Table 3. GRADE evidence profile of the efficacy and safety of prokinetics in critically ill adult patients receiving gastric feeding tubes.
| Certainty assessment | No. of patients | Effect | Certainty | Importance | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. of studies | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Prokinetics | Placebo | Relative (95% CI) |
Absolute (95% CI) |
||
| Effect on ICU length of stay | ||||||||||||
| 3 | randomized trials | serious 1 | not serious 2 | not serious 3 | serious 4 | none 5 | 96 | 90 | - | MD 2.03 days lower (3.96 days lower to 0.1 days lower) |
⨁⨁◯◯ LOW |
IMPORTANT |
| Effect on hospital length of stay | ||||||||||||
| 5 | randomized trials | serious 6 | not serious 7 | not serious 3 | serious 8 | none 5 | 141 | 109 | - | MD 3.21 days lower (5.35 days lower to 1.06 days lower) |
⨁⨁◯◯ LOW |
IMPORTANT |
| Effect on reported adverse events | ||||||||||||
| 7 | randomized trials | serious 9 | not serious 10 | not serious 3 | serious 11 | none | 105/320 (32.8%) | 120/437 (27.5%) |
RR 1.13 (0.92 to 1.38) |
40 more RAE per 1,000 patients (from 20 fewer RAE to 100 more RAE) |
⨁⨁◯◯ LOW |
IMPORTANT |
| Effect on all-cause mortality | ||||||||||||
| 6 | randomized trials | serious 12 | not serious 13 | not serious 3 | not serious 14 | none | 114/286 (39.9%) | 174/405 (43.0%) |
RR 0.96 (0.81 to 1.14) |
30 fewer deaths per 1,000 patients (from 100 fewer deaths to 40 more deaths) |
⨁⨁⨁◯ MODERATE |
CRITICAL |
CI: confidence interval; RR: risk ratio; MD: mean difference; RAE: reported adverse events.
1. We downgraded the quality of evidence for risk of bias by one level. Two of three included studies had a high or unclear risk of bias.
2. We did not downgrade for inconsistency, I2 = 0% and Chi2 = 1.45, P = 0.48.
3. Although the studies included any critically ill patient, we did not downgrade for indirectness.
4. We downgraded the quality of evidence for imprecision by one level because the total population size is less than 400. The 95% confidence interval contained a small benefit that did not meet the clinical decision threshold (min. one day).
5. We did not downgrade for publication bias, although we could not assess this category reliably due to the small number of eligible studies. Not all included studies showed benefits of the studied intervention.
6. We downgraded the quality of evidence for risk of bias by one level. Most studies had an unclear risk of bias. In addition, one study lacked allocation concealment and blinding.
7. We did not downgrade for inconsistency, I2 = 28% and Chi2 = 5.52, P = 0.24.
8. We downgraded the quality of evidence by one level for imprecision because the population size is less than 400.
9. We downgraded the quality of evidence for risk of bias by one level. Most studies had an unclear risk of bias. In addition, two studies lacked allocation concealment and/or blinding.
10. We did not downgrade for inconsistency, I2 = 0% and Chi2 = 1.64, P = 0.95.
11. We downgraded the quality of evidence for imprecision by one level because the 95% confidence interval around the pooled effect included both no effect and appreciable harm (a relative risk increase greater than 25%).
12. We downgraded the quality of evidence for risk of bias by one level. Most studies had an unclear risk of bias. In addition, two studies lacked allocation concealment, and two studies lacked blinding.
13. We did not downgrade for inconsistency, I2 = 0% and Chi2 = 4.52, P = 0.48.
14. We did not downgrade for imprecision because the 95% confidence interval around the pooled effect did not include both no effect and an appreciable benefit (a relative risk reduction greater than 25%) or appreciable harm (a relative risk increase greater than 25%).