Table 3.
Results of subgroup and sensitivity analyses for risk of 30-day mortality.
| No. of studies | No. of data | No. of patients | Heterogeneity | Pooled risk ratio (95% CI) | Significance | |
|---|---|---|---|---|---|---|
| Age of participants | ||||||
| <60 yr | 4a | 4a | 25,68 | I-squared = 0.0%, P = 0.402 | 0.975 (0.301, 3.162) | p = 0.967 |
| ≥60 yr | 13a | 13a | 62,150 | I-squared = 41.3%, P= 0.059 | 1.716 (1.244, 2.366) | p = 0.001 |
| Unknown | 3b, c | 5b, c | 132,986 | I-squared = 0.0%, P = 0.769 | 1.164 (1.025, 1.322) | p = 0.019 |
| Type of surgery | ||||||
| Orthopedic surgery | 10 | 10 | 49,474 | I-squared = 3.7%, P = 0.406 | 1.827 (1.165, 2.865) | p = 0.009 |
| Vascular surgery | 3b | 3b | d | I-squared = 69.3%, P = 0.038 | 1.560 (0.861, 2.827) | p = 0.143 |
| Ventral hernia repair | 1a | 2a | 7,733 | I-squared = 0.0%, P= 0.978 | 0.883 (0.479, 1.629) | p = 0.691 |
| Bariatric surgery | 1c | 2c | 89,373 | I-squared = 0.0%, P = 0.837 | 1.278 (0.881, 1.855) | p = 0.196 |
| Emergency abdominal surgery | 1 | 1 | 4,560 | – | 1.177 (0.998, 1.388) | p = 0.053 |
| Tonsillectomy | 1 | 1 | 590 | – | 5.377 (0.220, 131.419) | p = 0.302 |
| Major non-cardiac surgery | 1 | 1 | 360 | – | 2.940 (0.977, 8.847) | p = 0.055 |
| General surgery | 1b | 1b | e | – | 1.180 (1.001, 1.391) | p = 0.049 |
| Upper extremity surgery | 1 | 1 | 4,451 | – | 3.402 (1.049, 11.031) | p = 0.041 |
| RR/OR type | ||||||
| Crude RR | 16a | 17a | 69,669 | I-squared = 28.8%, P = 0.129 | 1.614 (1.218, 2.138) | p = 0.001 |
| Adjusted OR | 3 | 5b, c | 128,035 | I-squared = 0.0%, P = 0.584 | 1.161 (1.021, 1.319) | p = 0.022 |
| Comparability score of NOS | ||||||
| 0 point | 15a | 16a | 67,384 | I-squared = 13.8%, P = 0.296 | 1.802 (1.317, 2.466) | p < 0.001 |
| 1 point | 2b, c | 4b, c | 125,170 | I-squared = 0.0%, P = 0.832 | 1.156 (1.017, 1.314) | p = 0.026 |
| 2 points | 2 | 2 | 5,150 | I-squared = 0.0%, P = 0.352 | 1.182 (1.002, 1.393) | p = 0.047 |
a
Qin et al. (39) provided mortality information for open ventral hernia repair and laparoscopic ventral hernia repair, respectively.
b
Serio et al. (36) provided mortality information for vascular surgery and general surgery, respectively.
c
Leonard-Murali et al. (23) provided mortality information for laparoscopic sleeve gastrectomy and laparoscopic Roux-en-Y gastric bypass, respectively.
d
Serio et al. (36) did not provide demographic information for vascular surgery.
e
Serio et al. (36) did not provide demographic information for general surgery.