CONTEXT

Prior to merging with Doctor on Demand*, Grand Rounds Health (GRH) was an employer-based, virtual healthcare company dedicated to their mission of “raising the standard of healthcare for everyone, everywhere.” Its primary offerings to members included expert second medical opinions, assistance with healthcare navigation, and facilitating virtual health visits with clinicians.

Creating a virtual primary care (VPC) experience was already on the long-term roadmap for GRH, but the COVID-19 pandemic—which necessitated an almost instantaneous, coordinated effort among clinicians and insurance companies to deliver virtual care of many kinds due to shelter-in-place—shifted many healthcare companies’ priorities and dramatically accelerated the drive towards increasingly virtual health care.

GRH considered primary care to be a mainstay of overall health. Research amply demonstrates that those who have a dedicated, longitudinal relationship with a primary care physician (PCP) have better health outcomes (Ganguli). However, primary care utilization in the United States is low—and trending lower.

Before actively designing a VPC experience, the design and product teams wanted to conduct exploratory research to address the following, foundational questions:

• What keeps people from engaging in primary care?
• How might virtual primary care address the primary care gap?

RESEARCH PROJECT DESIGN, METHODS, AND PARTICIPANTS

The main hypothesis going into the research was that the (United States domestic) rural population would be best served by, and more eager for, VPC, given their greater distance from clinicians and healthcare generally, and due to fewer options in rural areas for doctors who “are like/understand me,” with regard to ethnoracial identity, sexual orientation, gender identity, and other sociological and demographic statuses and signifiers. This hypothesis existed in theoretical contrast to an inverse assumption about the urban population: that closer geographic proximity to clinicians, and a greater diversity of clinicians to choose from, would render VPC less necessary or desirable.

Additionally, the rural population was considered, and the participant population selected, to be a rough proxy for the sizable number of blue-collar, working-class employees which GRH served; for example, industrial farm, retail, and warehouse workers. The urban population was selected to align with the higher-wage, white-collar GRH member population, such as Silicon Valley tech workers and middle- and upper-management employees in Fortune 500 companies.

We conducted remote, deep-dive interviews with the rural population—that is, relatively lower socioeconomic status (SES) and, with a majority living in rural and small towns, those who are most likely to have to travel considerable distances to a primary care provider, and/or for whom financial and geographic access may be problematic.

We conducted focus groups with the “urban” population: possessing a bachelor’s degree or higher (roughly half of the 36 focus group participants had graduate/advanced degrees), relatively higher SES, employer-provided health insurance, and urban/suburban residence. In short, for these participants, access to healthcare along multiple dimensions is not an issue.

Due to the logistical constraints of the pandemic we facilitated the focus groups and interviews online. This also enabled us to recruit nationally.

Across populations we discussed participants’ personal definitions and practices of “health,” their experiences with clinicians and institutionalized healthcare, and about their experiences with and hopes for virtual care.

Table 1. Research Design Overview

A note on terminology: we define virtual primary care to be a combination of in-person services (labs, bloodwork, physical examination), and technology-based services (virtual/video-based consultations, messaging, scheduling, insurance navigation, health records, and provider identification). We refer to this set of services as VPC; it is important to note that participants were not given and did not use this definition so that their intuitive associations and mental model of virtual primary care could be uncovered.

FINDINGS

Not only were our original hypotheses debunked, they were found to be insufficient, in that they did not account for any kind of distrust of or skepticism in primary care. The urbanites far more enthusiastically anticipated the shift to a more, if not entirely, virtual care experience. Our rural participants were comparatively more skeptical, if not wary, of primary care, generally, and the potential for delivering it through virtual means.

Additionally, we found that socioeconomic status is the “meta” status that impacts other demographic signifiers and identities—specifically gender and ethnoracial identity—that in turn impact one’s personal calculus of the factors that affect engagement with primary care.

Socioeconomic status, not geographic location, has the greater impact on definitions of health, engagement with healthcare, and anticipation of virtual primary care. Thus, our two populations are no longer referred to as “rural” and “urban,” but “low SES” and “high SES.”

The differences between the low-SES and high-SES populations primarily lie in deeply entrenched beliefs about, and prior experiences with, institutional health care—and to a lesser extent, technological milieu— with the low-SES population likely needing far more persuasion to engage in institutionalized healthcare, and with virtual primary care, in the first place. These participants often do not see institutionalized healthcare, much less primary care, as valuable or trustworthy. Reducing access barriers (distance, time, cost) will not in itself persuade them to engage with institutionalized health care. Their low levels of trust lead them to see virtual primary care as a diminished form of in-person care; thus, even though they recognize the convenience offered by virtual care, they are unable to imagine a desirable future involving it. In contrast, the high-SES population was much more optimistic, and the healthcare futures they anticipated did not, unlike those of the low-SES population, involve dread.