Estimates from multivariable linear probability model controlling for the effects of age, sex, and race/ethnicity in year 1 and time-varying measures of household income as a percentage of the federal poverty line, self-reported health, and number of chronic conditions (coronary heart disease, angina, myocardial infarction, other heart disease, stroke, emphysema, high cholesterol, diabetes, arthritis, and asthma). Increase in unmet need for those losing health insurance in Year 2 was significantly greater than for the continuously insured.
Estimates from multivariable linear probability model controlling for the effects of age, sex, and race/ethnicity in year 1 and time-varying measures of household income as a percentage of the federal poverty line, self-reported health, and number chronic conditions (coronary heart disease, angina, myocardial infarction, other heart disease, stroke, emphysema, high cholesterol, diabetes, arthritis, and asthma). Decrease in unmet need for those gaining health insurance in Year 2 was significantly greater than for the continuously uninsured.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Yabroff KR, Kirby J, Zodet M. Association of Insurance Gains and Losses With Access to Prescription Drugs. JAMA Intern Med. 2017;177(10):1531–1532. doi:10.1001/jamainternmed.2017.4011
Prescription drugs can effectively treat many diseases, improving quality of life, life expectancy, and population health. However, prescription drug spending has been rising rapidly in the United States1 resulting in concerns about affordability and patient access. Health insurance is strongly associated with prescription drug access in cross-sectional studies,2,3 but estimates may partly reflect differences between individuals with and without insurance, rather than effects of insurance coverage. To address this limitation, we used longitudinal data from the nationally representative Medical Expenditure Panel Survey (MEPS) to assess the effects of insurance gains and losses on prescription drug access.
Using longitudinal MEPS data spanning from 2008 to 2014, we categorized adults aged 18 to 64 years by insurance coverage during 2-year panels: (1) continuously insured (n = 38 231); (2) insured year 1 and lost coverage for at least 6 months in year 2 (n = 1320); (3) continuously uninsured (n = 13 516); and (4) uninsured year 1 and gained coverage for at least 6 months in year 2 (n = 1619). Unmet need for prescription drugs was measured in year 1 and year 2 based on responses to questions about delays or inability to obtain needed prescription drugs. We compared unmet need in years 1 and 2 across the 4 insurance categories and controlled for patient characteristics associated with insurance coverage and medical need (ie, age, sex, race and ethnicity, and time-varying measures of self-reported health, number of chronic conditions, and income as a percentage of federal poverty line) with a multivariable linear probability model using STATA statistical software (version 14, STATA corp). The study was covered under an institutional review board agreement designed for the Medical Expenditure Panel Survey with regard to approval and patient written informed consent.
Among individuals with continuous coverage, the percent with unmet need for prescription drugs was low in years 1 and 2 (3.2% and 3.3%, respectively). In contrast, among individuals who had coverage in year 1 but lost it in year 2, the percent with unmet need more than doubled, from 3.1% to 6.6% (difference, 3.5%; 95% CI, 1.3 to 6.1; P < .05). Adjusted estimates were similar (Figure 1) and increases in unmet need for those losing insurance were significantly greater than for the continuously insured. Among individuals continuously uninsured, unmet need was similar in year 1 and year 2 (6.2% and 5.5%, respectively). However, initially uninsured individuals who gained coverage in year 2 had a 3.4 percentage point decline in unmet need (8.8% to 5.4%; 95% CI, −5.9 to −1.3; P < .05). Adjusted estimates were similar (Figure 2) and declines in unmet need for those gaining insurance were significantly greater than for the continuously uninsured. Findings were robust in sensitivity analyses of duration of insurance loss or gain and all combinations of time-varying self-reported health and individual chronic conditions.
Our findings that unmet need for prescription drugs declined among initially uninsured adults who gained coverage and doubled among initially insured adults who lost coverage provide longitudinal evidence that having and maintaining health insurance is a key protection against unmet need for prescription drugs in a nationally representative sample. However, having insurance does not guarantee coverage completeness or access to care. Patient cost sharing is increasing through higher deductibles, copayments, and coinsurance rates4 and medical financial hardship is increasingly documented in the United States,5 especially in relation to prescription drug use.
Although findings were robust in sensitivity analyses, the study was limited by self-reported measures and lack of cost-sharing information. Self-reported unmet need may not correspond exactly to objective clinical measures.
Prescription drug spending is projected to continue rising,1 increasing fiscal pressures on commercial, federal, state, and family budgets. For individuals with high drug costs, these trends may erode some of the protective effect of insurance coverage documented in this study. It is therefore imperative that research continue to monitor the relationship between insurance coverage and unmet need, assess spending and clinical outcomes, and that survey, administrative, and clinical data be available to do so.
Corresponding Author: K. Robin Yabroff, PhD, MBA, Division of Health Care Financing Policy, Office of Health Policy, Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, 200 Independence Ave, SW, Washington, DC 20201 (email@example.com).
Accepted for Publication: June 27, 2017.
Published Online: September 11, 2017. doi:10.1001/jamainternmed.2017.4011
Author Contributions: Dr Kirby had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: Yabroff, Kirby.
Drafting of the manuscript: Yabroff, Kirby.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: all authors.
Conflict of Interest Disclosures: None reported.
Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Agency for Healthcare Research and Quality or the Department of Health and Human Services.