Call edition 2012

A new edition of the Prize of the Belgian Development Cooperation has been launched. This call is open until March 31st, 2011. You can read in the regulations whether you comply with the criteria for participation.

I would like to submit my application

Regression Models for Disease Prevalence with Diagnostic Tests on Pools of Serum Samples

Stijn VANSTEELANDT researcher laureate
stijn.vansteelandt@rug.ac.be

°1967 Belgium
Licence in mathematics (applied mathematics), Universiteit Gent, Belgium, 1998

Regression Models for Disease Prevalence with Diagnostic Tests on Pools of Serum Samples

Contagious diseases remain the major health problem in developing countries. Strengthening health services and controlling these diseases, not in the least AIDS, are the key to improving public health and contribute towards poverty relief and development.

Surveillance constitutes an essential activity of disease control. However, the number of diagnostic tests that has to be carried out makes adequate surveillance expensive and hardly feasible in developing countries. Lowering the cost and at the same time maintaining quality hence becomes an important objective in its own right.

Performing diagnostic tests on pools of serum samples instead of individual samples can sometimes lead to 80% cost reduction. The present publication shows that this is possible without loosing quality, if the pools are carefully designed. Under certain conditions it may even result in more precise estimates of disease frequency. Furthermore, it remains possible to link the risk of infection to important prognostic factors like age and to obtain a correct estimate of trends over time. How pooling for epidemiological studies can practically be implemented is clearly indicated. The methodology can also be adapted for individual diagnosis.

Application of this important methodological contribution should lead to substantial cost savings for surveillance. This will permit to control communicable diseases, including AIDS, more efficiently and to improve the evaluation of the results of our efforts towards this end.

report by Prof. Dr. Patrick Van der Stuyft, Unit of Epidemiology, Department of Public Health, Institute of Tropical Medicine, Antwerp, Belgium