Bacteriological skin and subcutaneous infections in injecting heroin users-relevance for custody

Forensic medical examiners, custody nurses and police not infrequently come into physical contact with injecting heroin users whose injection sites maybe infected. Karch evaluated published reports on bacteriological findings at infected injection sites in drug users in 1996. Studies were then 10 years out of date. Recent concern has been about unexplained deaths in injecting heroin users, partly attributed to toxins produced by Clostridium novyii, Clostridium perfringens (welchii) and Clostridium botulinum. In response to this epidemic Lambeth, Southwark and Lewisham Health Authority (London, UK) urged investigation and treatment of injection site infection in heroin users within their area; circularising police stations and general practices with an official notice. In response, I followed up and treated in general practice 35 injecting heroin users with injection site infection presenting as detainees at two South London police stations. All swabs taken had a 48 hour incubation for anaerobic bacteria. At audit, no spore forming anaerobes were found. Just under 25% had non spore forming anaerobes such as bacteroides. Just over 50% had staphylococci in the injection site infection. Just under 20% had streptococci, half of which were beta haemolytic, a possible contact risk of infection being erysipelas, cellulitis, impetigo or conjunctivitis. No comparable studies were found but a study in New York in 1984 tabulated bacteriological findings of soft tissue infections in 29 hospitalised parenteral drug users. They found similar proportions of staphylococci, far more with streptococci, hardly any with non spore forming anaerobes and one case of spore forming anaerobic clostridium perfringens.