Table 5.
Association of tight glucose control vs usual care with outcomes among critically ill adults, stratified by whether using continuous glucose monitoring (CGM)
| Subgroup of trials | Number of trials | Event number/total number (%) | OR (95% CI) | I2 (%) | |
|---|---|---|---|---|---|
| Tight control | Usual control | ||||
| Mortality | |||||
| Using CGM | 3 | 97/1553 (12.9) | 86/1514 (9.2) | 1.13 (0.83–1.53) | 0 |
| Not using CGM | 3 | 13/441 (3.5) | 27/533 (4.2) | 0.70 (0.18–2.69) | 49 |
| Overall | 6 | 110/1974 (5.6) | 113/2047 (5.0) | 0.95 (0.62–1.45) | 40 |
| Any hypoglycemia (< 60 mg/dl) | |||||
| Using CGM | 3 | 259/1533 (30.4) | 99/1514 (11.4) | 3.00 (2.07–4.36) | 54 |
| Not using CGM | 2 | 24/392 (6.1) | 5/396 (4.7) | 5.23 (1.95–4.00) | 0 |
| Overall | 5 | 371/1925 (19.3) | 116/1910 (6.1) | 4.39 (2.39–8.06) | 83 |
| Severe hypoglycemia (< 40 mg/dl) | |||||
| Using CGM | 2 | 85/1533 (30.4) | 22/1514 (11.4) | 3.88 (2.41–6.26) | 0 |
| Not using CGM | 3 | 112/392 (28.6) | 17/396 (4.7) | 9.35 (1.49–58.83) | 88 |
| Overall | 5 | 109/1925 (18.4) | 27/1910 (5.8) | 4.11 (2.67–6.32) | 0 |
| Sepsis | |||||
| Using CGM | 3 | 91/1553 (12.9) | 101/1514 (9.2) | 0.88 (0.66–1.18) | 0 |
| Not using CGM | 3 | 125/441 (3.5) | 174/533 (4.2) | 0.76 (0.18–2.69) | 69 |
| Overall | 6 | 216/1947 (11.0) | 275/2047 (13.4) | 0.83 (0.63–1.08) | 32 |
| Dialysis | |||||
| Using CGM | 2 | 67/1184 (4.5) | 97/1165 (6.8) | 0.64 (0.46–0.89) | 0 |
| Not using CGM | 1 | 2/349 (4.5) | 6/351 (6.8) | 0.33 (0.07–1.65) | NA |
| Overall | 3 | 69/1533 (4.5) | 103/1516 (6.8) | 0.63 (0.45–0.86) | 0 |
| Seizures | |||||
| Using CGM | 3 | 31/1533 (2.0) | 31/1514 (2.0) | 0.98 (0.59–1.63) | 51 |
| Not using CGM | 0 | NA | NA | NA | NA |
| Overall | 3 | 31/1533 (2.0) | 31/1514 (2.0) | 0.98 (0.59–1.63) | 51 |
OR odds ratio, CI confidence interval, NA not applicable