Proxy adverse health outcomes comprising high-severity emergency department visit expenditures and high-severity hospitalization days (and low-severity outcomes for comparison), overall and among key HDHP subgroups, 1 year before and after an HDHP switch compared with contemporaneous control group members
| . | Cumulative annual rate at the end of1 . | Change in HDHP vs. control group, end of follow-up vs. end of baseline1 . | ||||
|---|---|---|---|---|---|---|
| HDHP group . | Control group . | Absolute Estimate (95% CI) . | Relative, % Estimate (95% CI) . | |||
| Baseline . | Follow-up . | Baseline . | Follow-up . | |||
| Overall (n = 23,493 HDHP and 192,842 control) | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 185.0 | 233.6 | 185.0 | 233.6 | ND | ND |
| Low severity2 | 170.3 | 213.0 | 165.5 | 212.3 | −10.6 (−29.3, 8.0) | −4.8 (−12.8, 3.3) |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 64.9 | 89.0 | 62.7 | 85.8 | 3.2 (−1.1, 7.4) | 3.7 (−1.4, 8.8) |
| Low severity2 | 41.3 | 55.3 | 41.3 | 53.7 | 1.6 (−1.6, 4.7) | 2.9 (−3.1, 8.9) |
| Low income (n = 8,453 HDHP and 65,468 control)3 | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 143.1 | 262.7 | 148.1 | 228.4 | 50.0 (40.6, 59.4) | 23.5 (18.3, 28.7) |
| Low severity2 | 158.2 | 230.3 | 166.8 | 243.6 | −13.3 (−19.5, −7.1) | −5.5 (−8.0, −3.0) |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 48.8 | 103.3 | 50.7 | 85.7 | 22.2 (14.3, 30.1) | 27.4 (15.5, 39.2) |
| Low severity2 | 38.4 | 62.6 | 40.8 | 58.7 | 3.9 (0.0, 7.7) | 6.6 (−0.2, 13.4) |
| High income (n = 14,841 HDHP and 124,479 control)4 | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 181.7 | 217.3 | 199.1 | 218.6 | 16.1 (1.3, 30.8) | 8.0 |
| Low severity2 | 162.5 | 199.8 | 148.8 | 186.2 | ND | ND |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 64.8 | 78.1 | 58.4 | 81.0 | −9.4 (−17.0, −1.7) | −10.7 (−18.8, −2.6) |
| Low severity2 | 39.9 | 51.4 | 37.8 | 45.5 | 3.8 (−3.8, 11.4) | 8.0 (−8.8, 24.8) |
| . | Cumulative annual rate at the end of1 . | Change in HDHP vs. control group, end of follow-up vs. end of baseline1 . | ||||
|---|---|---|---|---|---|---|
| HDHP group . | Control group . | Absolute Estimate (95% CI) . | Relative, % Estimate (95% CI) . | |||
| Baseline . | Follow-up . | Baseline . | Follow-up . | |||
| Overall (n = 23,493 HDHP and 192,842 control) | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 185.0 | 233.6 | 185.0 | 233.6 | ND | ND |
| Low severity2 | 170.3 | 213.0 | 165.5 | 212.3 | −10.6 (−29.3, 8.0) | −4.8 (−12.8, 3.3) |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 64.9 | 89.0 | 62.7 | 85.8 | 3.2 (−1.1, 7.4) | 3.7 (−1.4, 8.8) |
| Low severity2 | 41.3 | 55.3 | 41.3 | 53.7 | 1.6 (−1.6, 4.7) | 2.9 (−3.1, 8.9) |
| Low income (n = 8,453 HDHP and 65,468 control)3 | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 143.1 | 262.7 | 148.1 | 228.4 | 50.0 (40.6, 59.4) | 23.5 (18.3, 28.7) |
| Low severity2 | 158.2 | 230.3 | 166.8 | 243.6 | −13.3 (−19.5, −7.1) | −5.5 (−8.0, −3.0) |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 48.8 | 103.3 | 50.7 | 85.7 | 22.2 (14.3, 30.1) | 27.4 (15.5, 39.2) |
| Low severity2 | 38.4 | 62.6 | 40.8 | 58.7 | 3.9 (0.0, 7.7) | 6.6 (−0.2, 13.4) |
| High income (n = 14,841 HDHP and 124,479 control)4 | ||||||
| ED visit expenditures, $ per member | ||||||
| High severity2 | 181.7 | 217.3 | 199.1 | 218.6 | 16.1 (1.3, 30.8) | 8.0 |
| Low severity2 | 162.5 | 199.8 | 148.8 | 186.2 | ND | ND |
| Hospitalization days, per 1,000 members | ||||||
| High severity2 | 64.8 | 78.1 | 58.4 | 81.0 | −9.4 (−17.0, −1.7) | −10.7 (−18.8, −2.6) |
| Low severity2 | 39.9 | 51.4 | 37.8 | 45.5 | 3.8 (−3.8, 11.4) | 8.0 (−8.8, 24.8) |
ED, emergency department; ND, not detected.
Bold values indicate statistical significance.
1All rates and changes account for differing baseline trends between HDHP and control group members and are estimated with marginal effects methods using parameters from aggregate-level segmented regression analysis of cumulative interrupted time–series data that were adjusted for age, sex, race/ethnicity, education level, poverty level, U.S. region, ACG score, employer size, and calendar month of the index date.
2See research design and methods for definition of low- and high-severity emergency department visit expenditures and hospitalization days.
3Living in a neighborhood with below-poverty levels of ≥10%.
4Living in neighborhoods with below-poverty levels of <10%.