Abstract
Purpose :
Collagen crosslinking (CXL) is a promising new intervention for the treatment of keratoconus (KCN) that is seldom covered by insurers. Initial estimates for the incremental cost-effectiveness ratio (ICER) for CXL compared to standard treatment range from $4300 per quality-adjusted life year (QALY) to $60000 per QALY, yet the willingness to pay (WTP) of many payors is often within this range, making the overall cost-effectiveness of CXL unclear. Using a novel simulation of KCN progression, we tested the hypothesis that varying CXL costs and efficacy, as well as other parameters that differ regionally, would have significant effects on the cost-effectiveness of CXL.
Methods :
A stochastic Markov model with two identical simulated cohorts of 10,000 patients (20,000 eyes) each were created. One received CXL while the other received standard care. Eyes were classified according to visual acuity and the progression of KCN was simulated at one-year cycles in all eyes until death. QALYs and costs accumulated throughout the simulation were used to calculate ICERs for CXL versus control groups. Four parameters were tested by sensitivity analysis to determine their effect on ICER: cost of CXL, CXL effect duration, quality of life in KCN, and time discount rate.
Results :
At a WTP threshold of $50,000 and $100,000 per QALY, both CXL cost and the discount rate were statistically significant factors on the cost-effectiveness of CXL (p < 0.0005), but CXL duration and quality of life values in KCN were not. At a WTP threshold of $30,000, CXL cost and discount rate (p < 0.0005), as well as quality of life (p = 0.036), were significant. 96.6%, 97.9%, and 99.0% of parametric scenarios tested were cost-effective at a WTP of $30,000, $50,000, and $100,000, respectively.
Conclusions :
CXL is cost-effective compared to current standard management across a wide range of realistic economic scenarios. The cost of CXL and time discount rate are significant in the determination of cost-effectiveness. At lower WTP thresholds, more precise calculations of quality of life in KCN will be necessary to accurately determine the cost-effectiveness of CXL. Different estimates of long-term CXL efficacy do not appear to be significant in determining cost-effectiveness.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.