RESUMO
While there is high prevalence rate of anxiety and depressive illness in community their recognition rate in primary care varies from 17-30 percent. There is, therefore, need to detect these disorders in primary care settings. The study aims at comparing the reliability and validity of two frequently used instruments for their detection namely self-reporting questionnaire [SRQ] and Bradford SOMATIC inventory [BSI]. Two-stage study. Study was carried out in a rural community [Susral], about 75 km south-west of Rawalpindi, and its duration was 4 months. First stage involved administration of SRQ and BSI by trained raters to 664 adults. In the Second stage Psychiatrist blind to the first stage score carried out structured interview on stratified random sample population using Psychiatric Assessment Schedule [PAS] based on lCD-10 [Research diagnostic criteria] to 191 adults. The data was analyzed using SPSS.PV+V, 3-0 and ROCFIT Computer programme for receiver operative characteristics analysis. The validity co-efficients of SRQ and BSI for males and females were calculated. The best threshold for SRQ was 7/8 for women having a sensitivity of 78 percent and specificity of 81 percent. For men the best SRQ threshold was 3/4 having a sensitivity of 78 percent and specificity of 70 percent. The positive predictive values [PPV] were 95 percent and 62 percent respectively. In case of BSI the best threshold for women was 20/22 with sensitivity of 82 percent and specificity of 71 percent. For men the best threshold was 11/14 having a sensitivity of 59 percent and specificity of 62 percent. The PPVs were 95 percent and 44 percent respectively. It is safe to conclude that both SRQ-20 and BSI-44 are useful as screening instruments for probable psychiatric morbidity with SRQ - 20 emerging as a more versatile screening instrument