|
Rapid screening tests for meticillin-resistant Staphylococcus aureus at hospital admission: systematic review and meta-analysis |
Tacconelli E, De Angelis G, de Waure C, Cataldo MA, La Torre G, Cauda R |
|
|
CRD summary The review assessed the effect of methicillin-resistant Staphylococcus aureus (MRSA) detection, using rapid screening tests, on the incidence and acquisition rate of hospital-acquired MRSA; it found that data did not support use of rapid testing alone to identify MRSA carriers or reduce acquisition rate. Despite some limitations in the study selection and analysis, these conclusions are likely to be reliable. Authors' objectives To assess the effect of methicillin-resistant Staphylococcus aureus (MRSA) detection, using rapid screening tests, on the incidence and acquisition rate of hospital-acquired MRSA infections. Searching MEDLINE, EMBASE, Science Citation Index and the Cochrane Library were searched from July 1965 to February 2009. Search terms were reported. Searches were restricted to published articles and no attempt was made to obtain information from unpublished studies; there were no language restrictions. Bibliographies of retrieved studies were screened for additional articles. Study selection Randomised, non-randomised, or observational studies that compared incidence and/or acquisition rate of MRSA in hospitals or wards, in which active screening for MRSA carriers was done at admission, were eligible for inclusion. Studies were required to use rapid molecular tests, and to include control groups from hospitals or wards in which active screening was done with culture alone or not at all. Studies that reported only data on the diagnostic accuracy of rapid tests or culture for MRSA detection were excluded.
Primary outcomes were MRSA acquisition rate and incidence of infections (bloodstream and surgical) caused by MRSA. Secondary outcomes were sensitivity and specificity of the rapid tests, and median turnaround time from admission to reporting of the results for rapid tests compared with culture.
Included studies were conducted in a variety of settings, which included both medical and surgical wards. The polymerase chain reaction (PCR) rapid screening methods used were GeneOhm, or in-house methods. All included studies used nasal samples. In the majority of included studies, control groups received no screening (five studies), or screening using culture was undertaken only in high risk patients (two studies).
One reviewer screened studies for inclusion. Assessment of study quality The methodological quality of clinical trials was assessed using the Jadad and Chalmers scales. Criteria included adequacy of randomisation and blinding, and management of withdrawals. The maximum quality score was 5 points. Studies with a score higher than 2 points were considered to be of good quality.
The methodological quality of cohort studies was assessed using quality scale for observational studies. This assessed study design and selection bias (six items), information bias (five items), adjustment and matching for confounders (one item), and data collection and analysis (five items). Data analysis items were scored 0.5 and all other items were scored 1; the maximum possible score was 14.5.
Two reviewers independently scored methodological quality. Data extraction Data were extracted on the acquisition rate and incidence of bloodstream infections and surgical-site infections caused by MRSA in locations where active screening was done using rapid tests compared with those where screening was done with culture alone, or not at all. Acquisition rate was defined as the incidence of hospital acquired MRSA colonisation and infection per 1000 patient-days. Hospital-acquired colonisation and infection were defined as MRSA detection in screening swab specimen or clinical specimen 48 hours or more after admission. Incidence of hospital-acquired MRSA bloodstream infections was defined as number of cases per 1000 patient-days. Incidence of hospital acquired MRSA surgical-site infections was defined as number of cases per 100 surgical procedures. Authors were contacted for any missing/unclear data.
Relative risk (RR), with 95% confidence intervals (CIs) was calculated for each study and outcome.
Data were extracted independently by three reviewers, using standardised forms in a computerised database. Any disagreements were resolved by consultation with a fourth reviewer. Methods of synthesis Pooled estimates of relative risks, with 95% confidence intervals, were estimated using random-effects models. If relative risk was not directly available, it was calculated by dividing the absolute risk in patients who received rapid screening by the absolute risk in control patients.
Pooled weighted averages (weighted by sample size) were calculated for sensitivity and specificity where applicable.
The Q test (by inverse variance weights method) and I2 test were used to assess between study heterogeneity and heterogeneity was considered significant at p<0ยท10. Publication bias was assessed using the Begg funnel plot. Results of the review Ten studies, with over 175,000 participants, were included in the review. Sample size ranged from 338 to 101,556 participants. Most studies were interventional cohorts; two studies used a crossover design, one used a before-and-after design and one was a time-series analysis. Only one study was a cluster-randomised, unblinded, crossover trial. In five studies, the control group received methicillin-resistant Staphylococcus aureus (MRSA) screening using culture, and in the remaining five studies the control group was not screened at all. Rapid screening (polymerase chain reaction PCR) results were confirmed, using culture, in four studies. The only clinical trial scored 3 on the Jadad scale and 34.5 on the Chalmers scale. The median quality score of the interventional cohort studies was 7.5 (range 4.5 to 9.5).
Use of a rapid screening test did not significantly reduce the MRSA acquisition rate compared with screening using culture (RR 0.87, 95% CI 0.61 to 1.24) in four studies with significant between study heterogeneity (p=0.027). The use of a rapid screening test significantly reduced the risk of MRSA bloodstream infections compared with no screening (RR 0.54, 95% CI 0.41 to 0.71) in three studies with no significant between study heterogeneity (p=0.268), but did not significantly reduce the risk of MRSA surgical-site infections (RR 0.69, 95% CI 0.46 to 1.01) in five studies with significant between study heterogeneity (p=0.012).
The pooled estimate on accuracy of the rapid screening tests for MRSA showed a sensitivity of 87.7% (95% CI 83.4 to 91.9) and a specificity of 98.1% (95% CI 94.5 to 100), based on data from two studies.
Turnaround time was less than one working day for rapid screening polymerase chain reaction tests and two to four working days for conventional culture, based on data from four studies.
There was no evidence of publication bias. Authors' conclusions The authors concluded that their data did not support use of rapid testing alone to identify MRSA carriers or to reduce acquisition rate in wards in which active screening with enrichment cultures linked to contact isolation were already in use. Active screening for MRSA was more important than the type of test used. CRD commentary The review stated a clear objective, to assess the effect of MRSA detection, using rapid screening tests, on the incidence and acquisition rate of hospital-acquired MRSA infections. Appropriate inclusion criteria were defined and a range of sources were searched for relevant studies, without restriction by language. However, searches were restricted to published studies and publication bias was not assessed. The methodological quality of included studies was assessed and the results of this assessment were discussed in the text. Data extraction and quality assessment processes included measures to reduce error and bias, but the selection of studies for inclusion was completed by one reviewer, increasing the possibility of error and/or bias at this stage. The calculation of summary estimates of relative risk for heterogeneous data sets was of questionable value. However, the authors' overall conclusion, that the current data did not support the use of rapid testing alone to identify MRSA carriers or to reduce acquisition rate, is likely to be reliable. Implications of the review for practice and research Practice: The authors stated that, when considering decisions such as mandatory legislation for MRSA universal screening, policy makers should be aware of the limits and the heterogeneity of the available evidence.
Research: The authors stated that rigorous assessment using randomised controlled clinical trials is needed to determine the effectiveness of rapid tests and their dissemination and implementation strategies, and that a cluster randomised design would be appropriate. They further stated that future research should incorporate economic evaluations, and compare culture screening with rapid screening. It is also important to define the target population for screening (e.g. surgical, dialysed, or elderly patients) based on the incidence of MRSA. Bibliographic details Tacconelli E, De Angelis G, de Waure C, Cataldo MA, La Torre G, Cauda R. Rapid screening tests for meticillin-resistant Staphylococcus aureus at hospital admission: systematic review and meta-analysis. Lancet Infectious Diseases 2009; 9(9): 546-554 Indexing Status Subject indexing assigned by NLM MeSH Animals; Carrier State /diagnosis /drug therapy /microbiology; Clinical Trials as Topic; Cross Infection /microbiology /prevention & Diagnostic Tests, Routine /methods; Humans; Mass Screening /methods; Methicillin-Resistant Staphylococcus aureus /isolation & Sepsis /microbiology /prevention & Staphylococcal Infections /diagnosis /drug therapy /microbiology /prevention & Surgical Wound Infection /microbiology /prevention & control; control; control; control; purification AccessionNumber 12009108194 Date bibliographic record published 21/10/2009 Date abstract record published 17/02/2010 Record Status This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn. |
|
|
|