Education And Debate

Controversies: Is it time to stop searching for MRSA?

BMJ 1997; 314 doi: (Published 01 March 1997) Cite this as: BMJ 1997;314:664

Screening is still important

  1. Barry Cookson, director, Laboratory of Hospital Infectiona
  1. a Central Public Health Laboratory, Public Health Laboratory Service, Colindale, London NW9 5HT

    The number of patients acquiring methicillin and multiple antibiotic resistant Staphylococcus aureus (MRSA) infections while in hospital is increasing worldwide. This pattern varies between countries, within the same city, or even in different wards in the same hospital.1 2 3

    The revised British guidelines for MRSA control will shortly be circulated for comment.4 They will suggest different types of response and even screening strategies depending, for example, on local healthcare referral patterns, the type of MRSA, and the different categories of patients at risk.5 6

    In the United Kingdom, unlike in many other countries, when MRSA re-emerged in the early 1980s, epidemic strains (EMRSA–defined as MRSA isolated from two or more patients in at least two hospitals) were characterised and numbered at the Laboratory of Hospital Infection in Colindale.7 Initially they seemed to be confined to outbreaks in one region, but the isolates that have emerged in the 1990s (EMRSA-15 and EMRSA-16) are causing outbreaks of infection and colonisation in hospitals in more than one region,9 resulting in a fourfold increase in submissions of isolates for typing to the Laboratory of Hospital Infection in the past six years.

    Any MRSA that spreads in a hospital is a potential new EMRSA.8 9 Certain EMRSAs (strains 1, 3, 15, and 16) have spread more widely and over a longer period and perhaps could be termed “super” EMRSA. Patients infected with MRSA in hospitals are now involved in cycles of re-admission from the community to hospitals and causing renewed outbreaks.

    We cannot provide reliable predictors of virulence in the laboratory, and it is true that MRSA has rarely caused the primary sepsis seen in patients and healthcare workers caused by the infamous phage “80/81” S aureus of the 1950s and 1960s. Resistant strains can be as virulent as strains that are susceptible to antibiotics, but the virulence of both types can vary.4

    In Britain there is recent evidence that current strains are causing disease. In the second United Kingdom national prevalence survey conducted in 1993-4, MRSA comprised 5% of all infections, including 14 of the 228 surgical wound infections. The preliminary univariate analysis of this survey has shown that MRSA colonisation had the highest relative risk (5.09) for hospital acquired infection.10 The data from the Public Health Laboratory Service's bacteraemia reporting system also indicate worrying increases in MRSA infections.11 The incidence of MRSA infection had remained static (about 1.8%) between 1989 and 1991 but increased to 8.1% by 1994 and in the first half of 1995 was 13.5%. Other data on antibiotic susceptibility patterns indicate that these MRSA strains are most likely to be the current epidemic strains (EMRSA-15 and EMRSA-16). At least two studies (described by Wenzel et al12) have shown that these MRSA infections are additional to, and do not replace, strains that are sensitive to antibiotics. Controlling MRSA infection, therefore, will reduce overall infection rates.

    Screening for carriers rather than simply identifying infected patients has a major role in control of an outbreak and reduces the number of infections. Although detecting MRSA in routine clinical specimens provides important information, many studies show this to be inadequate.13 It will become more unsatisfactory as lengths of hospital stay shorten and the numbers of clinical specimens decrease.14 By switching to a strategy of identification and treatment of carriers in a large and prolonged MRSA outbreak in Spain, Coello and coworkers were able to reduce mortality from and infections with MRSA.15

    Searching and eradicating carriage of MRSA have been shown to be cost effective even in a large prolonged MRSA outbreak in a London tertiary referral hospital, in comparison with a large MRSA outbreak in Madrid.15 The screening and control measures resulted in a probable sevenfold reduction in the number of infected patients. Using recent costing data for British hospital infections, these researchers showed that the costs of extended lengths of stay and resources needed to treat the infected patients exceeded the costs of screening and control.16

    The re-emergence of MRSA as a serious problem has led to our relearning many of the lessons from 20 years ago. Infection control teams (which interact with the Laboratory of Hospital Infection) should encourage correct reflexive infection control practices and enable others to improve their quality of service.17 Infection control teams should also facilitate the review of antibiotic policies, as antibiotics have been identified as risk factors for acquiring MRSA, although recent data suggest that this might have been overestimated due to prolonged lengths of stay acting as a collinear confounding factor.18

    The new British guidelines will suggest ways in which MRSA rates can be used to encourage changes in infection control practices. Health workers should be given feedback on the number of new MRSA infections and colonisations detected by screening, as a measure of the success of control measures.12 19 The shortening lengths of hospital stay may mean that screening on discharge is required in certain situations.

    In certain parts of the world rates of MRSA colonisation are so high, and resources so scarce, that elimination is impossible. Damage limitation policies such as control of antibiotic prescribing are the sensible way forward. However, the increasing number of antibiotic resistances seen in some strains of MRSA, and the predicted eventual emergence of vancomycin resistance in such strains, makes MRSA control, and indeed antibiotic resistance in general, a matter of global importance.20


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    13. 13.
    14. 14.
    15. 15.
    16. 16.
    17. 17.
    18. 18.
    19. 19.
    20. 20.
    View Abstract

    Log in

    Log in through your institution


    * For online subscription