Abstract: AIM: To investigate the influence of dosage, therapeutic course, case, group (trial group and contrast group) and publication year on the effectiveness of randomized contrast trials.METHODS: Meta analysis literatures were retrieved from the fulltest data base of CNKI journals in the “Digital Library of Chinese Hospital”.Multiple elements analysis were carried out by binomial logistic regression.RESULTS: One standard-matched literature was got, including 26 clinical contrast trials for the treatment of hypertension by amlodipine, in which 23 cases were randomly divided and 9 cases were treated with amlodipine as the contrast.The course of treatment and groups were both factors for the effectiveness;The therapeutic advantage of trial groups was 2.445-3.690 times as that of contrast.CONCLUSION: In the random contrast test, the tested groups were possibly exaggerated, while the contrast groups depreciated.In the study on Meta analysis and evidence-based medicine, the attention should be paid for the group-bias influence.

Key words: group-bias, amlodipine, hypertension, random contrast clinical trial, Meta analysis, evidence-based medicine, proof study