Twelve RCTs met the inclusion criteria with 1,383 participants (730 randomised to LA and 653 randomised to AO). Eight studies were combined for the outcomes of length of operation and hospitalisation, while four studies were combined for the outcome of return to normal activities and ten studies reported data for complications.
Mean operating time in 8 trials was significantly longer with LA (weighted mean difference (WMD) 18.10 minutes, 95% CI: 12.87, 23.15 minutes, p < 0.05) with statistically significant heterogeneity (20.17, 7 df, p < 0.001). The overall conversion rate from LA to OA was 11% (range 5% to 20%).
Complications were reported in 10 trials with no statistically significant heterogeneity: there were fewer wound infections in LA (OR 0.40, 95% CI: 0.24, 0.69) but no significant differences in intra-abdominal abscess rates in 6 trials (OR 1.94, 95% CI: 0.68, 5.58), or rare but serious complications in 5 trials (OR 1.00, 95% CI: 0.27, 5.58).
There was no statistically significant difference in the mean length of hospital stay in 8 trials (WMD 0.16, 95% CI: -0.44, 0.15 days) and the included trials were statistically homogeneous.
There was no statistically significant difference in readmission rates of patients who underwent LA (6.4%) and OA (6.3%) (OR 1.16, 95% CI: 0.54, 2.48 days).
Return to normal activity outcome in 4 trials was statistically significantly earlier after LA (WMD 5.79, 95% CI: -7.38, -4.21 days) and these trials were statistically homogeneous.
Sensitivity analyses did not affect the results.
The overall quality of the 12 trials was 6.4 out of 10 with an inter- rater reliability of 0.81 as measured by an unweighted kappa statistic.