DID | Heterogenous DID | Using treatment only and control only patients | IPTW | |
Model 1 β(SE) | Model 2 β(SE) | Model 3 β(SE) | Model 4 β(SE) | |
Month Dummies | Yes | Yes | Yes | Yes |
Disease Dummies | Yes | Yes | Yes | Yes |
Patient Fixed Effect | Yes | Yes | Yes | Yes |
Post | 0.0235 (0.003)† | 0.0235 (0.003)† | 0.0238 (0.004)† | 0.0236 (0.004)† |
Treat | −0.3722 (0.277) | −0.3738 (0.276) | –- | −0.4257 (0.305) |
Post×Treat | −0.0299 (0.004)† | −0.0267 (0.004)† | −0.0270 (0.004)† | −0.0268 (0.004)† |
Post×Treat×Male | – | −0.0025 (0.001)* | −0.0026 (0.001)* | −0.0025 (0.001)* |
Post×Treat×Low-Income | –- | 0.0037 (0.002)* | 0.0035 (0.002)* | 0.0039 (0.002)* |
Post×Treat×High-Income | –- | −0.0005 (0.001) | −0.0005 (0.001) | 0.0005 (0.001) |
Post×Treat×Seoul Metro. Area | –- | −0.0052 (0.001)†† | −0.0052 (0.001)† | −0.0052 (0.001)† |
R-square | 0.006 | 0.006 | 0.006 | 0.006 |
Number of observations | 714 801 | 714 801 | 699 867 | 714 801 |
F-statistics | 84.22† | 77.49† | 77.57† | 79.21† |
The full estimation results are available in online supplemental material D.
*P<0.05.
†P<0.001.
DID, difference-in-difference; IPTW, inverse probability of treatment weighting.