Patient-Reported Outcomes

10. Applications

Decision makers and analysts wanting to measure PROs should first identify the problem or application of the measure. With this information, one can then identify the desired characteristics of existing measures to be included in the assessment.

For example, monitoring the health of populations and communities demands parsimonious instruments including global evaluations across a number of conditions and different population groups. For comprehensive evaluation in a clinical trial, health profiles or batteries are most appropriate according to the main effects intended and unintended or adverse consequences of treatment. For economic evaluation, utility measures are useful to produce a comparison across alternative treatments.

PROs from Children

PROs directly from children and youth are taking greater prominence among all interested parties, following similar development as that for adults and older adults (Starfield, 1996). Children and youth represent a special case, however, given knowledge of variation in how and when children develop, the wide variation in willingness and ability to self-report across the age spectrum, and the “special” language of children and adolescents in different cultures of the world. Rapid progress is being made in Europe and North America, to be followed and informed by work in other parts of the world, often less accessible to Western parties.

Cross Cultural PROs

Demand is also increasing for quality of life measures available for use in cross-cultural comparisons, which requires special attention to cultural adaptation and validation for each culture in which the measure is applied (Acquadro et al., 2003).

The most desirable means of development and validation is to have the goal of cross-cultural comparability in mind from the beginning.

Measures developed simultaneously in different cultures have the advantage of identifying as early as possible those domains and items that are more or less valid in a particular culture or population. Translating instruments developed in one culture for use in another is more common, but the danger of this approach is the assumption that the conceptual structure, domains, and items are cross-cultural.

Recognizing Cultural Differences

Assessments of functional status that use examples such as “ability to walk several blocks” run the danger that “blocks” have different if any meaning at all in different cultures. Response scale anchors, such as “quite a bit”, also do not translate easily into different languages.

The psychometric properties of cross-cultural measures is similar to that for instruments used within one language or cultural group, although standards for aggregation across sites have not been rigorously applied. When it is and is not valid to use measures in different populations and to pool data across different cultures remains an area for further investigation and debate.