Selective Fetoscopic Laser Photocoagulation or Expectant Management for Stage I Twin-Twin Transfusion: A Cost-Effectiveness Analysis

Abstract

Background: Selective fetoscopic laser photocoagulation (SFLP) is the preferred intervention for stage II-IV twin-twin transfusion syndrome (TTTS); however, there is no consensus on whether SFLP or expectant management (EM) is the preferred strategy to manage Quintero stage I TTTS. Objective: The objective of this study is to estimate whether SFLP or EM is the cost-effective strategy for management of Quintero stage I TTTS. Study Design: A decision-analysis (DA) model compared SFLP to EM for 1,000 pregnant people with monochorionic-diamniotic twins affected by stage I TTTS. All subjects were assumed to be appropriate candidates for either SFLP or EM. Probabilities, costs, and utilities were derived from the literature. The DA was conducted from a healthcare payor perspective, and the analytic horizon was over the course of an offspring’s lifetime, with primary outcomes of survivorship (i.e., no intrauterine fetal demise or neonatal death) and long-term neurodevelopmental impairment. The model incorporated Markov processes with 4-week cycles throughout pregnancy. Incremental cost-effectiveness ratios (ICER) for each strategy were calculated and compared to estimate marginal cost effectiveness. An ICER of USD 100,000 per quality-adjusted life year was used to define the cost-effectiveness threshold. One-way sensitivity and Monte Carlo analyses (MCA), as well as microsimulations, were performed. Results: For base-case estimates, SFLP was found to be cost-effective compared to EM in the management of stage I TTTS. In one-way sensitivity analysis, varying each variable along pre-specified ranges did not result in changes in the conclusion. MCA projects SFLP as the cost-effective strategy in 100% of runs. Conclusions: With base-case estimates, SFLP is estimated to be the cost-effective strategy for the treatment of Quintero stage I TTTS when compared with EM. This remained true across a wide range of inputs.