Research report
Utility of the Child Behavior Checklist in screening depressive disorders within clinical samples

https://doi.org/10.1016/j.jad.2010.08.011Get rights and content

Abstract

Background:

Previous studies analyzed the clinical validity of the Child Behavior Checklist (CBCL) for screening primary depressive disorders (major depression and dysthymia). Other psychiatric disorders with depressive symptomatology (e.g. adjustment disorders with depressive reaction) have not been a research focus to date. This study aims to examine the utility of the CBCL scales Anxious/Depressed and Affective Problems in screening both primary depressive disorders and other disorders with depressive symptomatology.

Methods:

The two samples consisted of 1445 outpatients and 698 inpatients aged 11–18 who were referred for child and adolescent psychiatric services. The predictive power of the CBCL scales was examined using ROC-analysis.

Results:

Both CBCL scales showed small to medium predictive power when screening for primary and other depressive disorders in clinically referred outpatients (AUCs = .694  .780). Within the inpatient sample, only primary depressive disorders were detected with small predictive power different from chance level (AUCs = .625  .673). Both CBCL scales did not differ significantly with regards to predicting depressive disorders.

Limitations:

A limitation of this study is the lack of reliability checks for consensus classification process of the ICD-10 diagnoses. However, comparable prior work using structured interviews resulted in similar AUC values, supporting our findings.

Conclusions:

In outpatient samples, the CBCL is suitable for screening other depressive disorders in addition to primary depressive disorders. This is not the case for inpatients due to a misclassification number of 40%; thus using CBCL scales screen for depressive disorders in such populations warrants caution.

Introduction

Depressive disorders are very common in youth psychopathology: about 10% of adolescents experience a depressive episode by the age of sixteen (Costello et al., 2003). Depressive disorders are associated with substantial psychosocial impairment, e.g. in school or social relationships (Essau et al., 2000, Lewinsohn et al., 2003). They commonly occur in conjunction with other mental disorders, such as anxiety, oppositional defiant disorder, substance use disorders, attention-deficit/hyperactivity disorders, and eating disorders (Costello et al., 2003). Furthermore, depressive symptoms are a relevant component of numerous psychiatric disorders such as adjustment disorders with depressive reaction, adjustment disorder with mixed anxiety and depressive reaction, mixed anxiety and depressive disorder, schizoaffective disorder—depressive type, bipolar affective disorders—current episode depression, or depressive conduct disorder (according to ICD-10; World Health Organization [WHO], 2008).

Early detection of depressive disorders facilitates treatment in time. Therefore, screening for depressive disorders is of vital importance. The Child Behavior Checklist (CBCL) is a cost- and time-efficient screening instrument developed for a dimensional assessment of competences as well as behavioral and emotional problems from the parents or caregiver's perspective. It is part of the Achenbach System of Empirically Based Assessment (ASEBA). Achenbach (1991) was able to show that the CBCL is effective in distinguishing between patients referred for mental health services and non-clinical populations and therefore has proven to be an effective screening instrument in clinical settings as well as in research. Its psychometric scales consist of a Total Problems Score, two broadband factors (Internalizing and Externalizing Problems) and eight different, statistically derived syndrome scales (e.g. Anxious/Depressed). The CBCL is easy to administer, has sound psychometric properties and is internationally validated (Achenbach, 1991, Doepfner et al., 1994, Rescorla et al., 2007).

Several studies have examined the concurrence between the CBCL and psychiatric diagnostic systems such as the Diagnostic and Statistical Manual of Mental Disorders (DSM-III-R/DSM-IV; American Psychiatric Association, 1987, American Psychiatric Association, 1994) and the International Classification of Diseases (ICD-10; WHO, 2008). A strong concurrence has been documented between the CBCL and externalizing disorders, such as conduct disorder as well as attention-deficit–hyperactivity disorders. In contrast, the concurrence rates for internalizing disorders, such as affective and anxiety disorders were reported to be substantially lower (Connor-Smith and Compas, 2003, Lengua et al., 2001, McGuire et al., 2000). The prediction of affective disorders based on one specific CBCL syndrome scale has not proved to be possible to a sufficient extent (Kasius et al., 1997).

In order to increase the congruence between the CBCL and the DSM, Achenbach and Rescorla (2001) developed new DSM-oriented scales based on expert ratings. The scale Affective Problems was constructed to detect the presence of major depressive disorders and dysthymia (Achenbach et al., 2003).

Besides the studies by Achenbach, as part of which the DSM-oriented scales were developed and standardized (Achenbach et al., 2003), there is only one study examining the validity of the Affective Problems scale for predicting affective disorders: Ferdinand (2008) analyzed the concurrence of the new Affective Problems scale with dysthymia and major depressive disorders. In order to validate the diagnosis, the results of the Anxiety Disorders Interview Schedule for Children (ADIS-C/P) were reviewed. Using Receiver Operating Characteristic analyses, the scale's predictive power proved to be moderate (AUC = .77). However, only 9 out of 255 patients suffered from dysthymia or major depressive disorders and 156 of the included patients showed no diagnosis at all. Accordingly, replications of these findings are necessary.

While the use of the CBCL as a screening tool for primary depressive disorders (major depressive disorders and dysthymia) has been investigated, its ability to screen for other ICD-10 psychiatric disorders with depressive symptomatology as mentioned above (e.g. adjustment disorders with depressive reaction) has not yet been examined. However, the prevalence of these disorders is not negligible (e.g. Baumeister, 2008, Das-Munshi et al., 2008, Schmidt et al., 2007). Therefore, it is important to study the CBCL's screening performance for this group of depressive disorders.

On the whole, the CBCL has been analyzed as an indicator of major depressive disorders and dysthymia (e.g. Achenbach et al., 2003, Ferdinand, 2008, Kasius et al., 1997), but it has not been studied with regard to screening other psychiatric disorders with depressive symptomatology. To date the extent to which these other depressive disorders can be predicted by using CBCL scales and which scale fits best is unknown. Furthermore, studies on the screening performance of the CBCL scale Affective Problems are scarce and need replication.

The aim of the present study is to examine the validity and utility of the CBCL scales Anxious/Depressed and Affective Problems in diagnosing primary depressive disorders as well as other psychiatric disorders with depressive symptomatology within both a clinically referred inpatient sample and an outpatient sample. The nature of the study is explorative.

Section snippets

Sample

Two samples were investigated, both consisting of adolescents (aged 11–18) referred for psychiatric services to the Child and Adolescent Psychiatry Unit of the Philipps-University Marburg, Germany. The only inclusion criterions were age and German-speaking parents, whilst the only exclusion criterion was referral for crisis intervention, consequently the samples are representative of an unselected clinically referred adolescent population. Patients, who were treated several times were only

CBCL

The German version of the CBCL (Working Group German Child Behavior Checklist, 1993) queries parents' assessments of child competencies as well as behavioral and emotional problems in children and adolescents aged 4 to 18 years. Parents rate their children on 118 specific problem items using a 3-point scale (0 = “not true”, 1 = “somewhat/sometimes true” and 2 = “very/often true”). As in the US version of the CBCL, a Total Problem score, two broadband scores (Internalizing and Externalizing Problems)

General symptom severity

The mean CBCL Total Problem Score in both samples was above the threshold of 63 T-points (inpatient: MT = 69.9, SD = 8.7; outpatient: MT = 66.7, SD = 9.6), and therefore in the clinical range. Symptom severity was significantly higher in inpatients than in outpatients (CBCL Total Problem Score, Mann–Whitney U test; Z = −7.359, p < .001). Looking at depressive patients, there was no significant difference between inpatients and outpatients (D-all; inpatient: MT = 70.4, SD = 8.2; outpatient: MT = 70.9, SD = 7.7; Z = 

Discussion

With regard to screening for primary depressive disorders (major depression and dysthymia), the CBCL scales Anxious/Depressed and Affective Problems has been recommended by several authors (Achenbach, 1991, Achenbach et al., 2003, Ferdinand, 2008). We examined the clinical utility of these scales applied routinely in outpatient and inpatient psychiatric settings for predicting primary depressive disorders. In addition, we studied the CBCL's predictive power with regard to other psychiatric

Role of funding source

No third-party funding has been provided for creating the submitted work.

Conflict of interest

The authors declare that they have no competing interests.

Acknowledgements

We thank U. König, who was involved in the data management, M. Heinzel-Gutenbrunner, who supervised the statistical analyses, and J. Pauschardt, who contributed to the revision of the manuscript. Furthermore, we thank all clinical psychologists and psychiatrists of the Department of Child and Adolescent Psychiatry, Philipps-University Marburg, who contributed to the data collection.

References (29)

  • J.K. Connor-Smith et al.

    Analogue measures of DSM-IV mood and anxiety disorders based on behavior checklists

    J Psychopathol Behav Assess

    (2003)
  • E.J. Costello et al.

    Prevalence and development of psychiatric disorders in childhood and adolescence

    Arch. Gen. Psychiatry

    (2003)
  • J. Das-Munshi et al.

    Public health significance of mixed anxiety and depression: beyond current classification

    Br. J. Psychiatry

    (2008)
  • E.R. DeLong et al.

    Comparing the areas under two or more correlated Receiver Operating Characteristic curves: A nonparametric approach

    Biometrics

    (1988)
  • Cited by (11)

    • Characteristics of the child behavior checklist in adolescents with depression associated with bipolar disorder

      2013, Journal of Affective Disorders
      Citation Excerpt :

      For example, Biederman et al. (1996a, 2005a, b), in two separate studies, found the Anxious/Depressed subscale to correlate well with unipolar major depression. Similarly, Eimecke and colleagues (2011) demonstrated low to medium predictive power of this subscale in both unipolar depressive disorder and other disorders with depressive symptomatology, including bipolar disorder. Although the CBCL has been extensively studied in bipolar youth during manic or mixed episode and in youth with unipolar depression, there is a paucity of data on the CBCL profile of youth with depression associated with bipolar disorder.

    • Psychometric properties of the Treatment-Emergent Activation and Suicidality Assessment Profile (TEASAP) in youth with OCD

      2013, Psychiatry Research
      Citation Excerpt :

      The CBCL yields a total problem score, two broadband scores (internalizing and externalizing) and eight syndrome scales (Achenbach, 2001). The CBCL has good psychometric properties (Aschenbrand et al., 2005; Dedrick et al., 2008) and clinical utility (Eimecke et al., 2011). The CBCL Externalizing Subscale was used to test the convergent validity of the total TEASAP score.

    • Can defense styles predict psychopathology in adolescents?

      2011, Procedia - Social and Behavioral Sciences
    View all citing articles on Scopus
    View full text