File
Davies L.1, Cook J.1, Price A.1, Beard D.1
1University of Oxford, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, Oxford, United Kingdom
Background: Surgical randomised controlled trials (RCTs) often present additional challenges in design, conduct and analysis to those evaluating pharmacological treatments. This is especially evident when the interventions being evaluated are quite different, for example, the comparison of surgery with a non-operative intervention such as physiotherapy. Specific methodological issues have been reported such as the impact of treatment preferences of both patients and participating clinicians on recruitment, difficulties with blinding and standardising interventions.
Purpose: The aim of this study was to systematically review the methodological challenges and limitations in musculoskeletal RCTs which compare a surgical and non-operative interventions (e.g. drug treatment or physiotherapy).
Methods: A search of the MEDLINE and CENTRAL databases between 2010 and 2015 was conducted. Randomised orthopaedic clinical trials which had a surgical and non-operative comparison such as physiotherapy were included. Data were extracted and summarised on the study characteristics such as target and actual recruitment, adherence to randomisation allocation, and in addition reported challenges to successfully conducting the trial amongst other details. A quality assessment was also conducted using the Cochrane Risk of Bias tool.
Results: In total, 6484 records were identified and 54 trials were included in the review. The majority of included studies were of fracture management (28, 52%). Twenty three (43%) were multi centre studies, with a sample size of 85 (median, IQR [50,150]). The main intervention comparisons of the included studies were of surgery and physiotherapy interventions (44, 81%). Methodological problems identified included a high proportion of participant crossover from their randomised intervention and difficulties reaching target sample size. Challenges reported by the authors included the impact of both patient and clinician preference for particular treatment on the number of potentially eligible patients recruited. In addition, various practical challenges were reported related to the nature of the interventions evaluated and health care systems in which the research was being conducted. This included difficulties implementing blinding of patients, care providers, and outcome assessors and post-randomisation delays in receiving an intervention.
Conclusion(s): This review demonstrated that trials which compare a surgical and non-surgical interventions have been successfully conducted for a number of conditions, however, several methodological issues were identified that have the potential to impact on the design, conduct and analysis of these trials.
Implications: If not adequately addressed, these issues may introduce significant bias and threaten the validity of the trial results. The planning and design of future studies should take into account and consider the specific challenges associated with the evaluation of surgical and non-surgical interventions.
Funding acknowledgements: Supported through funding from Health Education Thames Valley
Topic: Research methodology & knowledge translation
Ethics approval: Ethical approval was not required for this study
All authors, affiliations and abstracts have been published as submitted.
