Performance of primary health care workers in detection of mental disorders comorbid with epilepsy in rural Ethiopia

Background Timely detection and management of comorbid mental disorders in people with epilepsy is essential to improve outcomes. The objective of this study was to measure the performance of primary health care (PHC) workers in identifying comorbid mental disorders in people with epilepsy against a standardised reference diagnosis and a screening instrument in rural Ethiopia. Methods People with active convulsive epilepsy were identified from the community, with confirmatory diagnosis by trained PHC workers. Documented diagnosis of comorbid mental disorders by PHC workers was extracted from clinical records. The standardized reference measure for diagnosing mental disorders was the Operational Criteria for Research (OPCRIT plus) administered by psychiatric nurses. The mental disorder screening scale (Self-Reporting Questionnaire; SRQ-20), was administered by lay data collectors. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of PHC worker diagnosis against the reference standard diagnosis was calculated. Logistic regression was used to examine the factors associated with misdiagnosis of comorbid mental disorder by PHC workers. Results A total of 237 people with epilepsy were evaluated. The prevalence of mental disorders with standardised reference diagnosis was 13.9% (95% confidence interval (CI) 9.6, 18.2%) and by PHC workers was 6.3% (95%CI 3.2, 9.4%). The prevalence of common mental disorder using SRQ-20 at optimum cut-off point (9 or above) was 41.5% (95% CI 35.2, 47.8%). The sensitivity and specificity of PHC workers diagnosis was 21.1 and 96.1%, respectively, compared to the standardised reference diagnosis. In those diagnosed with comorbid mental disorders by PHC workers, only 6 (40%) had SRQ-20 score of 9 or above. When a combination of both diagnostic methods (SRQ-20 score?=?9 and PHC diagnosis of depression) was compared with the standardised reference diagnosis of depression, sensitivity increased to 78.9% (95% (CI) 73.4, 84.4%) with specificity of 59.7% (95% CI 53.2, 66.2%). Only older age was significantly associated with misdiagnosis of comorbid mental disorders by PHC (adjusted odds ratio, 95% CI?=?1.06, 1.02 to 1.11). Conclusion Routine detection of co-morbid mental disorder in people with epilepsy was very low. Combining clinical judgement with use of a screening scale holds promise but needs further evaluation.