Information Technology Aided Relapse Prevention Program in Schizophrenia (ITAREPS): Reliability and validity of the early warning signs questionnaire
Abstract
Background: The aim of the study is to determine psychometric properties of the Early Warning Signs Questionnaire, both Patient (EWSQ-10P) and Family Member (EWSQ-10FM) versions used within the Information Technology Aided Relapse Prevention Program in Schizophrenia (ITAREPS). The ITAREPS program employs mobile phone communication between the psychiatrist and the patient.
Rapid and targeted recognition of the most common warning signs of a relapse makes timely pharmacological intervention possible at the initial stages of relapse in accordance with the suggested Early Intervention Algorithm (EIA). Methods: Data analyses used Cronbach's Alpha to determine internal consistency reliability and Spearman's correlation to assess convergent validity of EWSQ-10 compared with Clinical Global Impression, Global Improvement subscale (CGI-I).
Results: During 5-months follow-up period, 22 patients reported 78 EWSQ-10P and 14 family members delivered 43 EWSQ 10FM reflecting occurrence of early warning sings, out of the total 66 patients and 48 family members previously enrolled in the ITAREPS program. The EWSQ-10 was internally consistent, with Cronbach's Alpha 0,69 in EWSQ-10P (Patient Version) and 0,75 in EWSQ-10FM (Family Member Version).
By means of CGI-Global Improvement subscale, outpatient psychiatrists assessed clinical deterioration in 66% EWSQ 10P and 33% EWSQ 10FM out of those questionnaires, using a web-based data capture system. These data did not find statistically significant convergent validity of both EWSQ Patient and the EWSQ Family Member Version with CGI Global Improvement subscale.
We explain this negative finding by superior sensitivity of the EWSQ over CGI-I in the detection of prodromal symptoms.