Healthcare use among persons with visual impairment: Difference-in-difference estimations.

BACKGROUND Studies on healthcare service use among persons with visual impairments in countries with different healthcare systems are needed for drawing global inferences. OBJECTIVE The goal was to assess the impact of visual impairment on healthcare service use. METHODS A retrospective cohort study on health insurance claims data from the National Health Insurance System (NHIS) was conducted. All data from 2002 to 2013 for 2206 and 8824 persons with and without visual impairments, respectively, were extracted and aggregated monthly, generating a total of 162,876 and 568,459 person-month observations, respectively. The dependent variable was total expenditures. Difference-in-difference estimations based on a multivariate log-normal random-effect regression were employed. RESULTS Persons with visual impairments spent 5.7% and 6.8% more on total healthcare expenditures per month and outpatient healthcare expenditures, respectively, than those without visual impairments, after adjusting for pre-impairment differences. Further results revealed an upward trend in expenditures, particularly among those with visual impairments; this increase was the highest during the year prior impairment (approximately 32% and 21% more for total and outpatient services, respectively, compared with baseline). Inpatient healthcare expenditures showed a similar increase (by approximately 26%) during the year prior impairment compared with baseline; however, the magnitude dropped to approximately 14% and 6% during the 1st and 2nd years of impairment, respectively. Conversely, the annual pattern of expenditure increase was parallel over time for persons without visual impairments, regardless of service. CONCLUSION Our findings suggest a need for health management and effective care use, particularly the year prior to visual impairment onset.