77 Patient cost-sharing, mental health care and inequalities care use in high deductible years was higher before reaching the exemption threshold and lower when paying the full deductible, compared to low deductible years. Observations in the range 217 to 227 months correspond to individuals partially exposed to the deductible in the year they turn 18. Appendix F, Figure A4 shows these discontinuities by year. Figure 1. Trends in mental health care use by age in months for the low and high deductible periods Notes: Each dot corresponds to a cohort of individuals born in January, February, March, etc. and represents the average of mental health care use (binary indicator) for that cohort, estimated from the raw data. The figure is obtained from the population born between 1988 and 1999, observed for the years of the 15th to 21st birthday (2,862,771 female and 2,973,588 male individuals). Depending on their birth cohort individuals might be represented in 1 to 6 of the dots (Appendix C). Quadratic age polynomials for female and linear age-trends for male. Only observations 204-216 and 228-239 (grey shaded areas) are used in the main analysis. Year-by-year age trends are available in Appendix F, Figure A4. Table 3 provides estimates for our baseline model (1). We find that a deductible increase of about 180 euros between the low and high deductible periods reduced mental health care use by 1.1 pp (-1.1, CI 95%: -1.8, -0.4) for females, corresponding to -13.6% (CI 95%: -22.1%, -5.2%) mental health care use. We find a small and negative but imprecisely estimated coefficient for males (-0.3, CI 95%: -0.7, 0.1), corresponding to -5.3% (CI 95%: -11.8%, 1.2%) in relative terms (panel A of Table 3). Our subgroup analysis by socioeconomic status shows that the impact is concentrated among lower income females, with -18.9% (CI 95%: -35.4%, -2.3%) use for the lowest income quartile and -21.3% (CI 95%: -36.7%, -5.9%) for the second lowest quartile, compared to smaller and imprecise negative coefficients for higher quartiles (panel B of Table 3). Treatment cessation among mental health care users at 17 The subgroup analysis of treatment cessation focuses on the subpopulation of adolescent users instead of the overall population. Both genders were more likely to stop treatment in the high deductible period (females: -19.9%, CI 95%: -28.5%, -11.3%; males:-26.5%, CI 95%: 3
RkJQdWJsaXNoZXIy MjY0ODMw