Article type
Year
Abstract
Introduction:
Objectives: To study whether the reporting of clinical outcomes measured on ordinal and interval scales are adequate in relation to meta-analysis.
Methods: Systematic review of randomised group comparative trials which compared two nonsteroidal, anti-inflammatory drugs in patients with rheumatoid arthritis.
Results: Frequency of optimal reporting, defined as data in the original, ordered categories for global evaluation and pain, and as mean and SD (or SE) for number of tender joints and grip strength, and if a visual analogue scale had been used to measure pain. Remits: 145 trials were included. The median sample size was 60 patients. Sample size increased over time (1966 to 1997) as did the quality of the reporting for three of the four variables. Global evaluation was optimally reported in 52 of the 128 trials (41%) in which it was recorded. Pain was optimally reported in 27 out of 98 trials (28%), number of tender joints in 41 out of 123 trials (33%) and grip strength in 34 out of 124 trials (27%). Even adopting rather broad criteria, only about half of the data were reported in a potentially useful way for a meta-analysis.
Discussion: It can be estimated that about 100,000 randomised trials of health care interventions have been reported in a way that is inadequate for meta-analysis. As most trials are grossly underpowered, this deficit needs urgent improvement.
Objectives: To study whether the reporting of clinical outcomes measured on ordinal and interval scales are adequate in relation to meta-analysis.
Methods: Systematic review of randomised group comparative trials which compared two nonsteroidal, anti-inflammatory drugs in patients with rheumatoid arthritis.
Results: Frequency of optimal reporting, defined as data in the original, ordered categories for global evaluation and pain, and as mean and SD (or SE) for number of tender joints and grip strength, and if a visual analogue scale had been used to measure pain. Remits: 145 trials were included. The median sample size was 60 patients. Sample size increased over time (1966 to 1997) as did the quality of the reporting for three of the four variables. Global evaluation was optimally reported in 52 of the 128 trials (41%) in which it was recorded. Pain was optimally reported in 27 out of 98 trials (28%), number of tender joints in 41 out of 123 trials (33%) and grip strength in 34 out of 124 trials (27%). Even adopting rather broad criteria, only about half of the data were reported in a potentially useful way for a meta-analysis.
Discussion: It can be estimated that about 100,000 randomised trials of health care interventions have been reported in a way that is inadequate for meta-analysis. As most trials are grossly underpowered, this deficit needs urgent improvement.