The role of race and ethnicity in health care crowdfunding: an exploratory analysis

Abstract

Medical crowdfunding is a key source of financing for individuals facing high out-of-pocket costs, including organ-transplant candidates. However, little is known about racial disparities in campaigning activity and outcomes, or how these relate to access to care. In this exploratory, nationwide, cross-sectional study, we examined racial disparities in campaigning activity across states and the association between US campaigners' race and ethnicity and crowdfunding outcomes using a novel database of organ-transplant-related campaigns, and an algorithm to identify race and ethnicity based on name and geographic location. This analysis suggests that there are racial disparities in individuals' ability to successfully raise requested funds, with Black and Hispanic campaigners fundraising lower amounts and less likely to achieve their monetary goals. We also found that crowdfunding among White, Black, and Hispanic populations exhibits different patterns of activity at the state level, and in relation to race-specific uninsurance and waitlist additions, highlighting potential differences in fundraising need across the 3 groups. Policy efforts should consider not only how inequalities in fundraising ability for associated costs influence accessibility to care but also how to identify clinical need among minorities.

Keywords: access to care; medical crowdfunding; organ transplantation; racial disparities.

Figure 1.

Campaigns and WLAs of solid organs (including heart, lungs, kidney, liver) per 100 000 race and ethnicity population. Includes linear fit lines per race and ethnicity, as well as the estimated slope and respective confidence intervals. To allow for sample consistency with Figure 2, excludes state-year pairs for which the uninsurance rate is not available due to small population size, according to KFF state data criteria. Abbreviations: CI, confidence interval; KFF, Kaiser Family Foundation; WLA, waiting-list addition.

Figure 2.

Campaigns and uninsurance rate per 100 000 race and ethnicity population. Includes linear fit lines per race and ethnicity, as well as the estimated slope and respective confidence intervals. Excludes state-year pairs for which the uninsurance rate is not available due to small population size, according to KFF state data criteria. Abbreviations: CI, confidence interval; KFF, Kaiser Family Foundation.

Figure 3.

Estimates of race- and ethnicity-adjusted IRR for each crowdfunding outcome. Amount raised and average donation models were estimated using negative binomial regressions, and success rate using Poisson regression. Estimated race-specific adjusted IRRs are shown with 95% confidence intervals, with White campaign organizer as the baseline. All adjusted models include (1) dummy variables for each year of creation (baseline: 2015); (2) campaign characteristics (social media shares [including a quadratic term], a dummy variable for a campaign organizer different from the beneficiary, a dummy variable for high fraud score, and dummy variable for solid organs); and (3) state characteristics (a dummy variable for fourth quartile of race-specific WLA per 100 000 and a dummy variable for fourth quartile of race-specific uninsurance rate). Abbreviations: CI, confidence interval; IRR, incidence rate ratio; WLA, waiting-list addition.