Letter to the Editor: The issue of retracted research
I am writing to highlight a potential issue with the article Beitz JM, Bolton LL. Systematic reviews and meta-analyses — literature-based recommendations for evaluating strengths, weaknesses, and clinical value. Ostomy Wound Manage. 2015;61(11):26–42. The article includes a study by Gethin and Cowman (2009) that may have been retracted (http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2702.2008.02558.x/abst...) around the time the review was published in your journal. I identified this discrepancy whilst conducting my own review and felt it was important to draw attention to the issue.
Richard Gray, RN, MSc, PhD (Lond.), FRSPH
Professor of Clinical Nursing Practice
La Trobe University, Melbourne, Australia; and Editor, Journal of Psychiatric and Mental Health Nursing
We thank Dr. Gray for his letter related to our publication and welcome his close perusal of the literature. We are pleased that our research endeavors are receiving scrutiny and raising questions to promote the best evidence-based literature.
In response to the concern explained in his letter to the editor, no change in the text, tables, or figures will be required if we cite the Gethin et al 2015 reference below. Both statements below and all figures remain factual wherever this study is mentioned, although it may require a codicil to page 34, noting that the article originally analyzed by Jull et al (reference 53) was retracted and a corrected analysis of the same dataset confirming statistical significance of the effects of honey on large, long-duration venous ulcers is now cited as:
Gethin G, Cowman S, Kolbach DN. Debridement for venous leg ulcers. Cochrane Database Syst Rev. 2015;9: CD008599. doi: 10.1002/14651858.CD008599.pub2.
In the text in Table 2 on page 34, column 1, illustrating homogeneity issues in meta-analyses, we wrote: “Two RCTs of clinically heterogeneous chronic venous ulcers53 were combined in a meta-analysis, despite clinically important differences. The RCTs differed in methodology (eg, 4 or 12 weeks’ treatment with honey-impregnated foam or alginate, compared to the same foam + hydrogel in 1 study or ‘usual care’ in the other study). Ulcers dissimilar in baseline area, duration, and slough were combined in a meta-analysis that obscured a clinically and statistically significant healing benefit of topical honey on larger, longer duration, less-likely-to-heal venous ulcers.62”
Likewise, on page 36 of our article we wrote: “In addition to the critical analysis of SR checklist flaws in Table 2, a pattern of 3 retractions was observed for honey interventions that may have potential to bias related literature and prevent access to RCTs, potentially biasing future SRs. Erasing from the literature a rigorously controlled RCT on 108 subjects reporting faster healing in large, sloughy venous ulcers dressed with foam dressings impregnated with honey compared to hydrogel62 presents a disservice to clinicians by rendering these clinically important data unavailable to inform practice or to include in future SRs. Without a clearer description of the rationale for retraction than ‘due to errors in the data analysis which affect the article’s findings,’ it would seem more appropriate to have published an erratum statement clarifying and correcting the data analysis error(s). This would allow the valuable clinical information gleaned from this important study to continue to inform clinical decisions.”
As noted above, on page 36 of our article, it was unusual to retract the full Gethin et al (2009) article rather than publish an appropriate statistical erratum. Inquiries regarding the source of the request for full retraction of the article were unable to rule out commercial interests. Instead, we focused on our purpose to inform readers how to find useful clinical results in systematic reviews relevant to their patients. Dr. Gray is invited to investigate this issue further if desired. Meanwhile, we trust that the substitution for our reference 62 of the above cited Gethin et al 2015 reference, which correctly analyzed the evidence from Gethin et al 2009, will address his concerns, and OWM has published a correction in this issue of the journal to this effect.
Despite the fact that in this instance other literature supports our conclusions, Dr. Gray raises one of the thorny issues and limitations associated with systematic reviews. As evidence changes, so do the nature and meaning of the subsequent reviews derived from it. It is a lesson worth remembering.
Janice M. Beitz, PhD, RN, CS, CNOR, CWOCN-AP, ANEF, FAAN
Professor of Nursing, WOCNEP Director
School of Nursing-Camden, Rutgers University
Laura Bolton, PhD
Adjunct Associate Professor of Surgery
Rutgers University Medical School
New Brunswick, NJ
These letters were not subject to the Ostomy Wound Management peer-review process.