Reports on surgical site infections and hospital acquired bacteraemias from the Nosocomial Infection National Surveillance Scheme
The Nosocomial Infection National Surveillance Scheme (NINSS) recently issued reports on the surveillance of surgical site infection and hospital acquired bacteraemia in English hospitals (1,2). NINSS is a system to monitor hospital acquired infection (HAI) and help hospitals to reduce their incidence of infection by identifying areas where clinical practice and infection control procedures may need to be reviewed. The scheme was set up and funded by England’s Department of Health and the Public Health Laboratory Service (3).
Ninety-six English hospitals in the eight NHS regions took part in surveillance of surgical site infection between October 1997 and September 19991. Data from 28 407 operations in 12 categories of surgery yielded reports of 1212 infections (4.3%), the incidence of which varied between hospitals for all surgical categories. Limb amputation and large bowel surgery most commonly led to infection in the nine surgical categories in which at least 10 hospitals participate. The more serious deep or organ/space (as opposed to superficial) infections accounted for at least 25% of surgical site infections in most categories. The incidence of surgical site infection rose with the number of risk factors from the US National Nosocomial Infections Surveillance System risk index (health of patient before surgery, duration of operation, wound class [likelihood of wound contamination during surgery]) for abdominal hysterectomy, coronary artery bypass grafts, hip prosthesis, large bowel surgery, and vascular surgery. For knee prosthesis, limb amputation, or open reduction of a long bone fracture the trend was less clear. Forty-seven per cent of the pathogens identified as causing infections were staphylococci: 81% of these were Staphylococcus aureus, 61% of which were resistant to methicillin (MRSA). MRSA was the commonest cause of surgical site infection in large bowel and vascular surgery, limb amputation, and open reduction of long bone fractures.
The 61 hospitals that took part in the surveillance of hospital acquired bacteraemia from May 1997 to April 1999 provided information on over a million patients, among whom 3824 episodes of bacteraemia were identified in 3629 patients (3.6 patients per 1000 admissions; mean of 0.6 bacteraemias per 1000 patient days) (2). Rates of hospital acquired bacteraemia varied widely between specialties within hospitals, the highest being in general and paediatric intensive care units (ICUs) and haematology units. Six specialties accounted for 72% of all bacteraemias: haematology, general ICUs, general medicine, general surgery, geriatric medicine, and nephrology. Nearly a half of the isolates causing hospital acquired bacteraemia were staphylococci: 24% were S. aureus, 47% of which were MRSA. More than one organism was identified in 11% of hospital acquired bacteraemias. Two thirds of bacteraemias whose source was known were associated with an intravascular device (most commonly central intravenous catheters) or with other devices, such as urinary catheters or ventilators.
Rates of HAI, stratified by operation type and risk index for surgical site infection and clinical specialty for hospital acquired bacteraemia, vary between hospitals. This variation within defined groups of patients provides an opportunity to review and compare clinical and infection control practice within and between hospitals. NINSS can advise on the interpretation and further investigation of results for hospitals with higher than average rates. Copies of these reports and details of the surveillance scheme are available from the Nosocomial Infection Surveillance Unit, PHLS Central Public Health Laboratory, 61 Colindale Avenue, London NW9 5HT.
Reported by Andrew Pearson, Nosocomial Infection Surveillance Unit, PHLS Central Public Health Laboratory, London, England. (Adapted from CDR Weekly 2000; 10: 213,6 (16 June).)