Validation of the Behavioral Health Checklist in Diverse Pediatric Primary Care Settings.

OBJECTIVE

Examine the validity and clinical utility of the Behavioral Health Checklist (BHCL), a screening tool with 2 forms (4-7, 8-12 years) developed for use with children of diverse backgrounds.

METHOD

At pediatric primary care appointments, the parents of 1274 children completed a demographic form, the BHCL, and the Child Behavior Checklist (CBCL). Concurrent validity was examined by conducting correlations between the BHCL and the diagnostic scales of the CBCL. Diagnostic prediction was examined by conducting logistic regression analyses and plotting receiver operating characteristics (ROC) curves. Clinical utility was investigated by examining sensitivity, specificity, and kappa corrections for total predictive power.

RESULTS

The pattern of correlations with the CBCL provided evidence of convergent and discriminant validity for both versions of the BHCL. ROC curve plots provided clear evidence of predictive validity (area under curve values ranged from .84 to .96 across factors and both age-determined versions). Cut-points achieving sensitivity and specificity values of at least .70 were identified for each BHCL factor for each version.

CONCLUSION

The BHCL was demonstrated to have strong construct and predictive validity. The predictive validity of each version was demonstrated across genders, socioeconomic status, and racial groups (black or African American and white). The BHCL has promise as a developmentally and culturally effective behavioral health screener for use in pediatric primary care practices.