Validation of a Computerized Parent Questionnaire for Identifying Child Mental Health Disorders and Implementing DSM-PC
BACKGROUND
Guidelines for child health supervision have called for early identification of childhood psychosocial problems and a new classification scheme (DSM-PC) has been developed and to help pediatricians in the identification process. Obstacles to implementation include the fact that the new system more than triples the number of potential diagnoses and pediatricians may not yet be familiar with criterion for even the most severe categories of psychiatric disorder. We have developed the Child Health and Development Interactive System (CHADIS), a computerized aid to identification and on-the-spot education for the pediatrician via an "intelligent electronic record."
OBJECTIVE
To assess the preliminary concurrent validity of the DSM-PC parent inventory component of CHADIS for identification of child psychiatric disorder in school aged children.
DESIGN / METHODS
The discriminant efficiency of decisions based on DSM-PC/C classifications was compared to use of three "gold standard" criterion: (1) the computerized, self-administered, parent version of the Diagnostic Interview for Children and Adolescents (DICA; Reich, 2000), (2) the paper and pencil version of the Child behavioral Checklist (CBCL; Achenbach & Rescorla, 2001), and (3) diagnoses by the behavioral-developmental pediatrician treating the patient who was blind to the CHADIS, CBCL and DICA results. The sample consisted of seventy-one caregivers of children being seen for a routine well child check-up (42 %) or a behavioral/developmental consult (33%) or who were regular patients of the study pediatricians and agreed to participate in the study but were not scheduled for an appointment (25%) at a pediatric clinic serving a 60% Medicaid population with 77% Caucasian, 20% African-American and 3% other. Children were between 6 to 12 years old (average = 10) with more boys (39) than girls (36). Administration of DICA and CHADIS were counterbalanced the order of administration. CHADIS asks parents to identify concerns from a list and In order to simulate the typical use of CHADIS during a pediatric visit, caregivers were only asked to rank their top 15 concerns (if they had that many) and to complete no more than five of the DSM-PC questionnaires representing the top five concern areas they had prioritized. Caregivers completed the CHADIS questionnaire associated with their top ranked concern in 12 minutes on average (It is estimated that many primary care applications will limit parents to only their top concern, and many non-referred families have one concern or fewer).
RESULTS
CHADIS appears to be sensitive, in terms of detecting a disorder when it is present according to another measure, specific, in terms of not diagnosing disorders that are not present according to other measures, and has good percent agreement when comparing it to the DICA (sens=.86, spec=.96, % agree=.91), CBCL (sens=.81, spec=.88, % agree=.85) and physician’s diagnosis (sens=.81, spec=.85, % agree=.83).
Sturner, R., Howard, B. J., Morrel, T., & Rogers-Senuta, K. (2003). Validation of a Computerized Parent Questionnaire for Identifying Child Mental Health Disorders and Implementing DSM-PC. Pediatric Academic Societies Meeting; abstract in Pediatric Research, 2003.
