Little et al pose an important question: would the universal offer of carrier screening for spinal muscular atrophy (SMA) during prenatal care be cost-effective, ie, provide good value for money spent? The rapid expansion of the availability of molecular genetic tests for rare disorders, with approximately 1700 tests currently available, poses a challenge to health care providers and payers. We as a society do not have sufficient funds to pay for all possible genetic tests, and prioritization in some form must be undertaken. Little et al and the American College of Obstetricians and Gynecologists are to be commended for addressing the balance of technological opportunity in genetic testing and opportunity cost of scarce health care resources implicit in the concept of cost-effectiveness.

The decision analysis prepared by Little et al estimated the net costs of screening for SMA, including reductions in costs of care associated with termination of affected fetuses, relative to gains in maternal quality-adjusted life years (QALYs). The approach used is similar to that used in cost-utility analyses of prenatal testing for fragile X syndrome and Down syndrome, the results of which appear to have gained acceptance in the obstetrics and gynecology literature. However, this approach has inherent limitations. The QALY was developed to measure the combined effects of premature death and disability on individuals who suffer from a disease or injury, and the relatively minor spillover effects on the health of other family members typically are ignored in cost-utility analyses. QALYs are not well suited to capture the personal utility or value either of diagnostic or genetic information in general or of prenatal carrier testing in particular. Because Little et al modeled the effects on only the woman of an affected fetus or child, it is not surprising that the projected gain in QALYs from prenatal diagnosis in their model was extremely modest. However, that is not the only possible method to calculate QALYs. For example, using a time-tradeoff method to assess the utility of prenatal screening for Down syndrome, Harris et al found that the QALYs associated with reassurance to women with normal findings dominated the negative effects of positive amniocentesis results. Similar results were found in another study that used the theoretically preferable standard gamble method to evaluate carrier testing for cystic fibrosis (CF). Had Little et al incorporated the same assumed utility of negative carrier test results for SMA as previously estimated in these studies of testing for Down syndrome or CF, the projected QALY gains, and hence the cost-effectiveness of the screening, would have appeared more favorable. It should be noted, however, that the value of reassurance from testing for an additional disorder during pregnancy can be determined only by asking women their perceptions about screening for multiple conditions.

The other major drawback of the approach used by Little et al was its emphasis on financial cost savings resulting from the early diagnosis and termination of affected fetuses. This general approach long has been criticized by health economists as inconsistent with economic theory. It implicitly prioritizes testing for disorders that have high lifetime medical costs, such as CF, over disorders, such as SMA, which typically result in early death and, hence, low lifetime medical costs. The prevention of diseases that are fatal in early childhood and have no treatment saves little money relative to disorders that are disabling but not lethal.

The results of a standard cost-effectiveness or cost-utility analysis of prenatal carrier screening are greatly influenced by the projected financial savings to the health care system resulting from the termination of affected fetuses. Such an analysis places little or no value on the ability of parents to prepare for the birth of an affected child, to avoid lengthy diagnostic “odysseys,” or to plan future childbearing. The justification of screening as cost-effective based on the frequency of terminations can seem difficult to reconcile with the clinical genetics paradigm of nondirective offer of counseling, carrier testing, and prenatal diagnosis.

A simpler economic approach–cost-comparison analysis–calculates the cost per carrier or case detected for different testing strategies. Carrier testing for SMA presumably costs more per case detected than does screening for CF because of the lower prevalence of SMA. The cost-comparison approach leaves it to policy makers to value the detection of carriers. Policy makers may consider the opportunity for informed choice by parents and the seriousness of the disorder as well as the potential reduction in health care costs. In addition to the numbers of cases detected, analysts should report the numbers of false-positive, false-negative, and true-negative test results and stakeholders also should take into account such outcomes of testing and their valuations.

The crucial challenge is to ascertain the value placed by prospective parents on screening for CF, SMA, and other disabling or life-threatening congenital disorders. The most direct approach is to ask respondents what they care about. For example, a recent study from The Netherlands found that most young women who were surveyed favored voluntary newborn screening–even for untreatable disorders–to avoid long diagnostic quests or odysseys after the birth of a child with a rare disorder. Economists often value interventions through willingness-to-pay estimates by asking respondents whether they would pay various specified amounts for prenatal tests for disorders such as CF and Down syndrome. However, that contingent valuation approach does not provide information as to what matters most to prospective parents in deciding for which disorders to screen. The newer method of conjoint analysis can be used to elicit information about hypothetical tests. Applications of conjoint analysis to prenatal tests for Down syndrome reported that women placed higher values on the rates of detection and miscarriage relative to testing earlier in pregnancy and that a less sensitive test with faster reporting of results was valued more highly than a more sensitive test with a longer reporting delay. This method also has been applied to voluntary newborn screening for rare disorders lacking effective treatments to elicit parental valuations on the relative importance of disorder prevalence, test sensitivity, developmental disability, physical disability, and life expectancy.

A rational decision-making process for carrier testing requires guidelines and criteria, as was recognized at a National Institutes of Health workshop held in February 2008. Consultation with ethicists, geneticists, social scientists, and other disciplines is essential for the assessment of population genetic screening programs, along with the engagement of policy makers and the public. At the present time, no federal advisory body evaluates or recommends disorders for carrier screening, as the Advisory Committee on Heritable Diseases in Newborns and Children does for proposed disorders for newborn screening. US newborn screening recommendations do not require demonstration of cost-effectiveness, nor do preventive services or public health policies in general.

In conclusion, it is premature to conclude that prenatal carrier screening for SMA either does or does not provide sufficient value for money to support a recommendation for the universal offer of testing. Research is needed on the perceived value to stakeholders, including the public, of carrier screening for SMA in particular and for rare disorders in general. More importantly, a process is needed to balance stakeholder interests, values, and ethical considerations in making recommendations on carrier screening that will be acceptable to heath care providers, payers, and consumers.