Effectiveness of Behaviorally Informed Letters on Health Insurance Marketplace Enrollment: A Randomized Clinical Trial

Abstract

Importance: Every year during the open enrollment period, hundreds of thousands of individuals across the Affordable Care Act marketplaces begin the enrollment process but fail to complete it, thereby resulting in coverage gaps or going uninsured.

Objective: To investigate if low-cost ($0.55 per person) letters can increase health insurance enrollment.

Design setting and participants: This intent-to-treat randomized clinical trial was conducted during the final 2 weeks of the 2015 open enrollment period among the 37 states on the HealthCare.gov platform. The trial targeted 744 510 individuals who started the enrollment process but had yet to complete it. Data were analyzed from January through August 2021.

Interventions: Study participants were randomized to either a no-letter control group or to 1 of 8 letter variants that drew on evidence from the behavioral sciences about what motivates individuals to take action.

Main outcomes and measures: The primary outcome was the health insurance enrollment rate at the end of the open enrollment period.

Results: Of the 744 510 individuals (mean [SD] age, 41.9 [19.6] years; 53.9% women), 136 122 (18.3%) were in the control group and 608 388 (81.7%) were in the treatment group. Most lived in Medicaid nonexpansion states (72.7%), and a plurality were between 30 and 50 years old (41.0%). For race and ethnicity, 3.0% self-identified as Asian, 14.0% as Black, 5.1% as Hispanic, 39.8% as non-Hispanic White, and 38.2% as other or unknown. By the end of the open enrollment period, 4.0% of the control group enrolled in health insurance coverage. Comparatively, the enrollment rate in the pooled treatment group was 4.3%, which demonstrated an increase of 0.3 percentage points (95% CI, 0.2-0.4 percentage points; P<.001), yielding 1753 marginal enrollments. Letters that used action language caused larger enrollment effects, particularly among Black individuals (increase of 1.6 percentage points; 95% CI, 0.6-2.7 percentage points; P = .003) and Hispanic individuals (increase of 1.5 percentage points; 95% CI, 0.0-3.0 percentage points; P = .046) in Medicaid expansion states.

Conclusions and relevance: This randomized clinical trial shows that letters designed with best practices from the behavioral sciences literature were a low-cost way to increase health insurance enrollment in the Affordable Care Act marketplaces. More research is needed to understand what messages are most effective amid the recently passed American Rescue Plan.

Trial registration: ClinicalTrials.gov Identifier: NCT05010395.

Figure 1.. CONSORT Flow Diagram of Included Individuals

Figure 2.. Effect of Letter on Affordable Care Act Enrollment Rate Pooled, by Arm, and Cost per Enrollee

Each point represents the average effect in percentage points. Error bars denote 95% CIs.

Figure 3.. Effect of Action Letters on Affordable Care Act Enrollment Rate by Race and Ethnicity and States’ Medicaid Expansion Status

Each row represents the average effect of a letter in percentage points. Error bars denote 95% CIs.