This is the authors' abstract. We don't add key points for this paper.
Systematic reviews and meta-analyses have become an enticing research venue for authors of Plastic and Reconstructive Surgery in recent years. The advantage of a systematic review and meta-analysis is that the data are presumed to be readily available in the literature. Thus, the culling of the data forms an efficient exercise. However, there are many misconceptions and misapplications of systematic reviews that we would like to highlight to assist authors in spending their time on pertinent study questions rather than engaging in systematic reviews that have no impact or relevance to the practice of plastic surgery. In contrast to the traditional nonsystematic review in which the authors choose the articles for assessment based on their preferences, a systematic review sets inclusion and exclusion criteria for selecting articles from the literature to minimize selection bias. Researchers can work with a medical librarian to ensure a comprehensive search strategy is developed to capture all applicable studies. From a semantic standpoint, a meta-analysis is a more rigorous systematic review. It collates data from randomized controlled trials or high-quality epidemiology studies so that disparate findings can be combined statistically to yield a unified summary statistic. A meta-analysis is frequently used in epidemiology studies when there are conflicting results from the literature that influence the risk of an exposure. For example, coffee was hypothesized to increase the risk of pancreatic cancer. However, the risk of pancreatic cancer is relatively small, whereas the consumption of coffee is prevalent. This led to a number of epidemiology studies presenting conflicting results. A meta-analysis was performed to combine these studies and increase the sample size to determine a unified assessment of risk.1 The reason that a meta-analysis could be conducted is because the data were derived from high-quality studies with relative homogeneity in exposure and outcomes. Thus, combining the data into a summarized statistical framework was suitable. For clinical trials conducted carefully with rigid inclusion and exclusion criteria and similar randomization schemes among studies, merging the data should be valid. In addition, researchers should use risk-of-bias assessments, such as the Cochrane risk-of-bias assessment for randomized, controlled trials, to assess the various domains of bias (selection, performance, detection, attrition, reporting, and any other potential biases) that may affect the quality of the study. These assessments can also be included as visual aids within a systematic review and meta-analyses to demonstrate the quality of the data to the readers. In contrast, combining data from less rigorous study designs such as case series, retrospective studies, and poorly performed clinical trials leads to faulty, imprecise conclusions. Therefore, if researchers want to present a meta-analysis, data quality must be carefully considered.2–4 When performing a systematic review, the study question needs to be judged carefully. One needs to understand what fatal biases can sink a systematic review project.5 For example, if a systematic review is designed to compare pressure sore reconstruction, the researchers should ensure that the indications for these operations and, even more notably, the clinical information, such as the wound size, wound characteristics, and surgeons’ expertise, are appropriately considered and controlled. In most cases, given the heterogeneity and inability to control clinical information, uniform data are not available within the literature, which renders conducting a systematic review not feasible. Furthermore, techniques can be surgeon dependent. Surgeon proficiency bias is when surgeons perform various procedures with different success rates. This is common within plastic surgery, as plastic surgeons may apply alternative surgical options to reconstruct a unique wound or defect. In postmastectomy breast reconstruction, a center may commonly perform deep inferior epigastric perforator flaps with excellent outcomes and only rarely perform profunda artery perforator flaps, with poor outcomes. To conclude that deep inferior epigastric perforator flap procedures have greater success and fewer complications than profunda artery perforator flaps is an oversimplification because the decision-making process inherent in plastic surgery is not considered. It is inappropriate to conduct a systematic review when the literature does not provide these key ingredients for analysis. An impactful systematic review is conducted when there is controversy regarding surgical risks and high-quality data are available to be reviewed within the literature. These high-quality data should be derived from prospective cohort studies, or better yet, clinical trials. Without the rigor of the high-level study designs in the articles reviewed, systematic reviews and meta-analyses often conclude that a valid conclusion cannot be made because of the poor-quality data in the literature and better studies need to be performed. When this kind of conclusion is proposed, the project has failed to reach a suitable outcome. This is not the fault of the authors, but rather the lack of sufficiently robust data from the literature. However, one should be able to identify fatal errors before conducting a systematic review and search for an alternative strategy or abandon the inordinate amount of time in performing a systematic review that predictably will yield no suitable results.6 The advent of systematic reviews and meta-analyses has enriched plastic surgery. However, too many articles are being proposed that have major flaws, leading to instant rejection. Rather than spending an enormous amount of time on unproductive projects, the authors need to study the guides on systematic reviews and meta-analyses that were published in Plastic and Reconstructive Surgery.2–4 It is also incumbent on our authors to solicit the help of statisticians with an expertise in systematic reviews and meta-analyses to assess the data critically. These two study designs are arduous and require great amounts of effort in data collection and analysis. With the help of experts in this field to construct funnel plots, organize confidence intervals, and measure the homogeneity of the data collected, the quality of these studies in our literature will improve. Ultimately, this will enhance the practice of plastic surgery as evidenced by the two rigorously performed systematic reviews championed by the American Association of Plastic Surgeons.7,8 As a philosopher, Bruce Lee once said: Practice makes perfect. After a long time of practicing, our work will become natural, skillful, swift, and steady. An even better observation is from Vince Lombardi: Practice does not make perfect. Only perfect practice makes perfect.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Chung et al. (2022) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: