gms | German Medical Science

MAINZ//2011: 56. GMDS-Jahrestagung und 6. DGEpi-Jahrestagung

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie e. V.
Deutsche Gesellschaft für Epidemiologie e. V.

26. - 29.09.2011 in Mainz

Design and statistical methods for rare disease studies: A literature review

Meeting Abstract

  • Christine Porzelius - Institut für Medizinische Biometrie und Medizinische Informatik, Universitätsklinikum Freiburg, Freiburg
  • Janbernd Kirschner - Zentrum für Kinder- und Jugendmedizin, Universitätsklinikum Freiburg, Freiburg
  • Dieter Hauschke - Institut für Medizinische Biometrie und Medizinische Informatik, Universitätsklinikum Freiburg, Freiburg
  • Martin Schumacher - Institut für Medizinische Biometrie und Medizinische Informatik, Universitätsklinikum Freiburg, Freiburg

Mainz//2011. 56. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 6. Jahrestagung der Deutschen Gesellschaft für Epidemiologie (DGEpi). Mainz, 26.-29.09.2011. Düsseldorf: German Medical Science GMS Publishing House; 2011. Doc11gmds050

DOI: 10.3205/11gmds050, URN: urn:nbn:de:0183-11gmds0504

Published: September 20, 2011

© 2011 Porzelius et al.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by-nc-nd/3.0/deed.en). You are free: to Share – to copy, distribute and transmit the work, provided the original author and source are credited.


Outline

Text

To obtain valid and convincing results in interventional and epidemiological studies, it is essential that the study population is of sufficient size. For rare diseases, it is often problematic to recruit enough patients for a study, despite long recruiting phases and incorporating multiple centers. Additional problems when considering rare diseases are in many times a lack of placebo controls due to ethical reasons and the absence of a gold-standard for therapy, as the course as well as treatment of the diseases are often not well-understood. From a statistical point of view, some of these problems can be overcome by using specific study designs and statistical methods for the data analysis, that typically aim on increasing the power of the study. In a literature review, we investigate the design and methods of recently published studies on rare diseases, as well as statistical literature and guidelines on studies of rare diseases or studies with small number of observations. As a first attempt, all research articles published in the “Orphanet Journal of Rare Diseases” and in “Rare Tumors” were examined. Many of them are case studies or include descriptive statistics only. Another big part are observational studies based on registry data, where the sample size is naturally limited. In almost all articles found, standard statistical methods are used for data analysis. In a second part, methodological papers were considered. Often, the statement can be found that there are no designs specifically for rare disease studies and that all designs, that are recommended for small samples, can theoretically be used for large samples as well, even though they are typically more complex, as e.g. adaptive designs or cross-over trials. Only few statistical methods are developed explicitly for the application in rare diseases, but some considerations and modifications on the application of standard methods to small sample sizes can be found. Based on the investigated articles published in medical journals, we think that amongst medical researchers there is only few awareness on the statistical possibilities to improve the evidence of small sample studies. Apart from that, we observed a lack of systematic investigation of the design and the statistical methods for rare disease studies, so further research would be desirable. However, general recommendations are difficult in the field of rare diseases, as situations and sample size numbers vary extremely.