Health outcomes assessment in vulnerable populations: measurement challenges and recommendations

With growing recognition that some population subgroups are particularly vulnerable to receiving suboptimal health care and achieving poor health outcomes, innovative techniques are required for collecting and evaluating health outcomes data. Research is also needed to better understand the causal pathways linking vulnerability with health outcomes. This article focuses on patients with a chronic illness (cancer) who also have low literacy and/or poor English language skills. We summarize the association among literacy, language, ethnicity, and health outcomes; describe innovative technologies to enhance communication; and discuss the advantages of using psychometric measurement models in health outcomes assessment. Results from our ongoing research projects are presented, including the development of an audiovisual computer-based testing platform for self-administration of questionnaires. Such innovative multimedia technologies allow patients with limited or even no reading ability to participate in outcomes assessment and have the potential to be incorporated into a clinical setting with minimal burden on staff and patients. Appropriate methods are also needed to evaluate measurement equivalence across diverse patient groups, that is, the extent to which items in a questionnaire perform similarly across groups. Item response theory measurement models provide a strategy for differentiating between measurement bias and real differences that may exist between groups. Recommendations for clinical practice and research are offered specifically to address medically underserved and vulnerable populations.