Table 3.
GRADE evidence profile – Comparison: Routine vs selective IOC. Population: laparoscopic cholecystectomy
| Certainty assessment | № of patients | Effect | Certainty | Importance | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| № of studies | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Routine IOC | selective IOC | Relative (95% CI) |
Absolute (95% CI) |
||
| Bile duct injury (laparoscopic cholecystectomy) (assessed with: RR) | ||||||||||||
| 3 | observational studies | very seriousa | not serious | not serious | not serious | publication bias strongly suspectedb | 4/1949 (0.2%) | 10/2114 (0.5%) |
RR 0.78 (0.25 to 2.41) |
1 fewer per 1 000 (from 4 fewer to 7 more) |
⨁◯◯◯ Very low |
CRITICAL |
| Major bile duct injury (laparoscopic cholecystectomy) (assessed with: RR) | ||||||||||||
| 2 | observational studies | very seriousa | not serious | not serious | not serious | publication bias strongly suspectedb | 1/1478 (0.1%) | 3/955 (0.3%) |
RR 0.39 (0.05 to 3.28) |
2 fewer per 1 000 (from 3 fewer to 7 more) |
⨁◯◯◯ Very low |
CRITICAL |
| Success rate of IOC (laparoscopic cholecystectomy) (assessed with: RR) | ||||||||||||
| 4 | observational studies | very seriousa | very seriousc | not serious | very seriousd | publication bias strongly suspectede | 2846/3127 (91.0%) | 372/414 (89.9%) |
RR 0.96 (0.86 to 1.06) |
36 fewer per 1 000 (from 126 fewer to 54 more) |
⨁◯◯◯ Very low |
IMPORTANT |
| Operation time (laparoscopic cholecystectomy) (assessed with: WMD) | ||||||||||||
| 3 | observational studies | very seriousa | very seriousc | seriousf | seriousg | publication bias strongly suspectedh | 857 | 1588 | - |
WMD 14.02 min. more (6.96 fewer to 35 more) |
⨁◯◯◯ Very low |
IMPORTANT |
CI confidence interval, RR risk ratio
aBias is likely due to the presence of confounding factors
bDue to the low number of articles publication bias was not assessed
cInconsistency is likely due to the presence of statistically significant heterogeneity
dConfidence intervals cross the benefit/harm line and 0−effect line
ePublication bias is likely due to funnel plot asymmetry
fIndirect population is likely due to the variable inclusion and exclusion criteria
gImprecision is likely due to the overall effect estimate lies between benefit and harm
hThe risk of publication bias was not assessed due to the low number of studies included