A User-Centered Design Approach to Information Sharing for Older Patients and Their Families

Abstract

The development of technology that drives and enables engagementwithpatients and their families is accelerating at an unprecedentedpace.Demand isdrivenbypatients seeking the best quality and experience as well as by health care organizations, payers, and policy makers in response to health care costs, especially for the older US population. Per-person personal health care spending in 2013 for individuals 65 years or olderwas 5 times higher than spending per child and 3 times higher than spending per working-age person in 2010.1 With nearly 60% of seniors reportinguseof the Internet in 2014 (up from just 35% in 2008),2 the importance of digital health care technology for the elderly population is paramount. This importance is appreciated by health care technologists fromSiliconValley entrepreneurs to giantmultinational companies.Withmore than$4.1 billion invested inhealth care technologies in2014,3 theappetite to innovateanddisrupt, and the anticipated payoff for doing so, seems virtually unbounded. In addition, government incentives, such as Centers for Medicare & Medicaid Services meaningful use rules, are spurring the use of patient portals for engagement with Medicare beneficiaries through functions such as the ability to view online, download, and transmit their health information. With such fuel and burning demand, yet little systematic evidence linking patient portals with outcomes and satisfaction,4 how can we know how to best serve the digital health care technology needs of elderly patients? In this issue of JAMA Internal Medicine, Crotty and colleagues5 explore the information sharing preferences of 30 elderlypatients and23of their family caregivers. Throughprofessionally moderated focus groups, information sharing was covered broadly, although the study necessarily contemplates the implications for software design for patient portals. Through their focus groups, Crotty and colleagues identified 7 lessons: (1) elders are concerned about placing the burden of their illnesses and problems, including through the use ofhealth informationandinformationtechnology,ontheircaregiverswhile caregiversare looking to thesame informationand technology as potential mechanisms for relief of these burdens; (2) elders have privacy concerns about sensitivemedical and relatedhistorical informationas itmaybeexposed to caregivers; (3)eldersappreciateelementsof their independenceand are concerned about information andmodern devices unnecessarilycompromising that independence; (4)elderswish to retain control of decisionmaking as long as possible; (5) transfer of this control from elders to caregivers occurs gradually depending on the elder’s health and functional status; (6) elders andcaregiversagreethat informationsharinganddecisionmaking is fluid,dynamic,anddependentonthecurrentcontext;and (7) preferences related to sharing, use, and control of personal health information varied by individual, situation, and elder/ caregiver relationships. These are important findings, not adequately represented in the design of current health information technology, andarguably they represent a call to action.With somuch investment in health information technology and the widespreaduseofportalsbyhealthcareprofessionals,patients, and their families, why haven’t we advanced further? The answer is at least partially evident in the findings from Crotty and colleagues’ focus groups. Specifically, we have acknowledged but failed to act on the fluidity and variability in preferences across individuals, situations, timing, relationships, and health context. What is needed to build on Crotty and colleagues’ research, in the approach to patient portals and in health information technology more broadly, is a response to this variability inpreferences throughestablished techniques, such as Lean UX6 (user experience) and user-centered design. Now foundational to modern consumer software and product design, several pillars of Lean UX are described by Gothelf6 as: • Design thinking—“innovation powered by...direct observationofwhatpeoplewantandneed in their livesandwhat they like or dislike about the way particular products are made, packaged, marketed, sold, and supported” (quoted from Brown7). • Agile softwaredevelopment—workingsoftware in favorofcomprehensivedocumentation; responding to change in favor of following a plan.6 • Lean startup method—the creation of rapid prototypes designed to test assumptions, and the use of customer feedback to evolve these prototypes quickly.6 Crotty andcolleagues’workhas aptlypointedout that one size does not fit all with regard to health information sharing andportal design for elders and caregivers. To address this issue,wemustobserve these individuals in theappropriate context, as they interact with health care professionals and each other, andespecially as they interactwith the informationand systems that we seek to develop to meet their needs. These systems should not be fully developed solutions. Indeed, the value is in the rapid exploration of user interactionwith early prototypes and in an agile response to the findings. A crucial step in the user-centered design process is the definition and refinement of multiple archetypes or personas, each of which represents a “fictional and super-typical characterization of a user created to represent a user group.”8(e255) Personas allow the design process to acknowledge the varied characteristics, preferences, contexts, and needs of individual users through their aggregation into clusters with common contextual elements. An output of several Related article page 1492 Research Original Investigation Information Sharing Preferences of Older Patients and Their Families