Time interval | Control | Intervention | D-I-D (95% CI) (intervention-control) | P value | ||
n | Estimated ∆SIxAge (95% CI) | n | Estimated ∆SIxAge (95% CI) | |||
(A) | ||||||
Before: all | 200 | −2.0 (−3.1, 0.9) | 151 | −3.0 (−4.2, 1.7) | ||
Post: 0–4 months | 73 | −1.2 (−3.0, 0.6) | 69 | −3.6 (−5.4, 1.7) | −1.4 (−4.4, 1.5) | 0.35 |
Post: 5–8 months | 62 | −1.0 (−2.9, 0.9) | 39 | −2.5 (−4.9, 0.0) | −0.5 (−3.9, 3.0) | 0.79 |
Post: 9–13 months | 98 | −1.3 (−2.8, 0.2) | 63 | −2.2 (−4.2, 0.3) | 0.0 (−2.9, 2.9) | 0.99 |
(B) | ||||||
Before: all | 64 | −5.8 (−8.7, 2.9) | 35 | −6.8 (−10.6, 3.0) | ||
Post: 0–4 months | 22 | −3.8 (−8.4, 0.9) | 19 | −12.4 (−17.6, 7.3) | −7.7 (−15.8, 0.3) | 0.06 |
Post: 5–8 months | 17 | −3.2 (−8.7, 2.3) | 10 | −9.7 (−16.7, 2.8) | −5.5 (−15.1, 4.1) | 0.26 |
Post: 9–13 months | 26 | −4.9 (−9.2, 0.6) | 13 | −4.9 (−10.9, 1.2) | 1.0 (−7.5, 9.4) | 0.82 |
∆SIxAge denotes the change in Shock Index x Age. A more negative delta Shock Index represents more improved shock.
D-I-D denotes the difference-in-differences computed as (change in ∆SIxAge in intervention group)−(change in ∆SIxAge in control group).
*Fifteen cases from the original sample of N=770 were excluded from this analysis due to missing data.