Using systematic reviews for evidence-based practice

AORN Journal,  Sept, 2005  by Peggy Edwards,  Allyson Lipp,  Victoria Steelman

• E-mail
• Print
• Link

We read with interest AORN's recently updated "Recommended practices for surgical attire" (1) and were disappointed that recommended practice III (1 (p224)) did not make any reference to a systematic review published in 2002 and updated in 2004 on the use of surgical face masks. (2) The review reinforces AORN's message that there are differences in the literature and that further research is needed. The review also highlights the fact that studies have been published that were not of sufficient quality to base any decisions upon, even though these studies frequently are cited in other articles.

Systematic reviews (ie, meta-analyses) are accepted as the highest form of evidence in the evidence-based hierarchy and should, wherever possible, form the basis for recommended practices. (3,4) We hope that our efforts to highlight the significance of systematic reviews to your readers will not go unnoticed.

PEGGY EDWARDS

RN, BSc, DIP OPERATING PRACTICE

PATIENT SAFETY MANAGER (WALES)

NATIONAL PATIENT SAFETY AGENCY

LONDON

ALLYSON LIPP

RN, RNT, MA, MSc, PHD

PRINCIPAL LECTURER

SCHOOL OF CARE SCIENCES

UNIVERSITY OF GLAMORGAN

GLYNTAFF, PONTYPRIDD, UNITED KINGDOM

NOTES

(1.) "Recommended practices for surgical attire," in Standards, Recommended Practices, and Guidelines (Denver: AORN, Inc, 2005) 223-227.

(2.) A Lipp, P Edwards, "Disposable surgical face masks for preventing surgical wound infection in clean surgery," in The Cochrane Database of Systematic Reviews, issue 2 (Chichester, United Kingdom: John J. Wiley and Sons, Ltd, 2005). Abstract available at http://www.cochrane.org/cochrane/revabstr/ab0 02929.htm (accessed 26 July 2005).

(3.) A Lipp, "A guide to developing a systematic review," AORN Journal 78 (July 2003) 90-107.

(4.) A Lipp, "The systematic review as an evidence-based tool for the operating room," AORN Journal 81 (June 2005) 1279-1287.

Response. Ms Edwards and Ms Lipp raise some good points. First and foremost, I would like to applaud their interest in and the value they place on evidence-based practice. It is very important that our members look at AORN's recommended practices and ask questions from an evidence-based practice perspective. This systematic process is the only way that AORN can achieve excellence in clinical decision making and ensure our recommended practices continue to be respected by other disciplines.

Second, they have identified a source of evidence, the Cochrane database of systematic reviews, that often is overlooked by clinicians who are searching for the best evidence to guide practice. The Cochrane Collaboration was founded in 1993 by British epidemiologist Archie Cochrane, CBE, FRCP, FFCM. (1) It is an international, nonprofit organization dedicated to sharing up-to-date, accurate information about the effects of health care decisions. The Cochrane Collaboration produces and disseminates systematic reviews of clinical trials and other research studies evaluating the effects of interventions. These systematic reviews are published quarterly as part of The Cochrane Library.

Ms Edwards and Ms Lipp also are correct that a meta-analysis is the strongest, or highest, level of evidence used for decision making. A meta-analysis is a research method that combines the data from multiple studies that meet strict criteria for inclusion. The data are analyzed collectively. It is a very rigorous research method that requires advanced skills and statistical methods. The findings of a meta-analysis provide the best overall answer to a research question to determine best practices.

A systematic review, however, is a review of research findings, rather than a research method. It does not carry the weight of a meta-analysis in the hierarchy of levels of evidence. In a systematic review, the results of studies can be placed into a table or described sequentially. The data are not pooled for analysis. This process can be duplicated by someone reviewing the same research studies, and that is what was done to update AORN's "Recommended practices on surgical attire."

The systematic review that Ms Edwards and Ms Lipp describe (2) discusses two randomized clinical trials that investigated the relationship between surgical masks and surgical site infections in patients undergoing surgical procedures in the United Kingdom. In 1984, researchers studied wearing versus not wearing masks in the OR during gynecological procedures. (3) The study was discontinued after seven weeks when the third of five patients in the "no mask" group developed a surgical site infection. Although organisms involved were not causally linked to OR personnel, the researchers terminated the study because of a perceived potential risk to participants. Results of the study were not meaningful because the sample size was too small to draw conclusions.

In 1991, a larger study was undertaken in Sweden. (4) Practices were randomized, with masks worn by OR personnel during some weeks but not others. Data were collected for one year. A power analysis determined that 3,000 patients would need to be included in the study to demonstrate a difference in surgical site infection rates, an outcome that has many contributing variables. A total of 3,088 patients undergoing clean, general surgery procedures were included in the study. Patients undergoing open heart, orthopedic, and urologic procedures were excluded. Procedures during which personnel had colds or rhinitis also were excluded. Patients were monitored until after discharge to determine if any visible pus or cellulitis developed. The methods for monitoring the patients after discharge were not identified.