Medical and health economic evaluation of prevention- and control measures related to MRSA infections or -colonisations at hospitals.

Introduction: Methicillin-resistant Staphylococcus aureus (MRSA) are dangerous agents of nosocomial infections. In 2007 the prevalence of MRSA is 20.3% in Germany (Oxacilline-resistance according to EUCAST-criteria [EUCAST = European Committee on Antimicrobial Susceptibility Testing]).

Objectives: Which measurements are effective in the prevention and control of MRSA-infections in the hospital?How effective are contact precautions, screening, decolonisation, education and surveillance?Which recommendations can be given to health care politics on the basis of cost-effectiveness studies?Have there been any adverse effects on patients and clinical staff?What kind of liability problems exist?

Methods: Based on a systematic review of the literature studies are included which have been published in German or English language since 2004.

Results: 1,508 articles have been found. After having surveyed the full text, 33 medical, eight economic and four ethical/juridical studies are included for the Health Technology Assessment (HTA) report. The key result of the HTA report is that different measurements are effective in the prevention and control of MRSA-infections in hospitals, though the majority of the studies has a low quality. Effective are the conduction of differentiated screening measurements if they take into account the specific endemic situation, the use of antibiotic-control programs and the introduction and control of hygienic measurements. The break even point of preventive and control measurements cannot be defined because the study results differ too much. In the future it has to be more considered that MRSA-infections and contact precautions lead to a psycho-social strain for patients.

Discussion: It is hardly possible to describe causal efficacies because in the majority of the studies confounders are not sufficiently considered. In many cases bundles of measurements have been established but not analyzed individually. The internal and external validity of the studies is too weak to evaluate single interventions. Hygienic measurements prove to be effective in combination with other measurements. But it cannot be said which of the single measurements (gloves, washing hands, wearing gowns or masks) has the strongest effect on the reduction of MRSA. It is irritating that there are high differences in the compliance concerning hand hygiene between different studies. A general decolonisation is questionable for different reasons: first because of the side-effects for patients, second because of the high rate of spontaneous remissions in the untreated control group, third because of the differentiated process from colonisation to infection. Severalfold Hawthorne effects have been reported. One of them is that the competition between hospitals to reduce MRSA-rates leads already to a reduction.

Conclusions: It is evident that selective screening programs of risk patients considering the particular MRSA-prevalence are of use. The application of rapid tests seems to be only recommendable for risk patients and a high MRSA-prevalence. The improvement of the compliance of hand hygiene should be the basis of any prevention strategy. Training of staff members (with feedback mechanisms) is effective to improve compliance and to optimise the use of antibiotics. Antibiotic management programs are effective as well. Obviously multimodal approaches can lead to overadditive effects. Therefore the catalogue of preventive and control measurements has to be further evaluated. Good cost-efficacy studies are missing in Germany. The psychosocial effects of MRSA-infections are not researched in Germany. There is only punctual information on the risk management of hospitals.