A probabilistic multicriteria decision analysis of robot-assisted versus laparoscopic sacrocolpopexy: a secondary analysis of published meta-analytic data
摘要
Objectives: Given that the overall comparative advantage between robot-assisted sacrocolpopexy (RASC) and laparoscopic sacrocolpopexy (LSC) remains unclear, a secondary analysis using multicriteria decision analysis (MCDA) was conducted to explore their overall differences across multiple outcomes. Methods: A secondary quantitative analysis was performed using pooled effect estimates from the meta-analysis by Chang et al. Ten outcomes were evaluated, such as the operative time, blood loss, perioperative complications, and anatomical measures. Each pooled estimate and its 95% confidence interval were approximated by a normal probability distribution under standard meta-analytic assumptions. Monte Carlo sampling was employed to generate estimated probabilities that RASC outperformed LSC. Outcomes were categorized into resource, safety, and efficacy domains. Four MCDA scenarios were examined: equal weighting and scenarios emphasizing resource, safety, or efficacy. Composite MCDA scores and the probability that RASC was preferred overall (Pr[RASC preferred]) were calculated for each scenario. Robustness was evaluated using weight-perturbation sensitivity analysis. Results: At the individual outcome level, RASC demonstrated a very high estimated probability of reducing blood loss and intraoperative complications, whereas LSC was highly likely to provide a shorter operative time. In MCDA, RASC was consistently favored across all weighting scenarios, with the Pr(RASC preferred) value ranging from 0.86 to 0.95 and expected MCDA scores from 0.66 to 0.76. Sensitivity analysis demonstrated only minor fluctuations in MCDA outputs, supporting robustness. Conclusions: When multiple outcomes were considered simultaneously, the overall profiles of RASC and LSC demonstrated differences across clinically plausible weighting structures.