Article type
Year
Abstract
Background: Randomized controlled trials (RCTs) provide the best evidence for efficacy of health-care interventions. Carefully planned and well-executed RCTs give us the best estimates of treatment effect and can thusly guide clinical decision making. Although, trials which lack methodological rigour cause over- or underestimations of treatment effect sizes due to bias or confounding factors. The methodological quality of trials directly influences results and conclusions of systematic reviews. At present, 21 systematic reviews have been completed within the Cochrane Back Review Group (CBRG) and are included in the Cochrane Library.
Objectives: The objectives of the present study are to describe the quality criteria used within CBRG systematic reviews and the quality of trials assessed within these reviews.
Methods: The following data were independently extracted by two individuals from 295 randomized controlled trials of spinal disorders included in 21 systematic reviews: 1. Criteria used within each review to judge the methodological quality of trials, 2. Quality ratings of individual trials as reported within the reviews. Discrepancies in extraction were resolved by consensus. One individual extracted inter-rater reliability ratings and number of high quality and low quality trials reported within reviews. Trials were divided into four groups: surgery, psychosocial, pharmacological, or physical medicine. Descriptive statistics calculated across all reviews included: 1. Percentage of total trials assessed on each of the methodological categories, 2. Percentage of Y, N and DK ratings for trials, 3. Percentage of high and low quality trials and, 4. Percentage of trials that met the minimum quality criteria as outlined in the method guidelines for the CBRG. A chi-square was used to test the difference in high verses low quality trials across decades (1960s - 2000s).
Results: Inter-rater reliability was 98.9% for methodological quality criteria extraction for 295 trials from the 21 reviews. Nine separate quality criteria scales/checklists were used across the 21 reviews. Less than 50% of the 295 trials blinded the outcome assessors, described the cointerventions, had equal cointerventions, had an adequate length of follow-up, were appropriately randomized, reported compliance, had allocation concealment, blinded patients, blinded care-providers, or reported adverse events. Surgery trials had the greatest number of high quality trials and psychosocial trials had the smallest number of high quality trials. The number of high quality of trials increased across time from the 1960s to the 2000s.
Conclusions: Continued efforts are required to improve both the methodological and reporting quality of trials testing interventions of spinal disorders. The quality of primary studies is essential to produce unbiased systematic reviews. It is suggested that those planning to perform controlled trials of spinal or neck disorders refer to the revised CONSORT statement for specifics regarding conduct and reporting. It is also suggested that reviewers of these trials refer to the CBRG updated method guidelines and the QUOROM statement in designing and conducting reviews.
Objectives: The objectives of the present study are to describe the quality criteria used within CBRG systematic reviews and the quality of trials assessed within these reviews.
Methods: The following data were independently extracted by two individuals from 295 randomized controlled trials of spinal disorders included in 21 systematic reviews: 1. Criteria used within each review to judge the methodological quality of trials, 2. Quality ratings of individual trials as reported within the reviews. Discrepancies in extraction were resolved by consensus. One individual extracted inter-rater reliability ratings and number of high quality and low quality trials reported within reviews. Trials were divided into four groups: surgery, psychosocial, pharmacological, or physical medicine. Descriptive statistics calculated across all reviews included: 1. Percentage of total trials assessed on each of the methodological categories, 2. Percentage of Y, N and DK ratings for trials, 3. Percentage of high and low quality trials and, 4. Percentage of trials that met the minimum quality criteria as outlined in the method guidelines for the CBRG. A chi-square was used to test the difference in high verses low quality trials across decades (1960s - 2000s).
Results: Inter-rater reliability was 98.9% for methodological quality criteria extraction for 295 trials from the 21 reviews. Nine separate quality criteria scales/checklists were used across the 21 reviews. Less than 50% of the 295 trials blinded the outcome assessors, described the cointerventions, had equal cointerventions, had an adequate length of follow-up, were appropriately randomized, reported compliance, had allocation concealment, blinded patients, blinded care-providers, or reported adverse events. Surgery trials had the greatest number of high quality trials and psychosocial trials had the smallest number of high quality trials. The number of high quality of trials increased across time from the 1960s to the 2000s.
Conclusions: Continued efforts are required to improve both the methodological and reporting quality of trials testing interventions of spinal disorders. The quality of primary studies is essential to produce unbiased systematic reviews. It is suggested that those planning to perform controlled trials of spinal or neck disorders refer to the revised CONSORT statement for specifics regarding conduct and reporting. It is also suggested that reviewers of these trials refer to the CBRG updated method guidelines and the QUOROM statement in designing and conducting reviews.