Artikel
Multicentre Clinical registries: a necessary element in health care research !? Example: The Berlin Myocardial Infarction Registry
Suche in Medline nach
Autoren
Veröffentlicht: | 14. September 2004 |
---|
Gliederung
Text
Introduction
The study of diseases, wheter in the framework of controlled clinical trials or observational studies, faces the problem of heterogeneity of subjects. Heterogeneity usually is tackled by the application of statistical methods which lead to general results valid for populations but not for single individuals. In statistical models, the true heterogeneity is viewed as a random variation produced by a stochastic mechanism. However, apart from systematic errors the statistical error produced by random variation of results may be considerable, especially for the analysis of subgroups. For specific problems, especially in the multifactorial settings, sample sizes of thousands of patient will be necessary to obtain sufficient results. This is true for controlled clinical trials but also for clinical epidemiology working with data from hospital based registries.
Methods
With the use of the Berlin Myocardial Infarction Registry (informations: www.herzinfarktregister.de) it will be shown how the cooperation between different hospitals can be organized, which kind of documentation and follow up data are necessary, how the quality assurance could be organized in a regional multicentre registry, which efficiency a remote data entry system should have, what is neccessary to fulfill the demands of the data protection agencies and which biometrical considerations and medical implications can be made.
Results
It will be shown that we need well designed and properly prospective registries continuing for several years [Ref. 1] to improve the quality of care and the outcome of the patients.
References
- 1.
- Zeymer U, Senges J. Why do we need prospective registries in patients with acute myocardial infarction? European Heart Journal 2003, 24: 1611-12.