Impact of high-deductible health plans on delays in metastatic cancer diagnosis.

Abstract

6503 Background: High-deductible health plans (HDHPs) have grown rapidly in recent years, and now cover over one-half of U.S. workers. Patients in HDHPs are liable for the costs of all cancer-related care until their annual deductible is met, with the exception of screening tests such as colonoscopy and mammography. Due to increased out-of-pocket obligations, patients may postpone presenting for concerning symptoms or diagnostic testing, leading to delayed diagnosis. We therefore assessed the impacts of HDHPs on the timing of metastatic cancer detection. Methods: Using a nationally representative cohort of privately insured members in a national commercial and Medicare Advantage database (2003-2017), we studied 345,401 individuals age 18-64 years whose employers mandated a switch from a low-deductible (≤$500) plan to a high-deductible (≥$1,000) plan. Our control group consisted of 1,654,775 contemporaneous individuals whose employers offered only low-deductible plans. Both groups had a 1-year baseline period when all members were enrolled in low-deductible plans, and we followed members for a maximum of 13.5 years. Participants were matched with respect to age, gender, race/ethnicity, morbidity (ACG) score, poverty level, geographic region, employer size, baseline primary cancer, baseline medical and pharmacy costs, and follow-up duration. We used a validated claims-based algorithm to detect incident metastatic cancer diagnoses. We assessed time to metastatic cancer diagnosis in the baseline period (pre-HDHP switch) and follow-up period (post-HDHP switch) using a weighted Cox proportional hazards model. Results: After matching, there were no systematic differences between the HDHP and control groups with regard to observable baseline characteristics (standardized differences < 0.1). The mean age of participants was 42 years and the mean ACG score was 0.75. 49% were female, 48% lived in low-income neighborhoods, and 62% were White. We detected 1,668 metastatic events over a mean follow-up period of 38 months. There were no differences in time to metastatic diagnosis in the baseline year, prior to the HDHP switch (HR 0.96, p = 0.67). After employer-mandated HDHP switch, HDHP participants had lower odds of metastatic cancer diagnosis (HR 0.88, p = 0.01), indicative of delayed detection relative to the control group. Conclusions: Compared with conventional health plans, HDHPs are associated with delayed detection of metastatic cancer. These findings imply that patients postpone seeking care for concerning symptoms or defer diagnostic testing when exposed to high cost-sharing. Given recent advances that have improved survival of patients with advanced-stage cancers, future research efforts should investigate the impacts of HDHPs on quality of life, engagement in palliative care, and use of treatments in this patient population.