gms | German Medical Science

6. Symposium Health Technology Assessment

Deutsche Agentur für HTA des DIMDI – DAHTA@DIMDI

03. bis 04.11.2005, Köln

The implications of metaanalytic techniques for Health Technology Assessment

Meeting Abstract

Search Medline for

  • author Stefan Sauerland - Medical Faculty at the University of Cologne, Biochemical and Experimental Department, Cologne, Germany

Deutsche Agentur für Health Technology Assessment des Deutschen Instituts für Medizinische Dokumentation und Information. 6. Symposium Health Technology Assessment. Köln, 03.-04.11.2005. Düsseldorf, Köln: German Medical Science; 2006. Doc05hta05

The electronic version of this article is the complete one and can be found online at: http://www.egms.de/en/meetings/hta2005/05hta05.shtml

Published: February 13, 2006

© 2006 Sauerland.
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

Abstract

An exhaustive review of the available scientific data is one of the cornerstones of health technology assessment (HTA). In this regard, HTA activities have close resemblance with the performance of systematic reviews. Metaanalysis is nothing more than a quantitative extension of a systematic review. If a high?quality systematic review or metaanalysis is available, it can ease HTA enormously. If no such review can be found, it is very informative to use metaanalytic techniques within the HTA report for three main reasons:

First, metaanalysis can be performed to obtain one summary estimate for relative efficacy, effectiveness or efficiency of a treatment. This is necessary, if the number of primary studies is large. Especially in those cases, where only many small and possibly underpowered trials have been conducted, it often requires metaanalysis to prove the superiority of a treatment statistically. The results of metaanalysis can be used for further analysis, such as economic modelling, formal decision analysis, and other simulation studies.

Secondly, it is possible to resolve contradictions between different primary studies by the use of metaanalysis. Every metaanalysis tries to group primary studies according to their methodological and medical characteristics. Such sensitivity analyses are necessary to examine to what extent the results of metaanalysis results are influenced by these characteristics. Differences between primary studies sometimes explain contradictory trial results. This allows HTA to identify important aspects of therapy closely correlated to therapeutic success. As many medical technologies are more or less efficacious in specific subgroups of patients, such sensitivity analyses are usually more valuable than the overall pooled result of a metaanalysis. Furthermore, it is possible to assess the influence of differences in study quality on overall conclusions.

Thirdly, the complex problems of publication bias can be investigated by metaanalytic techniques. As HTA has to rely on published data, the preferential publication of significant results can heavily distort overall results and conclusions. If a sufficiently large number of primary studies is available, metaanalysis will compare the results of smaller and larger studies (i.e. funnel plot). Any difference or asymmetry will cast a suspicion of publication bias on the nature of the published evidence.

In conclusion, many issues in HTA and metaanalyis complement each other. Close collaboration and exchange should be practised in order to achieve an optimum of analysis. Important HTA agencies therefore are closely collaborating with metaanalysts from the Cochrane Collaboration. Joint teaching programmes are completing the synergy between the two fields.


Notes

The complete lecture can be found on the website of DIMDI: http://www.dimdi.de/static/de/hta/symposien/2005/index.htm