Optimal recall periods for patient-reported outcomes: challenges and potential solutions

ABSTRACT

Objectives: As the role and importance of patient-reported outcomes (PROs) increase, the validity and reliability of PRO measures come under greater scientific and regulatory scrutiny. One key issue is selecting the ‘most appropriate’ recall period for capturing PROs in clinical trials. This paper draws on survey research, health-specific literature, and results from clinical trials to summarize factors that can influence recall and provide guidance on selecting an optimal recall period.

Methods: We conducted a systematic review of six databases and additional literature drawn from bibliographies of the selected articles.

Results: Six major factors can influence recall; these can be classified into two broad areas: characteristics of the recalled phenomenon (recency, attributes, complexity) and context or meaning of the recalled phenomenon (salience, patient experience, mood). Results of different recall periods for three classes of PROs are presented: health behaviors, symptoms, and health-related quality of life. We present findings on the effect of alternative recall periods for three commonly used PROs. Finally, we propose a heuristic model to link the concept under investigation with an optimal recall period.

Conclusions: No single recall period is best for all measures or all phenomena. The recall period must correspond to the characteristics of the phenomenon of interest and the purpose of the assessment. Recall period is an issue of internal validity. An incorrect recall period introduces measurement error that may reduce the chances of detecting a treatment effect. Researchers should consider recall period as seriously as they do other measurement properties.

Key words: :

• Measurement
• Patient-reported outcomes
• Recall bias
• Recall period

Acknowledgments

Declaration of interest: Funding for this research was provided by AstraZeneca. B. P. is an employee of AstraZeneca. D. E. S. and N. K. L. are employees of United BioSource Corporation (UBC), which provides consulting and other research services to pharmaceutical, device, government, and non-government organizations. In this salaried position, N. K. L. and D. E. S. work with a variety of companies and organizations. They receive no payment or honoraria directly from these organizations for services rendered. O. C. is an employee of Hôpital Saint Louis and University Denis-Diderot, Paris, France. He received no compensation for his participation in this research.

The authors would like to thank Meghan Werner and Laurie Smith of UBC for their assistance searching and organizing the literature, Laurie Burke of the US Food and Drug Administration for her helpful comments on an earlier draft of this paper, and two anonymous reviewers for excellent suggestions and comments that strengthened this paper.