Development and psychometric evaluation of the Decision Tool Anxiety Disorders, OCD and PTSD (DTAOP): Facilitating the early detection of patients in need of highly specialized care

Frédérique C.W. van Krugten, Meriam Kaddouri, Maartje Goorden, Anton J.L.M. van Balkom, Ed W. Berretty, Daniëlle C. Cath, Gert Jan Hendriks, Suzy J.M.A. Matthijssen, Henny A.D. Visser, Irene M. van Vliet, Werner B.F. Brouwer, Leona Hakkaart-Van Roijen

Research output: Contribution to journalArticle*Academicpeer-review

Abstract

Background Early identification of patients with an anxiety disorder, obsessive-compulsive disorder (OCD), or post-traumatic stress disorder (PTSD) in need of highly specialized care could facilitate the selection of the optimal initial treatment in these patients. This paper describes the development and psychometric evaluation of the Decision Tool Anxiety Disorders, OCD and PTSD (DTAOP), which aims to aid clinicians in the early identification of patients with an anxiety disorder, OCD, or PTSD in need of highly specialized mental healthcare. Methods A systematic literature review and a concept mapping procedure were carried out to inform the development of the DTAOP. To evaluate the psychometric properties of the DTAOP, a cross-sectional study in 454 patients with a DSM-IV-TR anxiety disorder was carried out. Feasibility was evaluated by the completion time and the content clarity of the DTAOP. Inter-rater reliability was assessed in a subsample of 87 patients. Spearman's rank correlation coefficients between the DTAOP and EuroQol five-dimensional questionnaire (EQ-5D5L) scores were computed to examine the convergent validity. Criterion validity was assessed against independent clinical judgments made by clinicians. Results The average time required to complete the eight-item DTAOP was 4.6 min and the total DTAOP was evaluated as clear in the majority (93%) of the evaluations. Krippendorff's alpha estimates ranged from 0.427 to 0.839. Based on the qualitative feedback, item wording and instructions were improved. As hypothesized, the DTAOP correlated negatively with EQ-5D-5L scores. The area under the curve was 0.826 and the cut-off score of ≥4 optimized sensitivity (70%) and specificity (71%). Conclusions The DTAOP demonstrated excellent feasibility and good validity, but weak inter-rater reliability. Based on the qualitative feedback and reliability estimates, revisions and refinements of the wording and instructions were made, resulting in the final version of the DTAOP.

Original languageEnglish
Article numbere0256384
JournalPLoS ONE
Volume16
Issue number8 August
DOIs
Publication statusPublished - Aug 2021

Cite this