Background People with severe mental illness (SMI) are at risk of contracting blood-borne viruses (BBVs) and other sexually transmitted infections (STI’s), however, this area of health has been neglected by policy and practice. A recent meta-analysis found the pooled prevalence of HIV in this population to be 6%. One explanation that has been suggested is engagement in high risk sexual behaviours such as sex work, sex trading and multiple partners. To investigate this under researched area we conducted a systematic review and meta-analysis to explore engagement in sexual risk behaviours in adults with SMI.

Methods We searched EMBASE, MEDLINE, CINAHL, AIDSLINE, PsychINFO and Web of Science using, ‘severe mental illness’, ‘blood-borne virus’ and ‘sexually transmitted infection’ terms without date or language restrictions. Grey literature resources were searched and reverse-citation and reference list searches undertaken. Studies were eligible if; the population were aged 18 or over, the exposure of interest was a SMI, comparator groups had no history of SMI and the outcome measured was engagement in sexual risk behaviours. Methodological quality of studies was assessed using the Newcastle Ottawa Scale. The risk behaviours identified were grouped together and meta-analyses conducted in STATA 13.

Results The search identified 10424 potentially relevant records. After duplicates were removed (2472) 7952 records remained. Overall 7824 studies were excluded leaving 128 full articles to be assessed. Six case-control studies met the eligibility criteria and were included in the review. A total of ten sexual risk behaviours were consistently reported. The sexual risk behaviour most significantly associated with SMI was ‘paid sex work’ with a pooled odds ratio of 5.36 (95% CI 2.15–13.38).

Conclusion Adults with SMI are more likely to engage in sexually risky behaviours compared to healthy controls and therefore, are at risk of contracting BBV’s or other STI’s. Several methodological limitations of the included studies were identified meaning the results should be interpreted with caution. SMI is unlikely to be a sole risk factor, substance misuse, gender, ethnicity and socioeconomic status should also be considered. Further robust epidemiological studies are needed to explore this relationship to help develop preventative strategies. Failure to address this could result in increased morbidity and mortality as a result of undetected and untreated infection. Therefore, mental health care providers should address sexual health and potential risk behaviours within routine care.