Measuring time use in health economics: a methodological perspective

Presenter: Catherine Pitt, London School of Hygiene & Tropical Medicine

Abstract

Background: Time use studies are employed in many fields, from consumer research and organizational productivity analysis, to gender equity and sociology. Within health economics, time use studies are primarily used to value time inputs in economic evaluations, to measure human resource requirements or excess capacity, and to conduct in-depth analyses aimed at improving staff efficiency and effectiveness. Time and motion studies are widely regarded as the “gold standard” in time use measurement, however, they are highly resource-intensive, prone to observer bias, and therefore sub-optimal in some study contexts. Other study designs have been employed and validated to varying degrees; they include activity sampling (random and systematic), self evaluations with time sheets or diaries, retrospective interviews and questionnaires, and various hybrids of these.

Objective: To review and assess the key methodological issues in the measurement of individuals’ time use, with particular emphasis on provider and patient/caregiver time use studies in economic evaluations.

Methods: The analysis will draw on evidence from the literature and on recent experiences with time use studies in three randomized controlled trials in West Africa.

Results: The presentation will define circumstances in which time use studies are (and are not) necessary; describe a number of possible study designs, weighing the advantages and disadvantages of each, and providing empirical examples; and consider strategies for maximizing data quality. The significance of the personal characteristics and incentives of subjects including employees, volunteers, patients and caretakers will be discussed, as will the significance of time use categories. When combined with an understanding of the resources available for a given study, these issues can help to elucidate the feasibility and likely sources of bias of potential study designs. Emphasis will also be placed on the precision and accuracy required for time measurements, and the degree of imprecision surrounding unit prices, which may be especially high when shadow prices are required.

Discussion: The choice of study design will depend on the levels of precision and accuracy required, the likely sources of bias, and feasibility and other practical concerns. High levels of accuracy cannot, however, be achieved without due attention to the research implementation process, which can have as great or greater an impact on the accuracy of findings as the choice of study design.