Characterizing Clinical Trials in Plastic & Reconstructive Surgery: A Systematic Review of ClinicalTrials.gov from 2007 – 2020
Lawrence Z. Cai, MD; Ashraf A. Patel, BS; Vineeth M. Thirunavu, BA; Nicholas F. Hug, BA; Siyou Song, BA; Jennifer Li, BS; Robert R. Barghout, MD; Christopher J. Magnani, MPhil; Brandon E. Turner, MD, MSc; Jecca R. Steinberg, MD; Gordon K. Lee, MD
Background: Clinical trials form the backbone of evidence-based medicine, but the state of clinical trials in plastic and reconstructive surgery (PRS) has not been comprehensively studied. To that end, we explored the distribution of therapeutic areas that are under investigation, impact of funding on study design and data reporting, and trends in research patterns of all clinical trials in PRS.
Methods: Using the ClinicalTrials.gov database, we identified and extracted all clinical trials relevant to PRS that were submitted between 2007 and 2020. Studies were classified based on anatomic locations, therapeutic categories, and specialty topics. Kaplan-Meier curves were used to visualize early discontinuation and results reporting by funder, and log-rank tests assessed differences between curves. Cox proportional hazard was used to calculate adjusted hazard ratios (HR) for early discontinuation and results reporting.
Results: 3,224 trials that included 372,095 participants were identified. PRS trials grew at an annual rate of 7.9% in contrast to 4.5% in non-PRS studies. The therapeutic classes most represented were Wound Healing (41.3%) and Cosmetics (18.1%), while Lymphedema was the least-studied category (4.0%). Funding for PRS clinical trials is largely provided through academic institutions (72.7%), while industry and US government constituted a minority. Industry-funded studies were more likely to be discontinued early than those funded by academics (HR 1.89) or government (HR 1.92) and more likely to be non-blinded and non-randomized. (Fig 1) Academic-funded studies were the least likely to report results data within three years of trial completion (OR 0.87, p<0.0001). Factors that were associated with results data reporting within three years included studies conducted in high-income countries (OR 1.10, p<0.0001), at two or more centers (OR 1.09, p<0.0001), and conducted in Cosmetics (OR 1.09, p<0.0001).
Conclusions: A gulf exists in the representation of different PRS specialties among clinical trials. This study highlights the role of funding source in trial design and data reporting to identify a potential source of financial waste in the form of early discontinuation, a phenomenon more prevalent in industry-funded trials that is worthy of further investigation. These findings also stress the need for continued appropriate oversight to maintain a high level of rigor in clinical trials to improve guidelines and disease management for patients.
Back to 2021 Abstracts