OBJECTIVE. On the basis of California’s experience implementing a pilot tandem mass spectrometry (MS/MS) screening program, an economic evaluation was conducted to determine the economic benefits and costs of a statewide MS/MS screening program.
METHODS. Cost-effectiveness, benefit/cost, and cost-utility analyses were conducted with a base-case set of assumptions. The base-case assumptions were varied by using a set of more-favorable and less-favorable assumptions to test the robustness of the analysis findings.
RESULTS. The total estimated, annualized, incremental costs of MS/MS screening of 540000 births in California were nearly $5.7 million; 83 affected newborns would be identified. Screening would reduce the expected lifetime costs of medical care for affected newborns by $7.2 million ($9.0 million in the best-case scenario and $1.8 million in the worst-case scenario). When all program costs and savings were considered, screening saved $1.5 million ($3.4 million saved in the best-case scenario and $3.8 million additional costs in the worst-case scenario). With only incremental program costs, the cost per life saved was $708000 and the cost per case detected was $68000. With consideration of the projected lifetime medical care costs, the total cost per case detected was $132000. MS/MS screening produced a benefit/cost ratio of $9.32 ($11.67 with the best-case set of assumptions and $4.34 with the worst-case set of assumptions). In this analysis, the benefits of screening exceeded total program costs by $47.1 million (the net incremental benefit). In the worst-case scenario, the net incremental benefit of screening was $18.9 million. Screening saved 949 quality-adjusted life-years (QALYs) and saved $1628 per QALY in the base case analysis. Under the worst-case scenario, the cost per QALY was $14922.
CONCLUSIONS. We found that the benefits of MS/MS screening outweighed the costs and that the net benefits were significant and robust in various scenarios with various conservative underlying assumptions.
The California legislature authorized and funded a pilot tandem mass spectrometry (MS/MS) screening program, which was conducted from January 7, 2002, through June 13, 2003. During that period, 755698 infants were born, 353894 newborns were screened, and 53 cases of disorders detectable with MS/MS were identified in the screened population (including 2 cases that were missed initially with MS/MS testing but would have been identified with our revised cutoff values). On the basis of this experience, the overall incidence in California is ∼1 case per 6700 newborns screened. We did not include phenylketonuria (PKU) because we were evaluating the marginal benefits of MS/MS screening, compared with the incremental cost of adding the new technology to the existing screening program, which already screens for PKU. Using this incidence figure, we project that statewide screening of 540000 annual births would, on average, identify 83 cases with disorders detectable through MS/MS screening.
Cost-effectiveness, benefit/cost, and cost-utility analyses were conducted to determine whether the benefits achieved through an expanded newborn screening program with MS/MS technology justify the costs of implementing the program. The analyses were conducted from the perspective of the costs of screening to the public (excluding additional costs borne by families), with several sets of assumptions that were varied to test the robustness of the conclusions.
Costs of Program Operation
The cost estimate includes the incremental costs associated with implementing MS/MS screening. The costs of the pilot project differ from the costs of a statewide MS/MS screening program because of differences between the limited scope of the pilot program (eg, PKU not included) and the ongoing statewide newborn screening program. However, the costs of the pilot project can be used to develop reasonable estimates of the costs of adding MS/MS technology to the basic newborn screening program. Direct program costs were allocated to the categories of personnel and administration (6 staff positions), equipment (12 MS/MS instruments plus ancillary equipment), supplies ($3.72 per test for 540000 tests), laboratory contracts (rental of space and labor at $3.46 per test for 540000 tests), and follow-up centers and were based on the actual experience with the personnel needs for a comprehensive program, including education, testing, interpretation, reporting, follow-up testing of false-positive and diagnosed cases, and quality control for clinical laboratory elements.
Valuing the Benefits of Screening
To estimate the direct medical care cost savings that might be realized through MS/MS screening from a payers’ perspective, first we needed to estimate the distribution of clinical outcomes expected with and without screening. We developed a model with the clinical outcome categories used by Insinga et al,1 ie, death, severe neurologic impairment, mild neurologic impairment, acute complications only, and asymptomatic. With a distribution of clinical outcomes similar to that reported by Schulze et al,2 we established our base-case scenario of expected outcomes without screening. Then a base-case scenario of outcomes expected with screening was developed with input from a panel of metabolic specialists in California.
For each clinical outcome category, we estimated the expected average lifetime medical care costs of treatment per individual. The total lifetime costs were calculated by multiplying the estimated lifetime medical care costs per newborn by the number of newborns within each clinical outcome category, with and without screening. To establish our base-case estimate of the lifetime medical care costs of MS/MS-detectable disorders with severe neurologic impairment resulting from inborn errors of metabolism that were not diagnosed through screening, we used (as a proxy) the $1014000 (2003 dollars) estimate published by the Centers for Disease Control and Prevention (CDC)3 for the lifetime costs for a person with mental retardation. A 3% inflation rate was applied to adjust this estimate to 2004 dollars as $1044420. The lifetime costs of moderate developmental delay ($77079, in 2004 dollars), as proposed by Carroll and Downs,4 were used as a proxy for the lifetime medical care costs for a moderately affected individual with a MS/MS-detectable disorder. We estimated the lifetime costs for acute MS/MS-detectable disorder complications only ($40000) and for asymptomatic individuals ($500). The difference between the total lifetime costs with MS/MS screening and the total lifetime costs without MS/MS screening equals the total lifetime medical care costs avoided.
In calculating the total benefits attributable to saving lives through MS/MS screening, we needed to value the lives saved. For our base-case analysis, we used a value derived from a report by the US Environmental Protection Agency5 that valued a life at $5.7 million ($5.5 million in 2003 dollars, adjusted to 2004 dollars with a 3% inflation rate).
Valuing the Costs of Screening
The incremental program costs for the annual cohort of births plus the estimated total lifetime costs of medical care for the affected births in the cohort (for each of the 3 scenarios, as described below), as detected through MS/MS screening, equal the total annualized incremental costs of MS/MS screening (the total marginal program costs). The incremental costs per life saved are calculated by dividing the incremental program costs by the number of lives saved. Similarly, the incremental costs per case detected are calculated by dividing the incremental program costs by the number of cases with MS/MS-detectable disorders diagnosed through screening in any given year. The benefit/cost ratio is calculated by dividing the total incremental program benefits by the total incremental program costs.
A cost-utility analysis calculates the incremental cost per quality-adjusted life-year (QALY), with the survival gains being adjusted for the quality of those lives. With this method, outcomes are assigned a relative value called “utility,” with perfect health being 1 and death being 0. We assigned utility values of 0.90 to newborns with no symptoms, 0.87 to those with acute complications, 0.65 to those with mild/moderate neurologic complications, and 0.39 to those with severe neurologic complications. The utility values for mild/moderate and severe neurologic complications were based on estimates by Bennett et al.6 The utility analysis also incorporates estimates of the expected lifespans of individuals with different levels of disability.7,8 For the utility analysis, we used the $5.7 million value per life saved estimate5 and not the alternative estimate of $4.5 million.9
In the sensitivity analysis, we varied multiple sets of assumptions to test the robustness of the base-case analyses. For the estimate of the lifetime medical care costs avoided through screening, we held constant the base-case clinical outcomes expected without screening and calculated the expected costs by using a best-case scenario and then a worst-case scenario. We also varied the estimate of the lifetime medical care costs for a MS/MS-detectable case that was not screened, using the $0.5 million to $1.5 million range suggested by Schulze et al,2 on the basis of projected expenses for hospitalizations and medications. The middle ($1 million) of this estimate range is consistent with the 2003 CDC estimate3 that was used in the base-case analysis ($1014000). Lastly, an alternative value per life saved of $4.5 million9 was used in the sensitivity analysis, as a more-conservative estimate than the $5.7 million used in the base-case analysis.
Costs of Program Operations
The first-year program startup costs would be $9.2 million. These costs would include $4 million in initial equipment costs. If the $4 million in equipment costs were depreciated over 8 years, then the ongoing incremental annualized direct costs to the California Newborn Screening Program for MS/MS technology would be almost $5.7 million (Table 1). The costs of collection of specimens, distribution of results, and tracking of cases would not be increased and are not included in the incremental costs.
Valuing the Benefits of Screening
Table 2 summarizes the base-case outcomes expected with and without screening. The base-case scenario with no MS/MS screening assumed that 10 deaths would occur. The base-case scenario with screening (scenario A) assumed that 2 deaths would occur. The sensitivity analysis used a best-case scenario with screening (scenario B) that assumed that no deaths would occur and a worst-case scenario (scenario C) that assumed that 6 deaths would occur. Compared with the base-case scenario, the best-case scenario assumed that more cases would be asymptomatic, whereas the worst-case scenario assumed that more cases would involve severe or mild/moderate impairment.
Table 3 presents the estimated total lifetime medical care costs avoided for the base-case analysis and the alternative scenarios. The base-case analysis (with the $1 million lifetime medical care cost estimate) showed that, with screening, more than $7.2 million in medical care costs would be avoided. In contrast, the best-case scenario (with an estimate of $1.5 million in lifetime medical care costs) produced an estimate of $12.7 million in medical care costs avoided, and the worst-case scenario (with an estimate of $0.5 million in lifetime medical care costs) produced an estimate of just more than $713000 in lifetime costs avoided with screening.
Table 4 presents several MS/MS screening economic outcomes for screening scenarios A, B, and C. In scenario A (the base-case analysis, assuming $1 million in lifetime medical care costs), 8 lives are saved with screening and the total costs saved are $1.5 million for the cohort of 83 cases identified through screening, compared with the estimated costs expected in the absence of screening. With the use of only incremental program costs, the costs per life saved are $708000 and the costs per case detected are $68000. With consideration of the projected lifetime medical care costs, the total costs per case detected are $132000. With the lower and higher estimates of lifetime medical care costs, the costs per case detected are $112000 and $148000, respectively.
The same set of economic measures were calculated for scenario B (best-case analysis) and scenario C (worst-case analysis) (Table 4). In scenario B, 10 lives are saved and the total costs saved through screening ranges between −$922000 (spent) and $3.4 million (saved), depending on which estimate of lifetime medical care costs is used. For scenario C, in which only 6 lives are saved, the total saved costs with screening range between −$2.9 million and −$4.9 million.
Table 5 presents the total incremental benefits, total incremental program costs, net incremental benefits, and benefit/cost ratio for all 3 scenarios, with 3 estimates of the lifetime medical care costs and 2 estimates of the value of lives saved. In the base-case scenario, the benefits of screening exceed program costs by $47.1 million (the net incremental benefit). In the worst-case scenario, the net incremental benefit of screening is $14.1 million; in the best-case scenario, the net incremental benefit of screening is $60.4 million. The benefit/cost ratio is $9.32 for the base-case scenario and ranges between $8.65 and $9.89 for the lower and higher lifetime medical care cost estimates, respectively (assuming a $5.7 million value of life). If a lower value ($4.5 million) for the lives saved is assumed, then the benefit/cost ratio is reduced, but not by much ($6.96 to $8.19). In the best-case scenario, the lowest benefit/cost ratio is $8.78 and the highest is $12.31; in the worst-case scenario, the benefit/cost ratio is between $3.30 and $4.50.
The results of the cost-utility analysis are shown in Table 6. Scenario A (base-case analysis) produces 949 additional QALYs and saves $1628 per QALY. With the lower estimate of $0.5 million in lifetime medical care costs, the cost per QALY is $2389. Scenario B (best-case analysis) produces 1221 additional QALYs, ranging between a savings of $5797 per QALY and a cost of $755 per QALY. In scenario C (worst-case analysis), 259 QALYs are achieved, at a cost of $11401 to $19129 per QALY.
The economic analysis presented here is a top-down approach designed for practical program needs. This analysis weighs the benefits of expanding the newborn screening program with MS/MS technology against the additional program costs required to screen all 540000 California newborns. The value of the analysis is to establish an understanding of how the benefits and costs of screening change with different assumptions of program effectiveness and expected costs.
Because the assignment of a value to a life saved is controversial, we used 2 different estimates, one that values a life saved at $5.7 million and one that values a life saved at $4.5 million. These produce benefit/cost ratios between 3.30 and 12.31 (with different assumptions of lifetime medical care costs). However, even if we assumed a much lower value of life, eg, $1 million, the (base-case) benefit/cost ratio of $2.68 is still quite acceptable.
Many health economists regard the benefit/cost ratio as too subjective, because it is based on the value assigned to lives saved. The cost-utility approach was developed so that all benefits could be presented with respect to QALYs. This analysis is robust even with lower estimates of the value of lives saved. In the worst-case scenario, which produces only 259 QALYs saved, the cost per QALY is between $11000 and $19000. This is similar to the results of the Wisconsin economic analysis of MS/MS screening (for the whole spectrum of MS/MS-detectable disorders that can be diagnosed through newborn screening), in which the cost per QALY was estimated to be $15252.1 Generally economists have accepted a standard that a cost per QALY of $50000 or less is considered a cost-effective investment.10 In this analysis, this threshold is easily met.
The treatment costs avoided through screening are based on the underlying distribution of outcomes expected with MS/MS screening, compared with not having a program. The distributions of possible outcomes were varied in the base-case, best-case, and worst-case scenarios, to account for the lack of certainty regarding the long-term clinical outcomes expected for the cohort of 83 patients with MS/MS-detectable disorders.
The CDC estimate of $1014000 in lifetime costs for a mentally retarded person includes a large proportion of indirect costs attributable to lost productivity. However, in this analysis, we used the same estimate to represent the expected lifetime medical care costs for a severely affected, neurologically impaired child with a MS/MS-detectable disorder that was not detected through newborn screening. To account for the uncertainty with respect to this estimate, which is by far the largest cost category, we varied the expected lifetime medical care costs within the range of $0.5 million to $1.5 million proposed by Schulze et al,2 which represents their estimate of the average lifetime costs of hospitalization and medication for an individual affected with a MS/MS-detectable disorder not detected through screening. As expected, the higher estimate ($1.5 million) produced the most favorable cost-effectiveness and benefit/cost ratio values. The higher estimate may be the most realistic one, on the basis of the experiences reported by several California parents of children born with MS/MS-detectable disorders that were not diagnosed through screening. It was not uncommon to have $1 million in medical costs reached within the first few years of life (for very long-chain acyl-CoA dehydrogenase deficiency and glutaric acidemia type 1 disorders).
In the base-case and best-case scenarios, MS/MS screening almost always saved money, compared with the costs that would have been incurred in the absence of screening. The range of benefit/cost ratios calculated with 9 different sets of assumptions showed that MS/MS screening makes sense even with the most conservative assumptions. MS/MS screening leads to sizeable cost savings at best and is a reasonably good value, according to standard health economic benchmarks, at worst.
This work was supported by a grant from the Health Resources and Services Administration (1 H46 MC 00199-03).
Special thanks go to Lisa A. Faulkner, PhD, Martin Kharrazi, PhD, Stephen M. Downs, MD, MS, and Richard Scheffler, PhD, for their valuable review and feedback. Special thanks also go to Bruce Barshop, MD, Susan Winter, MD, and Steven Cedarbaum, MD, who provided input in the development of the program effectiveness scenarios.
- Accepted December 27, 2005.
- Address correspondence to Lisa Feuchtbaum, DrPH, MPH, Genetic Disease Branch, California Department of Health Services, 850 Marina Bay Pkwy, F175, Richmond, CA 94804. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Schulze A, Lindner M, Kohlmuller D, Olgemoller K, Mayatepek E, Hoffmann GF. Expanded newborn screening for inborn errors of metabolism by electrospray ionization-tandem mass spectrometry: results, outcome, and implications. Pediatrics.2003;111 :1399– 1406
- ↵Carroll AE, Downs SM. Comprehensive cost-utility analysis of newborn screening strategies. Pediatrics.2006;117(5) . Available at: www.pediatrics.org/cgi/content/full/117/5/SE1/e287
- ↵US Environmental Protection Agency. Benefits of the Proposed Inter-State Air Quality Rule. Research Triangle Park, NC: US Environmental Protection Agency; 2004. Report EPA 452-03-001
- ↵Bittles AH, Petterson BA, Sullivan SG, Hussain R, Glasson EJ, Montgomery PD. The influence of intellectual disability on life expectancy. J Gerontol A Biol Sci Med Sci.2002;57 :M470– M472
- ↵Arias E, Smith BL. Deaths: preliminary data for 2001. Natl Vital Stat Rep.2003;51(5) :1– 44
- ↵Hirth RA, Chernew ME, Miller E, Fendrick AM, Weissert WG. Willingness to pay for a quality-adjusted life year: in search of a standard. Med Decis Making.2000;20 :332– 342
- Copyright © 2006 by the American Academy of Pediatrics