Cost-effectiveness of robotic-assisted versus conventional total knee arthroplasty: an analysis from a middle income country.

Background and purpose:  Robotic-assisted total knee arthroplasty (RATKA) can enhance surgical precision. In middle-income countries (MICs), constrained fiscal space and the double burden of rising demand for high-cost technologies and competing public-health priorities-unlike high-income countries with broader fiscal headroom and low-income countries with limited adoption of expensive innovations-make adoption decisions for RATKA particularly challenging. We aimed to evaluate the cost-effectiveness analysis (using a cost-utility framework) of RATKA vs conventional TKA (COTKA) from a societal perspective in Thailand as a MIC.

Methods:  A discrete event simulation model was employed to compare the cost-effectiveness of unilateral RATKA with COTKA over 4.5 years from a societal perspective, using patient-level data from January 2018 to June 2022 from an arthroplasty center in Thailand. Patients were propensity matched to balance comorbidities. Base case analysis assumed 1 robot performs 434 TKA cases per year with an anticipated lifespan of 12.5 years. We considered direct medical, non-medical, and indirect costs, alongside quality-adjusted life years (QALYs) gained from a societal perspective. We calculated incremental net monetary benefits (INMB) and cost-effectiveness ratios (ICERs) as the main outcome measures. Sensitivity analyses and 10 scenario analyses were performed exploring various possible settings. Threshold analyses determined combinations where RATKA could be cost-effective with positive INMB under the Thai cost-effectiveness threshold of US$4,888 per QALY gained.

Results:  The base case analysis involved 157 COTKA and 1570 RATKA matched cases with a mean age of 69 (standard deviation 8 years). The lifetime average outcomes per patient were: COTKA-US$5,031.9 and 9.07 QALYs; RATKA-US$5,666.9 and 9.16 QALYs. The incremental (RATKA-COTKA) differences were +US$633.6 (95% credible intervals [CrI] ~592-675) and +0.085 QALYs (CrI ~0.04-0.13), yielding an ICER of US$7,436.6/QALY. RATKA was not cost-effective compared with COTKA, with an INMB of -216.9 US$/patient. The probability of RATKA being cost-effective at the Thai cost-effectiveness threshold was 44.3%. For RATKA to be economically attractive, 1 robot must operate on at least 640 TKA cases/year over 12.5 years. 3 scenarios found RATKA to be cost-effective: (i) maximal robot utilization (850 cases/year); (ii) lowest capital costs (611,060 US$/robot) with high efficacy for RATKA (hazard ratio [HR] 0.6); and (iii) extreme efficacy of RATKA in reducing complications (HR 0.024).

Conclusion:  In the context of MIC, a broad adoption of RATKA is not economically attractive as treatment of end-stage knee osteoarthritis patients compared with COTKA.