Richens et al question whether the development of guidelines for the provision of post-exposure prophylaxis following sexual exposure (PEPSE) will cause harm, based on three areas of concern: doubts over efficacy, possible effects on sexual behaviour, and cost.

The efficacy of PEPSE has yet to be determined in a robust fashion. As Richens et al observe, data to support the UK¹ (and other) guidelines have been drawn from animal models, vertical transmission studies, retrospective data from healthcare workers exposed to HIV, and prospective (unrandomised) studies in men who have sex with men (MSM),² and individuals following sexual assault.³ In the latter two case-controlled studies HIV seroconversion in PEPSE recipients was 0.6% and 0% compared to 4.2% and 2.7% in controls, supporting its putative efficacy. It is highly unlikely that a prospective randomised study to address the fundamental question of efficacy would be possible (given the numbers involved to achieve sufficient power) or acceptable from an individual or ethical perspective. Whether the absence of such data should result in a failure to offer a potentially beneficial intervention is questionable. Although we live in an era where evidence based medicine is the holy grail, the majority of medical interventions are not supported by high levels of robust evidence from randomised studies. The provision of PEP to healthcare workers following actual or potential exposure to HIV is open to similar criticisms, but the routine adoption of PEP in this setting is not questioned by Richens et al, nor indeed by UK or other national policies. This is despite a marked difference in risk of transmission involved in comparing these scenarios. For example, the risk to a healthcare worker following a needlestick injury …