Objective: Research on HIV in South Africa has not reflected the impact of the disease on psychiatric patients. The aims of the study were: to compare the HIV prevalence among patient groups in Weskoppies Hospital; to compare psychiatric diagnoses of infected and non-infected patients; to assess intravenous drug use and high-risk sexual behaviour; to establish HIV-syphilis association; and to investigate the rapid test performance for screening, compared to the confirmatory ELISA test. Method: Onehundred-and-ninety-five patients were grouped into four categories according to their duration of admission and gender. HIV rapid testing, HIV ELISA, syphilis-RPR and TPHA testing were performed. Results: The HIV prevalence of 11% in the sample was significantly associated with 'gender-and-duration-of-admission' categories (p=0.003). No significant association between HIV infection and psychiatric diagnoses or intravenous drug use was found, but a significant association existed between HIV infection and high-risk sexual behaviour (p=0.002), and between HIV and syphilis (p=0.012). The HIV rapid screening test had a sensitivity of 91.7% and a specificity of 98.2%. Conclusion: The overall HIV prevalence at Weskoppies Hospital remains lower than the national average, but has increased since a previous local study. The rapid test for HIV had a lower sensitivity than was expected, and it is recommended that HIV ELISA testing be performed as a first line test in the setting of hospitalised patients. Due to the high prevalence of HIV and syphilis in the psychiatric population it is recommended that all patients be tested for both of these diseases.

<p><strong>Background.</strong> Syphilis research has neglected the prevalence of the disease among psychiatric patients, and traditional syphilis screening has been reported as inadequate.</p><p><strong>Objectives.</strong> (i) To assess the syphilis prevalence among psychiatric patients; (ii) to compare psychiatric diagnoses of syphilis-infected and -uninfected patients; (iii) to assess self-reported high-risk sexual behaviour; (iv) to establish syphilis/HIV co-morbidity; and (v) to investigate the performance of the rapid plasma reagin (RPR) test in syphilis screening, compared with the Treponema pallidum haemagglutination (TPHA) test.</p><p><strong>Methods.</strong> Psychiatric inpatients at Weskoppies Hospital, Pretoria, who consented to participate in the study (N=195) were categorised according to gender and length of admission (long-term or recent). Non-treponemal RPR, confirmatory TPHA, HIV-rapid and HIV enzyme-linked immunosorbent assay (ELISA) tests were performed. A reactive TPHA test was used to diagnose syphilis.</p><p><strong> Results.</strong> The estimated prevalence of syphilis was 11.7%. There was no significant association between TPHA sero-positivity and primary psychiatric diagnosis or self-reported high-risk sexual behaviour. Significant co-morbidity existed between syphilis and HIV (p=0.012). Compared with the TPHA test, the RPR test performed poorly, identifying only 2/23 patients who had a sero-positive TPHA test (8.7% sensitivity and 100% specificity).</p><p><strong>Conclusions.</strong> The prevalence of syphilis was higher than anticipated, supporting the need for routine testing. The significant co-morbidity and alarming prevalence of HIV and syphilis warrant testing for both conditions in all psychiatric admissions. Current syphilis screening with a single RPR test is inadequate; both RPR and TPHA tests should be performed.</p>