Abstract
Purpose: :
Cost-effectiveness of screening for amblyopia (AM) has been questioned in reports by Snowdon & Stewart-Brown (1997) and the German Institute for Quality & Cost-Effectiveness in Health Care (2007). Data from the Rotterdam AMblyopia Effectiveness Study (RAMSES), the Waterland Study (WS) and the Rotterdam Elderly Study (RES) allow estimation of its cost-effectiveness.
Methods: :
In the Netherlands, population-based preverbal screening is performed at 9, 14 and 24 months, preschool at 36, 45 or 54, and 60 months. Total examination time per child is ± 316s for all preverbal and ± 322s for all preschool exams, ± 2’07" per child annually. 1 minute screening costs 400 kE ann. at a birth rate of 180000. The RAMSES was a 7-year, prospective, birth-cohort (N=4626) study of AM. In the WS 1250 patients had been treated by a single orthoptist in the Waterland region in 1968-75. 471 of these were occluded for AM, 203 of these were contacted and 137 re-examined in 2003. In the RES, a population-based cohort study among elderly, we found that the period of bilateral visual impairment (BVI) in later life extends from 0.7 years in healthy 6-year-olds to 1.3 years in insufficiently treated AM.
Results: :
In RAMSES AM was detected at age 2.6 for strabismic AM, 2.5 for combined-mechanism AM and 3.9 for anisometropic AM. In the WS these ages were 5.1, 5.7 and 6.6. Approx. 0.8% of RAMSES children had VA > 0.2LogMAR at age 7 with a prevalence of confirmed AM of approx. 3%. The prevalence of occluded children in the WS was estimated at 4.8% (age at detection was 2.7 years later!). In the WS, 30% (estim. prev. 1.6%) of the treated AM cases had VA > 0.2 at age ± 8, 37% at age ± 35. Assuming that a person with BVI costs society 5 kE ann., the reduction of insufficiently treated AM from 1.6% to 0.8% by population-based screening saves 0.6 * 0.8% * 180000 * 5000 = 4320 kE per year. The costs of screening are estimated at 2’07" * 400 kE = 850 kE annually.
Conclusions: :
Screening for AM in The Netherlands is cost-effective even if only considering the costs resulting from loss of the better eye in insufficiently treated AM. The delay between cost of screening and yield in senescence is a concern, however. Reduction of QOL in BVI should next be estimated; its reduction in insufficiently treated AM patients who do not loose their better eye was recently estimated at ± 1% using utility analysis with the trade-off method, but was considerably lower with the standard-gamble method.
Keywords: amblyopia • screening for ambylopia and strabismus • clinical (human) or epidemiologic studies: health care delivery/economics/manpower