Table 3. Changes in health expenditure based on difference-in-differences within the event-study framework.
| Year | Total logarithm of health expenditures (95% CIa) | P value | Inpatient logarithm of health expenditures (95% CI) | P value | Outpatient logarithm of health expenditures (95% CI) | P value |
|---|---|---|---|---|---|---|
| 2017 | 0.0055 (−0.002 to 0.013) | .15 | 0.003 (−0.005 to 0.01) | .48 | 0.008 (−0.003 to 0.018) | .14 |
| 2018 | 0.003 (−0.002 to 0.007) | .22 | 0.001 (−0.004 to 0.007) | .59 | 0.0029 (−0.003 to 0.009) | .37 |
| 2019 | Reference | Reference | Reference | Reference | Reference | Reference |
| 2020 | −0.01 (−0.018 to -0.003) | .006 | −0.011 (−0.019 to -0.003) | .005 | −0.008 (−0.018 to 0.001) | .08 |
| 2021 | −0.006 (−0.019 to 0.0075) | .40 | −0.009 (−0.021 to 0.004) | .19 | −0.002 (−0.017 to 0.0127) | .78 |
| 2022 | −0.008 (−0.021 to 0.005) | .23 | −0.009 (−0.02 to 0.003) | .14 | −0.005 (−0.019 to 0.0097) | .51 |
CIs are based on the SEs that adjust for clustering at the prefecture level.