The investigation of salmonella outbreaks in hospitals

Abstract

Person-to-person spread of salmonellas in hospitals has been well documented (Datta and Pridie, 1960) but doctors and nurses are often unaware of the possibility and, when an outbreak does occur, they may be reluctant to accept that poor ward hygiene may be the cause. Although serially propagated salmonella outbreaks grumbling on for days or weeks are not as dramatic as explosive food-borne outbreaks they can seriously disrupt hospital services. For example, the 1980 outbreak of SalmonelZa lnuenchen at The Hospital for Sick Children, London, in which 28 children were infected lasted for 4 weeks and was only controlled after all wards were closed to cold admissions for 10 days and to emergency admissions for 5 days (Kumarasinghe et al., 1982). The relative importance of person-to-person and food-borne spread of salmonellas in hospitals was indicated by the Public Health Laboratory Service Salmonella Subcommittee’s review of the 552 hospital salmonella outbreaks in England and Wales from 1968 to 1977 (Abbott, Hepner and Clifford, 1980). Outbreaks were reported mainly in maternity, paediatric, psychiatric and geriatric units where faecal soiling of the ward is common. Of the 76 outbreaks where a source of infection was reported food was implicated in only 24 (30 per cent). In Scotland data collated at the Communicable Disease (Scotland) Unit shows that in the 11 years from 1973 to 1983 only 40 per cent of the 58 reported hospital salmonella outbreaks were food-borne. Fifty-one per cent of non-food-borne outbreaks occurred in maternity, paediatric and geriatric units compared with only 22 per cent of food-borne outbreaks. The Scottish data emphasize that salmonella outbreaks may be recurring problems in particular hospitals; 20 per cent of the affected hospitals had two outbreaks within a 6-year period (J. C. M. Sharp, personal communication). Salmonellas may also spread from a non-food common source such as contaminated carmine dye, pancreatin (Lang et al., 1967; Glencross, 1972) and inadequately sterilized endoscopes (Chmel and Armstrong, 1976). The course of these outbreaks tends to mimic person-to-person spread and the vehicle of infection may well be overlooked. But it seems that thorough investigations of salmonella outbreaks in hospitals are not often conducted. A recent prospective 2 year review (1980-82) in England and Wales