Table 6.
Summary meta-analysis results
| Length of stay | Time to surgery | In-hospital mortality | 1-year mortality | 30-day readmission rate | delirium | |
|---|---|---|---|---|---|---|
| MD; 95% CI I2, p n |
MD; 95% CI I2, p n |
RR; 95% CI I2, p n |
RR; 95% CI I2, p n |
RR; 95% CI I2, p n |
RR; 95% CI I2, p n |
|
| 3 Models | 1.55; 95% CI [− 2.53; -0.57] | − 0.23; 95% CI [− 0.46; 0.01] | 0.72; 95% CI [0.56; 0.92] | 0.86; 95% CI [0.76; 0.97] | 0.50; 95% CI [0.23; 1.12] | 0.81; 95% CI [0.71; 0.92] |
| 80%, p < 0.01 | 61%, p < 0.01 | 36%, p = 0.05 | 59%, p < 0.01 | 45%, p = 0.10 | 26%, p = 0.19 | |
| N = 24a,b | N = 13b,c | N = 23b | N = 13b | N = 6 | N = 13d | |
| ICM | − 2.12; 95% CI [−4.22; -0.03] | − 0.29; 95% CI [− 0.74; 0.16] | 0.65; 95% CI [0.48; 0.88] | 0.87; 95% CI [0.75; 1.01] | 0.46; 95% CI [0.16; 1.29] | 0.87; 95% CI [0.72; 1.05] |
| 81%, p < 0.01 | 69%, p < 0.01 | 28%, p = 0.15 | 60%, p = 0.01 | 55%, p = 0.06 | 0%, p = 0.52 | |
| N = 11a | N = 9c | N = 14 | N = 8 | N = 5 | N = 6d | |
| GCS | − 1.37; 95% CI [− 2.42; − 0.33] | − 0.40; 95% CI [− 0.52; − 0.28] | 0.59; 95% CI [0.27; 1.30] | 0.65; 95% CI [0.31; 1.38] | 0.82; 95% CI [0.23; 2.94] | 0.70; 95% CI [0.45; 1.08] |
| 61%, p = 0.01 | 0%, p = 0.94 | 0%, p = 0.51 | 23%, p = 0.27 | NA | 50%, p = 0.11 | |
| N = 8 | N = 3 | N = 5 | N = 3 | N = 1 | N = 4 | |
| GW | − 1.24; 95% CI [− 4.85; 2.37] | 0.05; 95% CI [-1.42; 1.51] | 1.07; 95% CI [0.52; 2.23] | 0.95; 95% CI [0.68; 1.32] | NA | 0.84; 95% CI [0.55; 1.26] |
| 88%, p < 0.01 | 16%, p = 0.27 | 32%, p = 0.21 | p0%, = 0.40 | NA | 35%, p = 0.22 | |
| N = 6 | N = 2 | N = 5 | N = 3 | N = 0 | N = 3 |
Bold values denote statistical significance at the p < 0.05 level
ICM integrated care model, GCS geriatric consultant service, GW geriatric ward, MD mean difference, RR relative risk, CI confidence interval, I2 test for heterogeneity, p significance for heterogeneity, n number of included studies in meta-analysis, NA not applicable
aDuaso et al. [35] and Naglie et al. [23] are excluded from the result of the meta-analysis because these studies were outlying on the overall LOS result
bBaroni et al. [32] investigated both ICM and GCS
cBiber et al. [43], Duaso et al. [35], and Gregersen et al. [46] are excluded from the result of the meta-analysis because these studies were outlying on the overall TTS result
dKusen et al. [39] was excluded from the result of the meta-analysis because the study was outlying on the overall delirium result