Nottingham Staphylococcus aureus population study: prevalence of MRSA among elderly people in the community

Participants, methods, and results

We used electoral ward level statistics from 1991 to combine the catchment areas of seven large general practices and provide a study population of which the demographic composition (age, sex, social class, ethnicity, and proportion of elderly people living alone) was representative of the Nottingham Health District, which in most respects is similar to England as a whole.4 With the general practitioners' cooperation, we invited 1615 randomly selected people aged 65 or over to take part in the study, excluding those whose permanent address was a residential or nursing home. We carried out the study at the time of administration of influenza vaccinations on the premises of the participating surgeries. Of those invited, 1047 (64.8%) elderly people presented for the investigation, and we enrolled 962 into the study. Samples consisted of swabs taken from both anterior nares. We used standard laboratory protocols to enrich and identify methicillin susceptible S aureus and MRSA and used SmaI macrorestriction to type them genetically. We collected data on baseline demographic variables as well as specific risk factors. We used the χ² test for univariate analysis of categorical variables, multivariate logistic regression to identify independent risk factors, and ward based Jarman deprivation scores to stratify data by socioeconomic characteristics.5

The sample comprised 1% of the population of the Nottingham Health District aged 65 and over. We found nasal carriage of methicillin susceptible S aureus in 257 people (26.7%, 95% confidence interval 24.1% to 29.8%). We isolated MRSA from eight people. The population prevalence of MRSA was 8 (3 to 14) per 1000 population. We identified no risk factors for carriage of methicillin susceptible S aureus. Carriage of MRSA was associated with several risk factors in univariate analysis (table), and two of these remained independently associated after logistic regression—hospital admission in the six months before the investigation (adjusted odds ratio 13.0, 2.5 to 68.2) and diabetes (6.8, 1.33 to 34.3). The presence of chronic skin ulcers was strongly associated, as a confounder, with both carriage of MRSA and previous hospital admission. Carriage of MRSA was independent of deprivation scores, indicating no association with lower living standards. All MRSA isolates were indistinguishable from the epidemic MRSA type 15 by genetic typing. This clone has become widely prevalent in English hospitals and was also the most common MRSA strain in the two major hospitals in Nottingham at the time of investigation.3

Comment

In contrast to its continuous occurrence as the major nosocomial pathogen in England, MRSA has not encroached into the community to a large extent, and the likelihood of finding healthy elderly people colonised with MRSA is relatively small. Carriage of MRSA can, however, be expected in patients who have chronic conditions or who have had recent admissions to hospital. MRSA in the community seems to be a consequence of a “spill over” of an uncontrolled hospital epidemic, and the few isolates found in the community are classic hospital strains. It continues to be safe to treat community acquired S aureus infections in England with conventional antistaphylococcal agents effective against methicillin sensitive strains, and third line antibiotics should be considered only when typical risk factors can be ascertained.