Background: Performance and efficiency in patient selection are essential for conducting clinical trials. Data on these are presented from a multinational trial.
Patients and methods: A randomized, double-blind, placebo-controlled, parallel-group study in asthma, with a screening phase followed after randomization by a treatment period, was selected. Number of patients screened and randomized by centre and country, centres achieving the minimum recruitment (> or = 10 patients randomized), and efficiency of investigators (randomized/screened x 100) were determined and compared.
Results: 564 patients, out of 836 screened, were randomized at 69 centres in 11 countries. Twenty-four centres (35%) randomized > or = 10 patients each, accounting for 70% (n = 395) of the total number recruited. Efficiency was significantly higher among these "high-performance" centres (81.4%; p < 0.001; OR, 4.7; CI 95%, 3.4-6.5) than in the remaining ones (48.1%). Five countries had > or = 2 "high-performance" centres. Efficiency was also significantly higher (p < 0.001) among those (370 randomized/455 screened, 82.1%) than in the remaining centres of the same countries (82/140; 58.6%; OR, 3.0; CI 95%, 2.7-4.7). A relevant number of centres (n = 17; 25%) randomized 0-1 patient (7 randomized/58 screened).
Conclusion: The analysis of patient selection in clinical trials showed that a minority of centres accounted for most of the patients recruited. Those are not only the most productive (more patients randomized) but also the more efficient (better quality of screening process).