To what extent do medication experts agree on diagnoses derived from psychotropic medication prescribed for children and adolescents in a state Medicare database?. (June 2021)

• Record Type: Journal Article
• Authors: O'Brien, Conor; Rapp, John T.
• Is Part Of: Developmental child welfare. Volume 3:Number 2(2021)
• Page Start: 150
• Page End: 165
• Publication Date: 2021-06
• Subjects: Diagnosis agreement -- foster children -- Medicaid database -- polypharmacy -- psychotropic medication; Child welfare -- Periodicals; Poor children -- Social conditions -- Periodicals; Poor youth -- Social conditions -- Periodicals; Poor children -- Health and hygiene -- Periodicals; Poor youth -- Health and hygiene -- Periodicals; Poor youth -- Social conditions; Poor children -- Social conditions; Poor children -- Health and hygiene; Child welfare; Electronic journals; Periodicals; 362.705
• DOI: 10.1177/25161032211019045
• Languages: English
• ISSNs: 2516-1032

Abstract

This study evaluated the extent to which psychotropic medication experts agreed on psychiatric/behavior diagnoses derived from 30 individuals' psychotropic medication regimens. Three medication experts reviewed the medication regimens and inferred one or more diagnoses based on the medication listed. Thereafter, we used kappa statistical analyses and category-by-category analyses to evaluate agreement of diagnoses (a) across all three reviewers for two time points (separated by 8 to 14 months) covered by each data set, (b) across each pairing of reviewers at the two time points, and (c) within each reviewer across both time points. Between-reviewer kappa statistical analyses of first- and last-regimen data yielded only five instances with excellent agreement and 13 instances of poor agreement. All remaining instances indicated various levels of disagreement. Similarly, within-expert kappa statistical analyses showed two instances with excellent agreement, four instances with poor agreement, and the remaining instances with various levels of disagreement. Overall, the highest kappa values were attached to low-count regimens, while most scores, regardless of medication count, were low and indicated potential disagreement. The category-by-category analyses yielded similar results. A secondary, conditional analysis revealed higher agreements between and within reviewers when medication regimens contained psychotropic medications typically prescribed to individuals diagnosed with Attention Deficit Hyperactivity Disorder.