Abstract
IMPORTANCE: High-deductible health plans (HDHPs) are a common insurance mechanism. OBJECTIVE: To evaluate whether HDHPs are meeting the medical needs of persons with chronical illness. DESIGN, SETTING, AND PARTICIPANTS: This longitudinal cohort study used 2016 to 2019 national administrative and claims data from MarketScan. Cohort members were adults aged 18 to 64 years with asthma, diabetes, hypertension, coronary artery disease, heart failure, or major depressive disorder. The treatment group was persons newly enrolling in an HDHP; the control group was persons continuously enrolled in non-HDHPs. Data were analyzed from October 2022 to April 2024, with revisions conducted between December and January 2025. EXPOSURE: New enrollment in an HDHP was instrumented by firms changing the plan choice options to promote HDHP enrollment (ie, restricted-choice firms). Analyses used difference-in-differences models combined with this instrumental variable. Entropy balancing was used to address residual confounding. MAIN OUTCOMES AND MEASURES: Use of annual recommended medical care, which was abstracted from disease-specific evidence-based clinical practice guidelines and included clinic visits, prescription drugs, laboratory tests, and an overall measure that combined all 3. Results were pooled across disease type to improve policy relevance of findings; disease-specific analyses were also conducted. Hypotheses were formulated prior to data collection. RESULTS: The cohort consisted of 343 137 adults (182 532 [53.20%] female; 149 760 [43.64%] aged 55-64 years [before entropy balancing]). Groups exhibited covariate balance after entropy balancing. Restricted-choice enrollment into an HDHP was associated with reduced use of recommended medical care, with persons in HDHPs reducing their use of recommended clinic visits by 3.1 (95% CI, -4.9 to -1.2) percentage points (P < .001), their use of recommended prescription drugs by 9.0 (95% CI, -11.8 to -6.2) percentage points (P < .001), and their use of recommended annual laboratory testing by 5.7 (95% CI, -8.2 to -3.2) percentage points (P < .001). Overall, HDHP enrollees were 4.7 (95% CI, -6.2 to -3.3) percentage points less likely to receive recommended medical care compared with non-HDHP enrollees (P < .001). CONCLUSIONS AND RELEVANCE: This longitudinal cohort study of 343 137 adults with chronic illness found HDHP enrollment was associated with reduced receipt of recommended medical care across a variety of conditions. These results have important implications for recently proposed federal legislation that proposes to exempt chronic illness management from HDHP deductibles.