Diagnostic accuracy at the first episode of psychosis in Uganda

ABSTRACT Background Correct clinical diagnosis at the first episode of psychosis may be difficult due to many non-specific symptoms. We aimed to determine the factors associated with a correct diagnosis among patients with a first episode of psychosis in Uganda. Methods A cross sectional study design was performed at the Butabika National Referral Mental Hospital in Uganda. We included treatment naïve participants aged 18 to 60 years with a diagnosis of a psychotic disorder. Patients with organic disorders like HIV/AIDS, syphilis and substance use disorders were excluded. The MINI international neuropsychiatric inventory was administered to confirm the clinical diagnosis. Concordance was based on the percentage agreement and kappa statistic between the admission chart diagnosis and the MINI diagnosis. Results 178 participants with a first episode of psychosis were included into the final analysis. The agreement between the MINI diagnosis and clinician diagnosis was 0.385, (P < 0.001) with a concordance of 49.5%. After controlling for nationality and the household’s source of income, duration of untreated psychosis, [p-value 0.028(95%CI: 0.07-0.89)], living with a primary family member, [p-value 0.038(95%CI:0.95-2.86)] and cadre of the clinician who made the initial diagnosis[Medical officer, [p-value 0.011(95%CI: 0.18-0.80)] were associated with a correct diagnosis. Conclusion We found low agreement between clinician diagnoses and MINI diagnoses at the first episode of psychosis. Improved training of staff while considering the duration of untreated psychosis and the living arrangements of the patient are required to improve diagnostic accuracy in this population.