Screening for fragile X syndrome: a literature review and modelling study.

OBJECTIVES To compare the effectiveness, estimate the associated costs, and summarise available evidence about the feasibility and acceptability of different screening strategies in England and Wales. Also to establish a model for estimating effectiveness and costs of these different strategies. DATA SOURCES Literature searches were restricted to MEDLINE and EMBASE, as well as citations in included papers. A broad search strategy was used involving all aspects of fragile X syndrome (FXS) and covered all relevant literature published between 1991 and 2001. REVIEW METHODS An assessment was conducted of published literature and efforts focused on the development of a model that could be used to synthesise data from various sources, estimate cost-effectiveness of different strategies, and conduct sensitivity analyses according to different assumptions. RESULTS The identified screening programmes were effective in detecting carriers, but a comparison of different strategies was not possible. Simulation results by the FXS Model showed that, over the first 10 years, 4% of premutation (PM) females and 70% of full mutation (FM) females could be detected by active cascade screening; it is 10% and 58%, respectively, by prenatal screening. The maximal detection rate for FM carriers by active cascade screening is higher than that by prenatal screening (91% versus 71%). However, the maximal rate of detection of female PM carriers by active cascade screening (6%) is much lower than that by prenatal screening (60%). During the first 10 years of simulation, the estimated direct cost per year to the NHS in England and Wales is 0.7-0.2 million pounds sterling by active cascade screening and 14.5-9.1 million pounds sterling by a programme of prenatal screening. The incremental cost per extra carrier detected (using current practice as the reference standard) is on average only 165 pounds sterling by active cascade screening and 7543 pounds sterling by prenatal screening. The incremental cost per FXS birth avoided is on average 8494 pounds sterling by active cascade screening and 284,779 pounds sterling by prenatal screening. CONCLUSIONS The empirical evidence suggests that both prenatal screening and cascade screening are feasible and acceptable. Population-based prenatal screening is more efficacious, but it will cost more than active cascade screening. The active cascade screening of affected families is more efficient, cheaper, but less effective than a population-based prenatal screening. It is suggested that both strategies be evaluated in large-scale trials, which might also help to determine whether and how the different strategies could be simultaneously or sequentially combined.