QUALITY MANAGEMENT FOR CLINICAL STUDIES ACCORDING TO
THE INTERNATIONAL BIOMETRICS REGULATIONS
How can statistical evaluations lead to wrong results?
Rózsa H. Nienhaus1 and Sándor Kiss2
1WWU Münster ZIV (Universitätsrechenzentrum), Germany
Röntgenstr. 9-13 D-48149 Münster
Tel. 0049-251-8331683
E-Mail. nienhau@uni-muenster.de
2DATAN Datenanalyse Havixbeck, Germany
Auf dem Blick 57 D-48329 Havixbeck
Tel. 0049-2507-7624 Fax. 0049-2507-4476
E-Mail. datan@t-online.de
www.datan-datenanalyse.de
"No statistic is better than the data used for evaluation"
(R. H. Nienhaus)
This lecture gives an overview about the most important requirements
which have to be kept in the course of a clinical study to ensure the desired
quality:
-
Answers to precisely formulated questions concerning advanced discoveries
of the effectiveness in routine applications of test drugs with meaningful
results
-
Confirmation of old knowledge as well as discovering new knowledge concerning
adverse drug reactions
-
To plan a project (a study)
-
To carry out a project (a study)
-
Data Management
-
Quality guarantee according to GCP, IBH, and SOP (Standard Operating Procedures)
-
Planning and preparation of a statistical evaluation
-
Statistical evaluation and statistical report according to the latest developments
in research of the involved disciplines
-
Graphical summary which shows - how statistical evaluations can lead to
wrong results - in which phases of carrying out a project (a study) there
still is the possibility of correcting any mistakes
References:
-
GCP (Good Clinical Practice)
-
IBH (International Biometry Harmionisation)
-
Statistical Evaluations (R.H. Nienhaus, lecture at the University of Münster,
1999)
-
Quality Management in Therapy-Surveillance-Studies from a Biometrical Point
of View (R.H. Nienhaus, MIDAS-conference, 1998