A Comprehensive Strategy for the Interpretation of Quality-of-Life Data Based on Existing Methods


Patrick Marquis, Mapi Values, 15 Court Square, Suite 620, Boston, MA 02108. E-mail: Patrick.Marquis@Mapivaluesusa.com


Objectives:  Health-related quality of life (HRQL) instruments generally undergo a rigorous development and validation process. In contrast, methods for interpreting HRQL data are varied, and no comprehensive widely applicable procedure exists. Determining whether differences are statistically significant is the most common method, but this yields conclusions that may be difficult to understand in a clinical context or which may be of no practical value. Consequently, there is a need for a comprehensive interpretation strategy that gives results that are meaningful to a variety of audiences, including patients, clinicians, and decision-makers.

Methods:  The review of available interpretation strategies revealed that not all methods are applicable to all questionnaires, and some strategies may be difficult to implement for interpreting trial results. In addition, the issues decision-makers may have when assessing HRQL results have not really been addressed: what is measured and what is the meaning beyond statistical significance?

Results:  A comprehensive stepwise strategy, based on the most effective methods available, has been developed to address the key interpretation issues of decision-makers. It is structured around several steps: understanding the content of the questionnaire; evaluating the  magnitude of changes and their statistical significance; determining whether results are clinically significant, e.g., whether the observed changes crossed ranges of established threshold for meaningful differences; comparing pre- and post-treatment scores distribution with norms of references; and relating score changes to other outcomes end points such as morbidity, death, compliance, resource utilization, or productivity.

Conclusions:  The proposed strategy should help to structure and successfully address interpretation issues and thus make HRQL results more convincing.