Table 3. Summary of the meta-analysis results.
| Analysis | NU | Model | OR (95% CI) | I2 (%) | P |
| Stroke associated infections | 6 | Rem | 0.87 [0.47, 1.60] | 88 | 0.65 |
| Subgroup1: Stroke type | |||||
| Ischemic stroke | 4 | Rem | 0.53 [0.18, 1.55] | 88 | 0.24 |
| Hemorrhage stroke | 1 | Fem | 0.22 [0.09, 0.57] | - | 0.002 |
| Both type | 1 | Fem | 1.92 [1.53, 2.42] | - | <0.00001 |
| Subgroup2: Study design | |||||
| Case-control study | 5 | Rem | 0.71 [0.33, 1.53] | 90 | 0.26 |
| Cohort study | 1 | Fem | 1.88 [1.16, 3.05] | - | <0.00001 |
| Subgroup3: NIHSS | |||||
| NIHSS>10 | 2 | Rem | 0.90 [0.20, 4.05] | 92 | 0.89 |
| 5<NIHSS<10 | 2 | Fem | 1.85 [1.51, 2.26] | 0 | <0.00001 |
| N | 2 | Fem | 0.28 [0.14, 0.56] | 0 | 0.00008 |
| Stroke associated pneumonia | 6 | Rem | 0.76 [0.38, 1.51] | 86 | 0.43 |
| Subgroup1: Stroke type | |||||
| Ischemic stroke | 4 | Rem | 0.77 [0.38, 1.54] | 84 | 0.46 |
| Hemorrhage stroke | 1 | Fem | 0.22 [0.09, 0.57] | - | 0.002 |
| Both type | 1 | Fem | 1.92 [1.39, 2.65] | - | <0.0001 |
| Subgroup2: Study design | |||||
| Case-control study | 5 | Rem | 0.61 [0.31, 1.20] | 79 | 0.15 |
| Cohort study | 1 | Fem | 1.92 [1.39, 2.65] | - | <0.0001 |
| Subgroup3: NIHSS | |||||
| NIHSS>10 | 2 | Rem | 0.69 [0.47, 1.02] | 79 | 0.06 |
| 5<NIHSS<10 | 2 | Fem | 1.85 [1.51, 2.26] | 0 | <0.0001 |
| N | 2 | Fem | 0.29 [0.14, 0.60] | 0 | 0.0008 |
| UTI | 3 | Fem | 1.69 [1.33, 2.14] | 0 | <0.0001 |
| Sepsis | 2 | Fem | 1.35 [0.71, 2.54] | 27 | 0.36 |
Abbreviations: NU: number of articles; Rem: random effects models; Fem: fix effects models; OR: odd ratios; CI: confidence interval; Both type: patients with ischemic or hemorrhage stroke; N: not given.