Multilevel models are applicable to both the quantitative data and categorical variables. We used the methods, including the multilevel models, analysis of covariance and CMH chi-square test, to analyse different types of data, to explore the application of multilevel models in the analysis of the multicenter clinical trial center effect. The results showed that the analysis of covariance is more sensitive to find the center effect for quantitative data, while multilevel models are more sensitive to categorical variables. It can be seen that results with different analytical methods for center effect are not the same, and the most appropriate method should be selected in accordance with the characteristics of data, the objective of research, and the applicable conditions of the various methods in practical use.
Objective To evaluate the quality of anticoagulant therapy in patients with mechanical prosthetic valve replacement during early period through time in therapeutic range(TTR) and fraction of time in therapeutic range(FTTR), and to provide an objective evidence for further improving quality of anticoagulant therapy. Methods All the patients were followed and registered in hospital and at outpatient clinic from July 2012 through April 2014, with a maximum of 6 months after surgery. Targeted international normalized rate (INR) was 1.60 to 2.20, acceptable INR was 1.50 to 2.30. And warfarin weekly dosage adjustment was used as the strategy of anticoagulation management. Adjusting the warfarin dosage when INR was beyond acceptable INR. Events of bleeding, thrombogenesis and thromboembolism, TTR and FTTR of these patients during the follow-up were collected to evaluate quality of anticoagulant therapy in these patients. Results A total 477 patients and 2 755 reports of INR values were included for data analysis. The follow-up time was 78 918 days. Values of INR varied from 0.92 to 7.72(1.83±0.64). Required weekly doses of warfarin in target INR and acceptable INR were 5.00–35.00(18.15±3.99) mg/week and 5.00–39.38(18.29±4.08) mg/week. TTR of target INR and acceptable INR was 36.85%(27 079.5 d/78 918.0 d) and 49.84% (39 331.5 d/78 918.0 d), respectively. FTTR of target INR and acceptable INR was 37.31% (1 028 times/2 755 times), 50.01% (1 378 times/2 755 times), respectively. TTR of target INR and acceptable INR was 46.04%(3 902.5 d/8 475.5 d), 59.49%(5 042 d/8 475.5 d) when the patients’ follow-up was up to six months and FTTR of target INR and acceptable INR value of these patients was 46.81%(206 times/440 times), 60.45%(266 times/440 times). During the follow-up, there were 3 thromboembolism events, 1 transient physical abnormal activity, and 1 thrombogenesis in the left atrium, and there was no bleeding and death events. Conclusion The strategy of anticoagulation management used in our study is reasonable. In order to further improve the patients’ quality of anticoagulant therapy, it is necessary to start anticoagulation after operation as soon as possible, to strengthen the education of patients with anticoagulant knowledge and to increase INR test frequency properly.