Reporting Characteristics of Negative Pressure Wound Therapy Systematic Reviews

Article type
Authors
Olaiya O1, Oyesile A2, Huynh M3, Alagabi A4, Mbuagbaw L5, McRae M3
1Michael G. DeGroote School of Medicine, McMaster University, Hamilton, Ontario
2University of Western Ontario Faculty of Health Science, London, Ontario
3Division of Plastic Surgery, Department of Surgery, McMaster University, Hamilton, Ontario
4Faculty of Health, York University, Toronto, Ontario
5Department of Health Research Methods, Evidence and Impact, McMaster University, Hamilton, Ontario
Abstract
Background:
Negative pressure wound therapy (NPWT) is a medical device used for wound management. Clinical indications for NPWT include chronic and acute wounds, dehisced incisions, diabetic wounds, pressure ulcers, meshed skin grafts, or flaps. Systematic reviews of randomized controlled trials offer the highest level of evidence. Reliable inferences can only be drawn from reviews that are methodologically sound and adequately reported. The quality of previously conducted NPWT systematic reviews are recondite.
Objectives:

Objectives:
We aim to compare Cochrane and non-Cochrane reviews and investigate the conduct and reporting characteristics of systematic reviews on the effectiveness of NPWT for treating various wounds.

Methods:
We searched MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews on January 06, 2020. Systematic reviews that included patients receiving wound care using NPWT were included and reporting characteristics were extracted by one reviewer and independently double checked by another reviewer.

Results:
We identified 30 systematic reviews for inclusion, after screening 646 title and abstracts and 168 full-text articles. There were five Cochrane reviews and 25 non-Cochrane reviews included. The median year of publication was 2017 (IQR 2013 – 2019) and 11 studies (36.7%) were published in 2019. Majority of the studies were from the Europe (53%) and very few from North America (17%). The included systematic reviews incorporated a median of nine studies (IQR 5-17) with a median of 872 patients (IQR 659 to 1689). There were very few non-Cochrane systematic reviews (12%, 3 studies) which reported registering their protocol. Only one-fifth of the non-Cochrane reviews specified the outcomes that were eligible for inclusion in the review. Nine non-Cochrane reviews (36%) failed to include outcomes related to harm in the review. Each Cochrane systematic reviews reported publishing their protocol, pre-specifying outcomes, and included adverse event outcomes. Cochrane reviews demonstrated higher odds of reporting adverse effects (OR = 33.3, 95% CI 5.5 to 1355.5). There was no difference in odds of performing meta-analysis between Cochrane and non-Cochrane reviews (OR = 0.390, CI 0.04 to 3.87). Few (16%) non-Cochrane systematic reviews used the GRADE approach to outline the certainty of the evidence. Risk of bias was assessed in 88% of non-Cochrane reviews but was only incorporated into the abstract in 40% of studies.

Conclusions:
While outcome data is being pooled and presented in both non-Cochrane and Cochrane reviews, non-Cochrane reviews on NPWT lack adequate conduct, reporting, and critical appraisal. Novel approaches are required to facilitate better conduct and reporting in non-Cochrane systematic reviews. The results of NPWT non-Cochrane systematic reviews should be used and interpreted with caution by patients or healthcare consumers.