteristics
Number of nonpsychotropic medications (mean ± SD) (444)	0.2 ± 0.7
Number of current psychotropic medications (mean ± SD) (444)	0.8 ± 1.1
Psychotropic medications (444)
Antidepressants, (n, %)	110 (24.8)
Antipsychotics; (n, %)	89 (20.1)
Mood Stabilizer; (n, %)	56 (12.6)
Stimulants; (n, %)	60 (13.5)
Anxiolytic/Hypnotic; (n, %)	18 (4.1)
Arrival mode (434^a)
Family/Self/Friend	379 (87.33)
EMS/NYPD (n, %)	49 (11.29)
Agency/School/Therapist/Inpatient MD	6 (1.38)
Referring agent (439^a)
Family/Self (n, %)	249 (56.72)
School (n, %)	100 (22.78)
Therapist (n, %)	33 (7.52)
Psychiatrist (n, %)	23 (5.24)
Non-psychiatrist MD	17 (3.87)
Other (n, %)	17 (3.87)
Clinician-rated presenting complaint (443^a)
Defiance (n, %)	102 (23.02)
Aggression (n, %)	97 (21.90)
Suicidal ideation (n, %)	100 (22.6)
Depression/Anxiety (n, %)	53 (11.96)
Suicide attempt (n, %)	31 (7.0)
Self-injurious behavior (n, %)	28 (6.32)
Psychosis (n, %)	21 (4.74)
Other (n, %)	11 (2.48)
Exacerbation of preexisting condition (n, %) (444)	246 (55.41)
Identifiable precipitant (441^a)
None (n, %)	194 (43.99)
Family conflict (n, %)	136 (30.94)
Peer conflict (n, %)	52 (11.79)
School conflict (n, %)	16 (3.63)
Trauma (n, %)	19 (4.31)
Treatment nonadherence	12 (2.72)
Other (n, %)	12 (2.72)
APPERC Scale ratings (0–7)
Symptom acuity (mean ± SD)	5.6 ± 1.9
Symptom severity (mean ± SD)	4.9 ± 1.5
Harm potential (mean ± SD)	4.5 ± 1.7
Inappropriateness of the ER visit (n, %) (444)	293 (66.0)
Outpatient disposition (437^a)
Inpatient (n, %)	106 (23.9)
Home (n, %)	320 (72.1)
Group home (n, %)	2 (0.5)
Day hospital (n, %)	9 (2.0)
Referral to outpatient non-psychiatrist physician (n, %) (435^a)	10 (2.3)
Referral to outpatient psychiatrist (n, %) (435^a)	218 (49.2)
Referral to outpatient therapist (n, %) (435^a)	199 (44.9)
Referral to outpatient (NYPD) (n, %) (435^a)	8 (1.8)

Some variables contain fewer than 444 patients visits, indicating missing or unknown data.

ADHD, attention-deficit/hyperactivity disorder; ACS, Administration for Children's Services; NOS, not otherwise specified; SCS, subjective caregiver strain; EMS, emergency medical service; NYPD, New York Police Department; APPERC, Appropriateness of Pediatric Psychiatric Emergency Room Contact.

Most patients were Caucasian (53.7%) or African American (29.1%). A high proportion had a family history of psychiatric illness (67.4%), 32.3% were in special education, 7.9% were adopted, and 6.0% were in foster care.

The most common primary psychiatric diagnostic group in our study sample was mood disorders (43.3%); that is, depressive disorders (26.8%), mood disorder not otherwise specified (9.5%), and bipolar disorders (7.0%). Other primary diagnoses included ADHD (17.8%), disruptive behavior disorder (14.2%), adjustment disorder (10.4%), anxiety disorders (4.5%), schizophrenia spectrum disorders (2.9%), pervasive developmental disorders (3.1%), lifetime substance abuse/dependence disorders (1.8%), and other disorders (2.3%), and 9.8% of patients had a history of current or past physical abuse and 5.8% had a history of current or past sexual abuse. The mean ± SD GAF score was 50.0 ± 12.5. Most patients were brought to the PPERS by family, self, or a friend (87.3%). The chief complaints were defiance (23.0%), suicidal ideation (22.6%), and aggression (21.9%).

Extent and correlates of SCS

The total SCS score reported by caregivers was 3.1 ± 0.7. Internalized SCS was significantly higher than externalized SCS (3.8 ± 0.9 vs. 2.4 ± 0.8; p < 0.001).

Table 2 displays associations of CS measures and parental as well as child demographic characteristics, child's trauma history, and illness and treatment variables, as well as visit and disposition characteristics; these associations were only screened to select factors for the multivariable analyses.

Table 2.

Interactions of Subjective Caregiver Strain with Demographic, Family, Clinical and Treatment Characteristics

		Mean total SCS			Mean externalized SCS			Mean internalized SCS
Baseline characteristic	Subgroup	Variable present	Variable absent	p value ^a	Variable present	Variable absent	p value ^a	Variable present	Variable absent	p value ^a
Sex	Male (n = 230)	3.1 ± 0.74	3.1 ± 0.69	0.72	2.5 ± 0.85	2.4 ± 0.82	0.48	3.8 ± 0.88	3.8 ± 0.84	0.93
Age range	13–17 yrs (n = 289)	3.1 ± 0.68	3.1 ± 0.77	0.54	2.4 ± 0.82	2.5 ± 0.85	0.56	3.8 ± 0.80	3.7 ± 0.94	0.11
	6–12 yrs (n = 152)	3.1 ± 0.77	3.1 ± 0.68	0.40	2.4 ± 0.84	2.4 ± 0.83	0.77	3.7 ± 0.94	3.8 ± 0.81	0.088
	≤5 yrs (n = 3)	3.7 ± 1.1	3.1 ± 0.71	0.16	3.3 ± 0.90	2.4 ± 0.83	0.086	4.1 ± 1.3	3.8 ± 0.85	0.52
Ethnicity				0.11			0.049			0.19
	Caucasian (n = 225)	3.2 ± 0.65	3.1 ± 0.77		2.4 ± 0.78	2.4 ± 0.87	0.87	3.9 ± 0.79	3.7 ± 0.91
	African American (n = 122)	3.1 ± 0.77	3.1 ± 0.69		2.5 ± 0.86	2.4 ± 0.81	0.73	3.6 ± 0.89	3.8 ± 0.83
	Hispanic (n = 48)	3.1 ± 0.76	3.1 ± 0.70		2.5 ± 0.91	2.4 ± 0.81	0.77	3.8 ± 0.12	3.8 ± 0.4
	Asian (n = 19)	2.8 ± 0.78	3.1 ± 0.70		2.1 ± 0.70	2.5 ± 0.83	0.042	3.6 ± 1.16	3.8 ± 0.84
	Multiethnic (n = 5)	3.7 ± 0.81	3.1 ± 0.71		3.3 ± 0.99	2.4 ± 0.82	0.019	4.1 ± 0.77	3.8 ± 0.85
Family situation	Living in foster care (n = 26)	2.9 ± 0.81	3.1 ± 0.70	0.18	2.4 ± 0.80	2.4 ± 0.83	0.94	3.4 ± 1.01	3.8 ± 0.84	0.030
	Adopted (n = 34)	2.8 ± 0.67	3.1 ± 0.71	0.01	2.2 ± 0.68	2.4 ± 0.84	0.079	3.4 ± 0.89	3.8 ± 0.85	0.010
Insurance status				0.47			0.39			0.049
	Private (n = 260)	3.1 ± 0.65	3.0 ± 0.81		2.4 ± 0.91	2.4 ± 0.91		3.8 ± 0.80	3.6 ± 0.92	0.032
	Medicaid (n = 70)	3.0 ± 0.75	3.1 ± 0.69		2.3 ± 0.85	2.4 ± 0.81		3.7 ± 0.83	3.8 ± 0.84	0.35
	None (n = 32)	3.0 ± 0.94	3.1 ± 0.67		2.6 ± 1.03	2.4 ± 0.79		3.5 ± 1.08	3.8 ± 0.81	0.036
Education	Special education (n = 134)	3.1 ± 0.73	3.1 ± 0.70	0.99	2.5 ± 0.82	2.4 ± 0.84	0.14	3.7 ± 0.85	3.8 ± 0.86	0.15
Psychiatric family history	Any family psychiatric history (n = 275)	3.2 ± 0.70	3.1 ± 0.73	0.30	2.4 ± 0.82	2.5 ± 0.85	0.62	3.9 ± 0.81	3.7 ± 0.89	0.026
	Anxiety disorders (n = 91)	3.3 ± 0.71	3.1 ± 0.70	0.066	2.5 ± 0.89	2.4 ± 0.82	0.72	4.1 ± 0.77	3.8 ± 0.85	0.006
	Mood disorders (n = 160)	3.1 ± 0.72	3.1 ± 0.70	0.95	2.4 ± 0.83	2.5 ± 0.84	0.69	3.9 ± 0.87	3.8 ± 0.81	0.61
	Schizophrenia (n = 43)	3.1 ± 0.77	3.2 ± 0.70	0.51	2.4 ± 0.82	2.4 ± 0.84	0.88	3.7 ± 0.93	3.9 ± 0.82	0.33
	ADHD (n = 66)	3.2 ± 0.67	3.1 ± 0.71	0.72	2.4 ± 0.79	2.4 ± 0.84	0.97	3.9 ± 0.81	3.8 ± 0.84	0.53
	Disruptive behavior disorders (n = 49)	3.3 ± 0.64	3.1 ± 0.71	0.047	2.6 ± 0.83	2.4 ± 0.83	0.14	4.0 ± 0.63	3.8 ± 0.86	0.062
	Substance abuse (n = 140)	3.1 ± 0.66	3.1 ± 0.73	0.73	2.3 ± 0.84	2.5 ± 0.83	0.11	4.0 ± 0.71	3.8 ± 0.89	0.030
ACS involvement	History of ACS involvement (n = 31)	3.2 ± 0.82	3.1 ± 0.71	0.48	2.6 ± 0.94	2.4 ± 0.82	0.14	3.75 ± 0.95	3.8 ± 0.85	0.79
History of abuse	Any abuse (n = 55)	3.0 ± 0.80	3.1 ± 0.70	0.40	2.3 ± 0.96	2.4 ± 0.80	0.17	3.8 ± 0.87	3.8 ± 0.87	0.93
	Physical abuse (n = 42)	3.0 ± 0.80	3.1 ± 0.70	0.32	2.2 ± 0.93	2.4 ± 0.81	0.11	3.8 ± 0.91	3.8 ± 0.86	0.92
	Sexual abuse (n = 25)	3.1 ± 0.78	3.1 ± 0.71	0.70	2.4 ± 1.03	2.4 ± 0.82	0.85	3.7 ± 0.76	3.8 ± 0.87	0.64
Primary psychiatric diagnosis	Depressive disorder (n = 119)	3.0 ± 0.59	3.2 ± 0.75	0.044	2.2 ± 0.73	2.5 ± 0.85	0.0002	3.8 ± 0.70	3.8 ± 0.91	0.83
	Mood disorder NOS (n = 42)	3.0 ± 0.70	3.1 ± 0.72	0.54	2.3 ± 0.74	2.4 ± 0.84	0.37	3.8 ± 0.91	3.8 ± 0.85	0.88
	ADHD (n = 79)	3.3 ± 0.76	3.1 ± 0.70	0.021	2.6 ± 0.88	2.4 ± 0.82	0.015	3.9 ± 0.86	3.8 ± 0.86	0.14
	Disruptive behavior disorders (n = 63)	3.3 ± 0.78	3.1 ± 0.70	0.0047	2.8 ± 0.91	2.4 ± 0.81	0.0020	3.9 ± 0.92	3.8 ± 0.84	0.27
	Adjustment disorder (n = 46)	3.1 ± 0.75	3.1 ± 0.71	0.61	2.5 ± 0.85	2.4 ± 0.83	0.69	3.6 ± 0.97	3.8 ± 0.84	0.22
	Bipolar disorder (n = 31)	3.0 ± 0.54	3.1 ± 0.73	0.61	2.3 ± 0.65	2.4 ± 0.84	0.29	3.8 ± 0.71	3.8 ± 0.87	0.87
	Anxiety disorder (n = 20)	2.8 ± 0.72	3.1 ± 0.71	0.027	2.1 ± 0.68	2.4 ± 0.84	0.046	3.5 ± 1.05	3.8 ± 0.84	0.082
	Psychotic spectrum disorder (n = 13)	2.8 ± 0.65	3.1 ± 0.71	0.10	2.2 ± 0.75	2.4 ± 0.83	0.23	3.4 ± 0.71	3.8 ± 0.86	0.11
	Other disorder (n = 10)	2.7 ± 0.93	3.1 ± 0.71	0.085	2.5 ± 0.81	2.4 ± 0.83	0.93	3.0 ± 1.2	3.8 ± 0.84	0.0030
	Pervasive developmental disorders (n = 13)	3.3 ± 0.86	3.1 ± 0.71	0.47	2.5 ± 1.1	2.4 ± 0.82	0.75	4.0 ± 0.77	3.8 ± 0.86	0.38
	Substance abuse (n = 8)	3.8 ± 0.24	3.1 ± 0.71	0.0066	3.2 ± 0.51	2.4 ± 0.83	0.0065	4.4 ± 0.59	3.8 ± 0.86	0.061
Comorbid substance abuse/Dependence	Any substance abuse/dependence disorder (n = 54)	3.2 ± 0.70	3.1 ± 0.71	0.35	2.4 ± 0.85	2.4 ± 0.83	0.85	3.9 ± 0.76	3.8 ± 0.87	0.17
	Alcohol abuse/dependence (current/past) (n = 15)	3.3 ± 0.48	3.1 ± 0.72	0.29	2.7 ± 0.70	2.4 ± 0.83	0.54	4.0 ± 0.74	3.8 ± 0.86	0.24
	Cannabis abuse/dependence (current/past) (n = 37)	3.3 ± 0.65	3.1 ± 0.72	0.041	2.6 ± 0.80	2.4 ± 0.84	0.21	4.1 ± 0.72	3.8 ± 0.86	0.028
Past suicidal history	Suicide attempt history (n = 65)	3.1 ± 0.62	3.1 ± 0.73	0.60	2.3 ± 0.78	2.4 ± 0.84	0.29	3.8 ± 0.72	3.8 ± 0.87	0.87
	Self-injurious behavior history (n = 94)	3.2 ± 0.62	3.1 ± 0.74	0.40	2.4 ± 0.78	2.4 ± 0.86	0.96	3.9 ± 0.77	3.8 ± 0.88	0.18
Psychiatric treatment	History of psychiatric ER visit (n = 67)	3.0 ± 0.71	3.1 ± 0.72	0.050	2.2 ± 0.81	2.5 ± 0.84	0.0084	3.7 ± 0.84	3.8 ± 0.85	0.49
	History of inpatient psychiatric treatment (n = 100)	3.2 ± 0.67	3.1 ± 0.71	0.47	2.5 ± 0.86	2.4 ± 0.82	0.47	3.8 ± 0.73	3.8 ± 0.88	0.62
	History of outpatient psychiatric treatment (n = 207)	3.2 ± 0.67	3.0 ± 0.75	0.010	2.5 ± 0.82	2.3 ± 0.85	0.012	3.9 ± 0.77	3.7 ± 0.91	0.066
	Current outpatient psychiatric treatment (n = 319)	3.1 ± 0.68	3.1 ± 0.79	0.50	2.4 ± 0.80	2.4 ± 0.91	0.95	3.8 ± 0.81	3.7 ± 0.94	0.24
Psychotropic medication	Antipsychotic (n = 89)	3.1 ± 0.64	3.1 ± 0.73	0.99	2.4 ± 0.80	2.4 ± 0.84	0.84	3.8 ± 0.72	3.8 ± 0.89	0.82
	Mood stabilizer (n = 56)	3.2 ± 0.69	3.1 ± 0.72	0.54	2.5 ± 0.81	2.4 ± 0.84	0.62	3.9 ± 0.79	3.8 ± 0.87	0.58
	Stimulant (n = 60)	3.2 ± 0.67	3.1 ± 0.72	0.25	2.5 ± 0.84	2.4 ± 0.83	0.34	3.9 ± 0.81	3.8 ± 0.86	0.32
	Anxiolytic/hypnotic (n = 18)	3.1 ± 0.59	3.1 ± 0.72	0.95	2.3 ± 0.78	2.4 ± 0.84	0.48	3.9 ± 0.76	3.8 ± 0.86	0.56
Mode of arrival to ER				0.73			0.66			0.56
	Family/friend/self (n = 379)	3.1 ± 0.70	3.2 ± 0.79		2.4 ± 0.83	2.5 ± 0.86		3.8 ± 0.84	3.8 ± 0.93
	EMS/NYPD (n = 49)	3.2 ± 0.80	3.1 ± 0.70		2.5 ± 0.87	2.4 ± 0.83		3.8 ± 0.92	3.8 ± 0.84
	Agency/school (n = 6)	3.3 ± 0.77	3.1 ± 0.71		2.4 ± 0.88	2.4 ± 0.84		0.98 ± 0.40	0.85 ± 0.04
Referring agent				0.16			0.41			0.056
	Family/Self (n = 249)	3.2 ± 0.68	3.0 ± 0.75		2.4 ± 0.82	2.4 ± 0.86		3.9 ± 0.76	3.7 ± 0.94
	School (n = 100)	2.9 ± 0.77	3.2 ± 0.69		2.3 ± 0.87	2.4 ± 0.82		3.6 ± 1.02	3.9 ± 0.78
	Therapist (n = 33)	3.1 ± 0.70	3.1 ± 0.71		2.3 ± 0.85	2.4 ± 0.83		3.8 ± 0.86	3.8 ± 0.85
	Psychiatrist (n = 23)	3.1 ± 0.65	3.1 ± 0.72		2.5 ± 0.69	2.4 ± 0.84		3.7 ± 0.75	3.8 ± 0.86
	Other (n = 17)	3.2 ± 0.89	3.1 ± 0.71		2.7 ± 1.01	2.4 ± 0.83		3.7 ± 0.97	3.8 ± 0.85
	Non-psychiatrist MD (n = 17)	3.2 ± 0.72	3.1 ± 0.71		2.5 ± 0.85	2.4 ± 0.84		3.9 ± 0.86	3.8 ± 0.85
Clinician-rated presenting complaint				0.0022			<0.0001			0.34
	Defiance (n = 102)	3.2 ± 0.77	3.1 ± 0.70	0.091	2.6 ± 0.92	2.4 ± 0.80	0.012	3.8 ± 0.82	3.8 ± 0.86
	Suicidal ideation (n = 100)	2.9 ± 0.63	3.1 ± 0.73	0.047	2.2 ± 0.71	2.5 ± 0.85	0.0005	3.8 ± 0.82	3.8 ± 0.87
	Suicidal attempt (n = 31)	3.0 ± 0.53	3.1 ± 0.73	0.38	2.3 ± 0.15	2.4 ± 0.04	0.29	3.7 ± 0.67	3.8 ± 0.87
	Aggression (n = 97)	3.3 ± 0.75	3.1 ± 0.69	0.0011	2.7 ± 0.89	2.3 ± 0.80	0.0002	3.9 ± 0.86	3.8 ± 0.85
	Depression/Anxiety (n = 53)	3.0 ± 0.68	3.1 ± 0.71	0.14	2.3 ± 0.79	2.5 ± 0.84	0.11	3.7 ± 0.88	3.8 ± 0.84
	Self-injurious behavior (n = 28)	3.1 ± 0.68	3.1 ± 0.72	0.84	2.4 ± 0.67	2.4 ± 0.84	0.91	3.8 ± 0.89	3.8 ± 0.85
	Psychosis (n = 21)	2.8 ± 0.69	3.1 ± 0.71	0.046	2.2 ± 0.18	2.4 ± 0.04	0.14	3.5 ± 0.91	3.8 ± 0.85
	Other (n = 11)	3.1 ± 0.81	3.1 ± 0.71	0.91	2.4 ± 0.25	2.4 ± 0.04	0.87	3.8 ± 1.4	3.8 ± 0.84
Exacerbation of preexisting condition	Exacerbation (n = 246)	3.1 ± 0.70	3.1 ± 0.74	0.67	2.4 ± 0.83	2.4 ± 0.84	0.68	3.8 ± 0.82	3.8 ± 0.90	0.76
Identifiable precipitant				0.74			0.48			0.90
	Family conflict (n = 136)	3.1 ± 0.71	3.1 ± 0.71		2.4 ± 0.85	2.4 ± 0.82		3.8 ± 0.81	3.8 ± 0.88
	Peer conflict (n = 52)	3.0 ± 0.63	3.1 ± 0.72		2.2 ± 0.66	2.4 ± 0.85		3.7 ± 0.93	3.8 ± 0.85
	School conflict (n = 16)	3.2 ± 0.57	3.1 ± 0.72		2.6 ± 0.81	2.4 ± 0.83		3.7 ± 0.69	3.8 ± 0.86
	Trauma (n = 19)	3.0 ± 0.73	3.1 ± 0.71		2.2 ± 0.82	2.4 ± 0.83		3.9 ± 0.95	3.8 ± 0.85
	Other (n = 12)	3.2 ± 0.56	3.1 ± 0.72		2.4 ± 0.71	2.4 ± 0.83		3.9 ± 0.49	3.8 ± 0.87
	Treatment nonadherence (n = 12)	3.1 ± 0.71	3.1 ± 0.71		2.6 ± 0.56	2.4 ± 0.84		3.7 ± 1.01	3.8 ± 0.85
	None (n = 194)	3.2 ± 0.75	3.1 ± 0.68		2.5 ± 0.88	2.4 ± 0.79		3.8 ± 0.89	3.8 ± 0.83
Appropriateness of PPER visit^b	Very inappropriate (444)	3.2 ± 0.79	3.1 ± 0.67	0.13	2.6 ± 0.86	2.3 ± 0.81	0.0030	3.8 ± 0.96	3.8 ± 0.80	0.81
Patient disposition	Locus of care			0.59			0.63			0.66
	Inpatient (n = 106)	3.2 ± 0.63	3.1 ± 0.74		2.5 ± 0.80	2.4 ± 0.84		4.0 ± 0.73	3.7 ± 0.89
	Home (n = 320)	3.1 ± 0.04	3.0 ± 0.21		2.4 ± 0.85	2.4 ± 0.81		3.7 ± 0.87	3.6 ± 0.87
	Group home (n = 2)	3.4 ± 0.91	3.1 ± 0.73		2.9 ± 0.18	2.4 ± 0.84		3.8 ± 1.65	3.7 ± 0.87
	Day hospital (n = 9)	2.9 ± 0.82	3.1 ± 0.73		2.3 ± 0.88	2.4 ± 0.84		3.5 ± 0.79	3.8 ± 0.87
	Outpatient referrals
	Referral to outpatient psychiatrist (n = 218)	3.1 ± 0.68	3.1 ± 0.74	0.58	2.4 ± 0.81	2.4 ± 0.86	0.56	3.8 ± 0.80	3.8 ± 0.91	0.72
	Referral to outpatient therapist (n = 199)	3.2 ± 0.75	3.1 ± 0.69	0.24	2.5 ± 0.88	2.4 ± 0.78	0.19	3.8 ± 0.85	3.8 ± 0.86	0.49
	Referral to NYPD (n = 8)	3.6 ± 0.68	3.1 ± 0.71	0.043	3.3 ± 0.85	2.4 ± 0.82	0.0043	4.0 ± 0.77	3.8 ± 0.86	0.54
		t Ratio	R²	p value ^a	t Ratio	R²	p value ^a	t Ratio	R²	p value ^a
Number of psychiatric diagnoses (mean ± SD)		1.48	0.0049	0.14	1.35	0.004125	0.18	1.15	0.002989	0.25
Current Global Assessment of Functioning (GAF) Scale score (mean ± SD)		−2.07	0.0099	0.039	−0.07	0.000012	0.94	−3.42	−2.07	0.0099
Number of current psychotropic medications (mean ± SD)		0.14	0.000047	0.89	−0.17	0.000062	0.87	0.40	0.00037	0.69
Number of nonpsychotropic medications (mean ± SD)		0.13	0.000038	0.89	−0.03	0.000002	0.98	0.24	0.00013	0.81
Symptom acuity (mean ± SD)		0.01	0.000000017	0.99	0.35	0.00027	0.73	−0.32	0.00024	0.75
Symptom severity (mean ± SD)		1.99	0.0088	0.048	1.34	0.0040	0.18	2.01	0.0090	0.045
Harm potential (mean ± SD)		2.18	0.011	0.030	1.10	0.0027	0.27	2.56	0.015	0.011

Other primary diagnoses = no diagnosis given (10), eating disorder (7), parent-child relationship problems (6), Tourette's syndrome (3), grief reaction (2), personality disorder (2), nightmare disorder (2), mutism (1), amnesia disorder NOS (1), delirium (1), language disorder (1), encopresis (1), conversion disorder (1), pseudoseizures (1).

Bolded values indicate statistical significance of variables that were entered into the initial multivariable models.

Measured with the Appropriateness of Pediatric Psychiatric Emergency Room Contact (APPERC) (Soto et al. 2009) that takes into account severity, acuity, and harm potential ratings described previously.

ACS, Administration for Children's Services; ADHD, attention-deficit/hyperactivity disorder; EMS, emergency medical service; NOS not otherwise specified; NYPD, New York Police Department; PPER, pediatric psychiatric emergency room; SCS, subjective caregiver strain.

The following factors were not associated with CS: history of Administration for Children's Services (ACS) involvement, history of abuse, past suicidal history, number of current psychotropic medications, number of current non-psychotropic medications, number of psychiatric diagnoses at the time of the ER visit. Several associations were found, with only some being retained in multivariable models.

Multivariable analyses of factors independently associated with SCS

Significantly elevated levels of total CS were associated with a complaint of aggression (p = 0.0024), youths' non-adoption status (p = 0.0057), substance abuse/dependence (p = 0.0062), youths' lower GAF scores (p = 0.0201), disruptive behavior disorder diagnosis (p = 0.0221), and those discharged into New York Police Department (NYPD) custody (p = 0.0396) (r ² = 0.096, n = 416, p < 0.0001) (Table 3).

Table 3.

Multivariable Analyses: Variables Independently Associated with Subjective Strain of Caregivers of Youth Presenting to a Pediatric Psychiatric Emergency Room

Significant variables in multivariate analyses (in order of relative weight)	Estimate ^a	Standard error	Sum of squares	F ratio	p value
Total SCS
Total model estimate: r² = 0.096, n = 416, p < 0.0001
Complaint: Aggression	0.1273	0.0416	4.3339	9.3662	0.0024
Family situation: Adoption	−0.1723	0.0620	3.5780	7.7327	0.0057
Diagnosis: Substance abuse/dependence disorder	0.3580	0.1302	3.5002	7.5644	0.0062
Current GAF score	−0.0063	0.0027	2.5213	5.490	0.0201
Diagnosis: Disruptive behavior disorders	0.1170	0.0509	2.4427	5.2790	0.0221
Patient disposition: NYPD	0.2514	0.1218	1.9731	4.2641	0.0396
Externalized SCS
Total model estimate: r² = 0.147, n = 388, p < 0.0001
Complaint: Aggression	0.1516	0.0534	11.7408	19.1285	<0.0001
PPERS visit rated as inappropriate^b	0.1492	0.0436	7.2032	11.7358	0.0007
Diagnosis: Substance abuse/dependence disorder	0.4685	0.1627	5.0921	8.2962	0.0042
Diagnosis: Disruptive behavior disorder	0.1516	0.0626	3.6035	5.8710	0.0159
History of psychiatric ER visit	−0.1267	0.0530	3.5067	5.7133	0.0173
Chief complaint: Defiance	0.1125	0.0506	3.0323	4.9403	0.0268
Patient disposition: NYPD	0.3270	0.1499	2.9217	4.7601	0.0297
Internalized SCS
Total model estimate: r² = 0.050, n = 418, p < 0.0001
Current GAF score	−0.0126	0.0033	9.6400	13.927	0.0002
Family situation: Adoption	−0.2356	0.0757	6.7106	9.6933	0.0020

Positive estimate indicates direct relationship, negative estimate indicates inverse relationship.

Measured with the Appropriateness of Pediatric Psychiatric Emergency Room Contact (APPERC) (Soto et al. 2009) that takes into account severity, acuity, and harm potential ratings.

ER, emergency room; GAF, Global Assessment of Functioning; NYPD, New York Police Department; PPER, Pediatric Psychiatric Emergency Room; SCS, subjective caregiver strain.

Significantly elevated levels of externalized strain were noted in caregivers of youth who presented with aggression (p < 0.0001) or defiance (0.0268), whose ER visit was rated as inappropriate (p = 0.0007), who had a diagnosis of substance use disorders (p = 0.0042) or disruptive behavior disorders (p = 0.0159), who had not made a previous ER visit (p = 0. 0173), and who were discharged into NYPD custody (p = 0.0297) (r ² = 0.147, n = 388, p < 0.0001).

Significantly elevated levels of internalized CS were associated with youths' lower GAF scores (p = 0.0002) and non-adopted status (p = 0.0020) (r² = 0.050, n = 418, p < 0.0001).

Discussion

Previous studies highlighted the relevance of parental strain resulting from caring for youth with emotional and behavioral disorders on clinical outcomes and mental health service use, but most of them were performed in the community or in research settings (Angold et al. 1998; Garland et al. 2005; Blader 2006; Chavira et al. 2009; Evans et al. 2009). To our knowledge, this is the first study that measured SCS at time of presentation of youth to a PPER, and that explored sociodemographic variables and a wide array of clinical correlates of SCS and its components.

In this cohort of families presenting to a PPER, externalized SCS was most closely associated with externalizing behaviors (especially disruptive behavior disorder diagnosis and aggression as the presenting complaint) and with substance abuse/dependence diagnosis. Higher internalized SCS was associated with lower global functioning of the child. Surprisingly, parents who had adopted their children experienced lower levels of total and internalized SCS independently of the child's symptoms. In line with our study hypothesis and with earlier literature (as discussed in detail subsequently), we found that the child's psychopathology was predominantly predictive of CS. However, contrasting with our study hypothesis, we failed to identify parental characteristics in the prediction of CS.

It is possible that this lack of contribution of parental characteristics to CS results from the parental perspective in the acute situation of an ER. The array of presenting symptoms matched those reported earlier for PPER samples (Cloutier et al. 2010). The PPER represents a unique help-seeking situation, and, it is hoped, a successful entry point to mental health services for caregivers of patients with a wide range of different disorders and with different complaints. These complaints are perceived to be acute mostly by those accompanying or sending the child, with the majority of ER visiting parents fearing a worsening of symptoms (Kahn et al. 1973). In fact, this PPER sample included a third of families who had not been involved in psychiatric treatment earlier. As access to care is often delayed in psychiatry (Altamura et al. 2008; Fraguas et al. 2014; Hardy et al. 2015), the impact of CS in obtaining help can be interpreted as facilitating service use. Appreciating parental concern can thus be used productively to improve pathways into care and to engage caregivers positively.

In this regard, in our study, externalized CS was also associated with the estimated inappropriateness of the PPER visit. The perspective of the ER as one low-threshold entry to mental healthcare seems to conflict with the categorization of visits to be inappropriate. However, among the major reasons for inappropriate PPERS visits were direct referrals from school as well as visits resulting from the unavailability of a specialist appointment (Soto et al. 2009). Therefore, this association may be reflective of the shortage in adequate low-threshold services.

High levels of externalized CS in externalizing disorders have been reported in various other settings as well (Anastopoulos et al. 1992; Angold et al. 1998; Brannan et al. 2012; Rockhill et al. 2013). In several studies in adult populations, aggressive behavior and destructiveness were among the variables that most clearly associated with CS (Winefield and Harvey 1993; Dyck et al. 1999; Fischer et al. 2004; Kjellin and Ostman 2005; Rodrigo et al. 2013; Koutra et al. 2015). In children and adolescents with ADHD, defiance, along with oppositional and law-breaking behaviors, was more robustly related to CS than symptom severity (Evans et al. 2009). Finally, analyzing CS over time, Evans et al. (2009) identified externalizing behaviors as the most important predictors of high levels of CS, suggesting further that the prevention and timely treatment of externalizing behaviors in youth could reduce parental CS. In this vein, a bidirectional relationship has been suggested (Mash and Johnston 1983; Pierce et al. 1999), in that reduced CS levels might also be beneficial in reducing youth's externalizing problems (Blader 2006), but high levels of CS also predicted lower treatment responsivity (Accurso et al. 2015).

In the majority of earlier studies of CS in psychiatric conditions, data were obtained from cohorts of patients with one specific disorder, such as ADHD, anxiety disorder, Tourette's syndrome, or bipolar disorder (Perlick et al. 2001; Bussing et al. 2003b; Cooper et al. 2003; Mugno et al. 2007; Kalra et al. 2008); therefore, the independent effect of psychiatric diagnosis on CS has rarely been examined (Abbeduto et al. 2004; Cadman et al. 2012). Contrasting autism spectrum disorder and ADHD, Cadman and colleagues (2012) showed, that “unmet needs most associated with caregiver burden in both groups included depression/anxiety and inappropriate behavior.” Interestingly, whereas “inappropriate behavior” represents the shared feature of these two conditions and would be expected to highly contribute to CS, depression and anxiety are not part of the disease-defining symptoms for ADHD and autism, but still dominated the perceived parental burden in both conditions. This observation is in line with our findings, as the psychopathological phenomenology was more predictive of caregiver strain than the diagnosis per se.

Consistent with prior data in both community and clinical samples (Angold et al. 1998; Barksdale et al. 2009; Chavira et al. 2009) as well as data from parents caring for children with organic brain disorders (Kuratsubo et al. 2008; Stancin et al. 2008; Manskow et al. 2015), we found a significant inverse relationship between youth global functioning and both internalized and total CS. This association illustrates that healthcare systems fail to fully appreciate and address the needs of families caring for children with disability that is often chronic.

Caring for adopted children emerged as a predictor of lower total and internalized SCS. This observation is in line with an earlier study, which compared the experience of CS in biological parents to other relative caregivers, and also found lower levels of internalized CS in the latter group (Heflinger and Taylor-Richardson 2004). Opposite interpretations of an association of lower levels of biological closeness with lower SCS are possible. These include on the part of the adoptive parents either 1) lower levels of investment; that is, disengagement or 2) higher tolerance levels and acceptance of the behavioral and emotional problems. Further research should focus on adopted youth and adoptive parents studying risk and resilience factors for positive youth outcomes.

In line with earlier studies, we also found that the internalized aspect of CS exceeded the externalized (Brannan et al. 1997, Bussing et al. 2003a; María Brannan and Heflinger 2005). Whereas in adults objective strain seems to play a prominent role, in psychiatrically ill youth previous research indicated that caregivers are more prone to report subjective strain than objective burden (Brannan et al. 1997). In the study conducted by Angold et al. (1998) in a representative sample of 1015 children, caregivers most commonly acknowledged negative effects on their well-being and stigma, although the authors did not differentiate between the internalized and externalized dimensions. Our results underscore the need during a psychiatric assessment of a child in a PPER to explore parent-level factors with a specific focus on internalizing symptoms, such as worry, guilt, sadness. and fatigue possibly experienced by caregivers.

Limitations

Several limitations should be considered in evaluating the results of this study. Because of the retrospective nature of this study, we had to rely on the documentation that was part of clinical care. Nevertheless, the routine use of a 12 page institutional semistructured assessment package assured a certain standardization and increased chances of completeness of the data capture. Further, data on variables such as duration of illness and number of sick days in the previous year as well as the concurrent presence of a physical disorder shown in the literature to be associated with CS (Ostacher et al. 2008), were not collected in our sample. However, most studies underlined that, independently of these variables, symptomatology and functional impairment were the strongest predictors of CS.

In our multivariate analyses, the variables that remained significant in our model explained only a relatively small percentage of the variance of the three dimensions of SCS, implying the need for future studies exploring the possible association of other relevant variables. We did not obtain data regarding parental coping strategies, or on formal or informal support systems used by the parents, although these factors have been shown to influence parental strain in earlier studies (Luescher et al. 1999; Higgins et al. 2005; Meadan et al. 2010). The finding of an association between substance use disorder diagnoses and higher total and externalizing CS strain was based on very few patients and may need replication.

Moreover, the data were collected in 2002. However, there is no indication to assume that the variables associated with SCS would be particularly sensitive to changes that occurred within the time period that has passed since the data collection, as these results are largely in line with earlier studies. Finally, we did not treat repeat users as a separate event, and this could have biased demographic and other characteristics towards the repeating patients. However, repeat users represented only 15.1% of all interviewees, and no association of repeat use and CS was noted.

Conclusions

Even though EDs are clearly not equipped to fully address CS, the knowledge of a particularly high burden in relatives of children with externalizing disorders and in families with low-functioning youth should trigger PPER staff to provide efficient information on appropriate, regular outpatient treatment for the child and on support facilities for the parents.

Clinical Significance

Our findings underscore that in psychiatrically ill youth referred to a PPER and presenting with externalized behaviors and/or those with impaired functioning, well-being should be assessed in both the presenting youth and the caregivers. A comprehensive psychiatric assessment of a child needs to include a detailed evaluation of factors that are not only related to the child, but also to the child's relationship to the primary caregivers as well as to the caregiver's resources and weaknesses (King 1995; Josephson and AACAP Work Group on Quality Issues 2007).

Footnotes

Disclosures

Dr. Correll has been a consultant and/or advisor to or has received honoraria from: AbbVie, Alkermes, Bristol-Myers Squibb (BMS), Eli Lilly, Genentech, Gerson Lehrman Group, IntraCellular Therapies, Janssen/J&J, Lundbeck, MedAvante, Medscape, Otsuka, Pfizer, ProPhase, Reviva, Roche, Sunovion, Supernus, and Takeda. He has received grant support from BMS, Otsuka, and Takeda. Dr. Carbon has the same conflict of interest as Dr. Correll because of a family relationship. The other authors have nothing to disclose.

References

Abbeduto

, Seltzer

, Shattuck

, Krauss

, Orsmond

, Murphy

: Psychological well-being and coping in mothers of youths with autism, Down syndrome, or fragile X syndrome. Am J Ment Retard, 109:237–254, 2004.

Accurso

, Garland

, Haine-Schlagel

, Brookman-Frazee

, Baker-Ericzén

: Factors contributing to reduced caregiver strain in a publicly-funded child mental health system. J Emot Behav Disord, 23:131–143, 2015.

Altamura

, Dell'osso

, Vismara

, Mundo

: May duration of untreated illness influence the long-term course of major depressive disorder?. Eur Psychiatry, 23:92–96, 2008.

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association; 1994.

Anastopoulos

, Guevremont

, Shelton

, DuPaul

: Parenting stress among families of children with attention deficit hyperactivity disorder. J Abnorm Child Psychol, 20:503–520, 1992.

Anderson

: Predictors of parenting stress in a diverse sample of parents of early adolescents in high-risk communities. Nurs Res, 57:340–350, 2008.

Angold

, Messer

, Stangl

, Farmer

, Costello

, Burns

: Perceived parental burden and service use for child and adolescent psychiatric disorders. Am J Public Health, 88:75–80, 1998.

Barkley

, Karlsson

, Pollard

: Effects of age on the mother-child interactions of ADD-H and normal boys. J Abnorm Child Psychol, 13:631–637, 1985.

Barksdale

, Walrath

, Compton

, Goldston

: Caregiver strain and youth suicide attempt: Are they related?. Suicide Life Threat Behav, 39:152–160, 2009.

10.

Blader

: Which family factors predict children's externalizing behaviors following discharge from psychiatric inpatient treatment?. J Child Psychol Psychiatry, 47:1133–1142, 2006.

11.

Brannan

, Athay

, de Andrade

ARV

: Measurement Quality of the Caregiver Strain Questionnaire-Short Form 7 (CGSQ-SF7). Adm Policy Ment Health, 39:51–59, 2012.

12.

Brannan

, Heflinger

, Bickman

: The Caregiver Strain Questionnaire measuring the impact on the family of living with a child with serious emotional disturbance. J Emot Behav Disord, 5:212–222, 1997.

13.

Bussing

, Gary

, Mason

, Leon

, Sinha

, Garvan

: Child temperament, ADHD, and caregiver strain: Exploring relationships in an epidemiological sample. J Am Acad Child Adolesc Psychiatry, 42:184–192, 2003a.

14.

Bussing

, Zima

, Gary

, Mason

, Leon

, Sinha

, Garvan

: Social networks, caregiver strain, and utilization of mental health services among elementary school students at high risk for ADHD. J Am Acad Child Adolesc Psychiatry, 42:842–850, 2003b.

15.

Cadman

, Eklund

, Howley

, Hayward

, Clarke

, Findon

, Xenitidis

, Murphy

, Asherson

, Glaser

: Caregiver burden as people with autism spectrum disorder and attention-deficit/hyperactivity disorder transition into adolescence and adulthood in the United Kingdom. J Am Acad Child Adolesc Psychiatry, 51:879–888, 2012.

16.

Chavira

, Garland

, Yeh

, McCabe

, Hough

: Child anxiety disorders in public systems of care: comorbidity and service utilization. J Behav Health Serv Res, 36:492–504, 2009.

17.

Cloutier

, Kennedy

, Maysenhoelder

, Glennie

, Cappelli

, Gray

: Pediatric mental health concerns in the emergency department: Caregiver and youth perceptions and expectations. Pediatr Emerg Care, 26:99–106, 2010.

18.

Cooper

, Robertson

, Livingston

: Psychological morbidity and caregiver burden in parents of children with Tourette's disorder and psychiatric comorbidity. J Am Acad Child Adolesc Psychiatry, 42:1370–1375, 2003.

19.

Duchovic

, Gerkensmeyer

, Wu

: Factors associated with parental distress. J Child Adolesc Psychiatr Nurs, 22:40–48, 2009.

20.

Dyck

, Short

, Vitaliano

: Predictors of burden and infectious illness in schizophrenia caregivers. Psychosom Med, 61:411–419, 1999.

21.

Estes

, Munson

, Dawson

, Koehler

, Zhou

X-H

, Abbott

: Parenting stress and psychological functioning among mothers of preschool children with autism and developmental delay. Autism, 13:375–387, 2009.

22.

Evans

, Sibley

, Serpell

: Changes in caregiver strain over time in young adolescents with ADHD: The role of oppositional and delinquent behavior. J Atten Disord, 12:516–524, 2009.

23.

Ferro

, Verdeli

, Pierre

, Weissman

: Screening for depression in mothers bringing their offspring for evaluation or treatment of depression. Am J Psychiatry, 157:375–379, 2000.

24.

Fischer

, Kemmler

, Meise

: Burden - distress - quality of life [in German]. Psychiatr Prax, 31:57–59, 2004.

25.

Fraguas

, Del Rey-Mejías

, Moreno

, Castro-Fornieles

, Graell

, Otero

, Gonzalez-Pinto

, Moreno

, Baeza

, Martínez-Cengotitabengoa

, Arango

, Parellada

: Duration of untreated psychosis predicts functional and clinical outcome in children and adolescents with first-episode psychosis: A 2-year longitudinal study. Schizophr Res, 152:130–138, 2014.

26.

Garland

, Lau

, Yeh

, McCabe

, Hough

, Landsverk

: Racial and ethnic differences in utilization of mental health services among high-risk youths. Am J Psychiatry, 162:1336–1343, 2005.

27.

Gerkensmeyer

, Perkins

, Day

, Austin

, Scott

, Wu

: Maternal depressive symptoms when caring for a child with mental health problems. J Child Fam Stud, 20:685–695, 2011.

28.

Hardy

, Hackett

, Murphy

, Cooper

, Ford

, Conroy

: Mental health screening and early intervention: clinical research study for under 5-year-old children in care in an inner London borough. Clin Child Psychol Psychiatry, 20:261–275, 2015.

29.

Harrison

, Sofronoff

: ADHD and parental psychological distress: Role of demographics, child behavioral characteristics, and parental cognitions. J Am Acad Child Adolesc Psychiatry, 41:703–711, 2002.

30.

Heflinger and Taylor Richardson: Caregiver strain in families of children with serious emotional disturbance: does relationship to child make a difference?. J Fam Soc Work, 8:27–45, 2004.

31.

Higgins

, Bailey

, Pearce

: Factors associated with functioning style and coping strategies of families with a child with an autism spectrum disorder. Autism, 9:125–137, 2005.

32.

Horwitz

, Reinhard

: Ethnic differences in caregiving duties and burdens among parents and siblings of persons with severe mental illnesses. J Health Soc Behav, 36:138–150, 1995.

33.

Johnston

, Mash

: Families of children with attention-deficit/hyperactivity disorder: review and recommendations for future research. Clin Child Fam Psychol Rev, 4:183–207, 2001.

34.

Josephson

, AACAP Work Group on Quality Issues: Practice parameter for the assessment of the family. J Am Acad Child Adolesc Psychiatry, 46:922–937, 2007.

35.

Kahn

, Anderson

, Perkoff

: Patient's perceptions and uses of a pediatric emergency room. Soc Sci Med, 7:155–160, 1973.

36.

Kalra

, Kamath

, Trivedi

, Janca

: Caregiver burden in anxiety disorders. Curr Opin Psychiatry, 21:70–73, 2008.

37.

Kazdin

, Whitley

: Treatment of parental stress to enhance therapeutic change among children referred for aggressive and antisocial behavior. J Consult Clin Psychol, 71:504–515, 2003.

38.

King

: Practice parameters for the psychiatric assessment of children and adolescents. J Am Acad Child Adolesc Psychiatry, 34:1386–1402, 1995.

39.

Kjellin

, Ostman

: Relatives of psychiatric inpatients–do physical violence and suicide attempts of patients influence family burden and participation in care?. Nord J Psychiatry, 59:7–11, 2005.

40.

Kouros

, Garber

: Dynamic associations between maternal depressive symptoms and adolescents' depressive and externalizing symptoms. J Abnorm Child Psychol, 38:1069–1081, 2010.

41.

Koutra

, Triliva

, Roumeliotaki

, Lionis

, Vgontzas

: Identifying the socio-demographic and clinical determinants of family functioning in Greek patients with psychosis. Int J Soc Psychiatry, 61:251–264, 2015.

42.

Kuratsubo

, Suzuki

, Shimozawa

, Kondo

: Parents of childhood X-linked adrenoleukodystrophy: High risk for depression and neurosis. Brain Dev, 30:477–482, 2008.

43.

Luescher

, Dede

, Gitten

, Fennell

, Maria

: Parental burden, coping, and family functioning in primary caregivers of children with Joubert syndrome. J Child Neurol, 14:642–648, 1999.

44.

Manskow

, Sigurdardottir

, Røe

, Andelic

, Skandsen

, Damsgård

, Elmståhl

, Anke

: Factors affecting caregiver burden 1 year after severe traumatic brain injury: A prospective nationwide multicenter study. J Head Trauma Rehabil, 30:411–423, 2015.

45.

María Brannan

, Heflinger

: Child behavioral health service use and caregiver strain: comparison of managed care and fee-for-service Medicaid systems. Ment Health Serv Res, 7:197–211, 2005.

46.

Mash

, Johnston

: Parental perceptions of child behavior problems, parenting self-esteem, and mothers' reported stress in younger and older hyperactive and normal children. J Consult Clin Psychol, 51:86–99, 1983.

47.

Meadan

, Halle

, Ebata

: Families with children who have autism spectrum disorders: Stress and support. Except Child, 77:7–36, 2010.

48.

Mugno

, Ruta

, D'Arrigo

, Mazzone

: Impairment of quality of life in parents of children and adolescents with pervasive developmental disorder. Health Qual Life Outcomes, 5:22, 2007.

49.

Oruche

, Draucker

, Al-Khattab

, Cravens

, Lowry

, Lindsey

: The challenges for primary caregivers of adolescents with disruptive behavior disorders. J Fam Nurs, 21:149–167, 2015.

50.

Ostacher

, Nierenberg

, Iosifescu

, Eidelman

, Lund

, Ametrano

, Kaczynski

, Calabrese

, Miklowitz

, Sachs

, Perlick

, STEP-BD Family Experience Collaborative Study Group: Correlates of subjective and objective burden among caregivers of patients with bipolar disorder. Acta Psychiatr Scand, 118:49–56, 2008.

51.

Pelham

, Lang

, Atkeson

, Murphy

, Gnagy

, Greiner

, Vodde-Hamilton

, Greenslade

: Effects of deviant child behavior on parental distress and alcohol consumption in laboratory interactions. J Abnorm Child Psychol, 25:413–424, 1997.

52.

Perlick

, Rosenheck

, Miklowitz

, Chessick

, Wolff

, Kaczynski

, Ostacher

, Patel

, Desai

, STEP-BD Family Experience Collaborative Study Group: Prevalence and correlates of burden among caregivers of patients with bipolar disorder enrolled in the Systematic Treatment Enhancement Program for Bipolar Disorder. Bipolar Disord, 9:262–273, 2007.

53.

Perlick

, Rosenheck

, Clarkin

, Raue

, Sirey

: Impact of family burden and patient symptom status on clinical outcome in bipolar affective disorder. J Nerv Ment Dis, 189:31–37, 2001.

54.

Pierce

, Ewing

, Campbell

: Diagnostic status and symptomatic behavior of hard-to-manage preschool children in middle childhood and early adolescence. J Clin Child Psychol, 28:44–57, 1999.

55.

Rockhill

, Violette

, Vander Stoep

, Grover

, Myers

: Caregivers' distress: youth with attention-deficit/hyperactivity disorder and comorbid disorders assessed via telemental health. J Child Adolesc Psychopharmacol, 23:379–385, 2013.

56.

Rodrigo

, Fernando

, Rajapakse

, De Silva

, Hanwella

: Caregiver strain and symptoms of depression among principal caregivers of patients with schizophrenia and bipolar affective disorder in Sri Lanka. Int J Ment Health Syst, 7:2, 2013.

57.

Sayal

: The role of parental burden in child mental health service use: Longitudinal study. J Am Acad Child Adolesc Psychiatry, 43:1328–1333, 2004.

58.

Soto

, Frederickson

, Trivedi

, Le

, Eugene

, Shekher

, Weiskopf

, Allen-Dicker

, Dicker

, Fornari

, Correll

: Frequency and correlates of inappropriate pediatric psychiatric emergency room visits. J Clin Psychiatry, 70:1164–1177, 2009.

59.

Stancin

, Wade

, Walz

, Yeates

, Taylor

: Traumatic brain injuries in early childhood: initial impact on the family. J Dev Behav Pediatr, 29:253–261, 2008.

60.

Weissman

, Pilowsky

, Wickramaratne

, Talati

, Wisniewski

, Fava

, Hughes

, Garber

, Malloy

, King

, Cerda

, Sood

, Alpert

, Trivedi

, Rush

, STAR*D-Child Team: Remissions in maternal depression and child psychopathology: A STAR*D-child report. JAMA, 295:1389–1398, 2006.

61.

Winefield

, Harvey

: Determinants of psychological distress in relatives of people with chronic schizophrenia. Schizophr Bull, 19:619–625, 1993.

62.

Woods

: The economic burden of bipolar disease. J Clin Psychiatry 61 Supp, 13:38–41, 2000.

63.

, Hoven

, Cohen

, Liu

, Moore

, Tiet

, Okezie

, Wicks

, Bird

: Factors associated with use of mental health services for depression by children and adolescents. Psychiatr Serv, 52:189–195, 2001.