Abstract
Purpose:
A recent clinical trial demonstrated the efficacy of home telemonitoring for early detection of incident choroidal neovascularization (CNV) among patients with age-related macular degeneration (AMD). We conducted an economic evaluation of the potential cost-effectiveness and government budgetary impacts of adoption of this technology among eligible Medicare patients.
Methods:
We developed a computer simulation model using effectiveness data from the AREDS2-HOME study, treatment data from the Treat and Extend Wills Eye Hospital study (T&E), and AMD progression from the AREDS study. The model estimated the societal cost, 10-year federal budgetary impact, and the incremental cost-effectiveness ratio (ICER) of telemonitoring patients with CNV in one eye or large drusen and pigment abnormalities in both eyes. We used alternative scenarios, univariate and probabilistic sensitivity analyses to test the effect of uncertainty in data and assumptions.
Results:
Compared to the AREDS2-HOME study control group we estimated home telemonitoring would cost $1,707 from the societal perspective, cost $1,923 per patient over 10 years from the federal government’s perspective, and result in an ICER of $40,286 (95% C.I. savings-$257,517) per QALY gained. Compared to CNV detection as observed in the T&E study, telemonitoring leads to all-payer savings of more than $6,285per person and a 10-year government costs of $946. Cost-effectiveness was highly sensitive to visual acuity at CNV diagnosis in the control group and the risk of bilateral CNV. When compared to real-world observed CNV detection rates, monitoring patients with existing CNV in one eye was cost-saving, while monitoring patients without CNV led to an ICER of $95,825. Telemonitoring was cost-saving in scenarios in which we assumed early detection resulted in a reduction of one or more ranibizumab injection over 10 years.
Conclusions:
Home telemonitoring of AMD patients at risk for CNV was cost-effective when compared to scheduled exams, but increased governmental costs by approximately $2,000 per person over 10 years. Comparing the AREDS2-HOME study results to visual outcomes observed in a nonclinical trial community setting improved the ICER dramatically. The true cost-effectiveness of home telemonitoring is likely to be driven largely by patients’ risk for CNV, their frequency of examination without telemonitoring, and the ability of early detection to avoid one or more injections of ranibizumab.