Table 2.
|
|
Secular trend |
|
|
|
|
Pre-post difference |
||
---|---|---|---|---|---|---|---|---|---|
Model | Group | Period | HR | 95% CI | P-value | P#b | HR | 95% CI | P-value |
Full ITSaModel |
<65/non-disabled |
before policy change |
1.00 |
0.97-1.02 |
0.681 |
0.364 |
0.91 |
0.83-1.00 |
0.063 |
|
|
after policy change |
1.02 |
0.98-1.06 |
0.418 |
|
|
|
|
|
≥65/disabled |
before policy change |
1.05 |
1.02-1.08 |
<0.001 |
0.515 |
0.94 |
0.85-1.04 |
0.220 |
|
|
after policy change |
1.03 |
0.99-1.08 |
0.123 |
|
|
|
|
Intermediate Model |
<65/non-disabled |
overall |
1.00 |
0.98-1.02 |
0.989 |
- |
0.93 |
0.85-1.01 |
0.099 |
|
≥65/disabled |
overall |
1.04 |
1.02-1.07 |
<0.001 |
- |
0.93 |
0.84-1.02 |
0.135 |
Final Model |
<65/non-disabled |
- |
- |
- |
- |
- |
0.93 |
0.89-0.97 |
0.002 |
≥65/disabled | overall | 1.04 | 1.02-1.07 | <0.001 | - | 0.93 | 0.84-1.02 | 0.135 |
aITS = interrupted time series.
bP# = P-value for the equality of the within-group trends before and after policy change.