: To evaluate the cost-effectiveness of osteoporosis treatments for women at high fracture risk and estimate the population-level impact of providing bisphosphonate therapy to all eligible high-risk US women.

: Fractures, healthcare costs, and quality-adjusted life-years (QALYs) were estimated over 10 years using a Markov model.

: No therapy, risedronate, alendronate, ibandronate, and teriperatide (PTH) were compared among 4 risk groups. Sensitivity analyses examined the robustness of model results for 65-year-old women with low bone density and previous vertebral fracture.

: Women treated with a bisphosphonate experienced fewer fractures and more QALYs compared with no therapy or PTH. Total costs were lowest for the untreated cohort, followed by risedronate, alendronate, ibandronate, and PTH in all risk groups except women aged 75 years with previous fracture. The incremental cost-effectiveness of risedronate compared with no therapy ranged from cost saving for the base case to $66,722 per QALY for women aged 65 years with no previous fracture. Ibandronate and PTH were dominated in all risk groups. (A dominated treatment has a higher cost and poorer outcome.) Treating all eligible women with a bisphosphonate would cost an estimated additional $5563 million (21% total increase) and would result in 390,049 fewer fractures (35% decrease). In the highest risk group, the additional cost of therapy was offset by other healthcare cost savings.

: Osteoporosis treatment of high-risk women is cost-effective, with bisphosphonates providing the most benefit at lowest cost. For highest risk women, costs are offset by savings from fracture prevention.

Osteoporosis has been defined as a skeletal disorder characterized by compromised bone strength predisposing a person to an increased risk of fracture.

 The National Health and Nutrition Examination Survey (NHANES) III estimated that 13% to 18% of women in the United States have osteoporosis.

 The prevalence of postmenopausal osteoporosis (PMO) increases with age from approximately 6% at age 50 years to more than 50% above age 80 years.

 The high osteoporosis prevalence coupled with its significant health consequences makes effective prevention and treatment leading public health concerns.

Current treatment options include bisphosphonates (eg, alendronate, risedronate, ibandronate), calcitonins, parathyroid hormone (PTH), and selective estrogen receptor modulators. 

Cost-effectiveness analysis provides important information about the value of alternative therapies that may assist decision makers who seek to equitably allocate constrained resources to achieve maximum healthcare benefits on a population level. Although several articles have addressed the economic value of specific agents,

 the cost-effectiveness of all currently available bone-specific agents approved for treatment of PMO in the United States has not been reported. Another point of differentiation is that our study includes a population-based budget impact analysis to evaluate a policy to extend treatment to all eligible patients within defined risk groups. Budget impact analysis estimates the cost of treatment on an annual basis for budgeting purposes. 

The objective of this study was to evaluate (from the healthcare system perspective) the cost-effectiveness of risedronate therapy compared with alendronate, ibandronate, and PTH for the treatment of women with PMO at high fracture risk. A secondary objective was to estimate the population-level impact of a decision to provide bisphosphonate treatment to all eligible US women not currently treated. 

For all model parameters, published data sources were used wherever possible and consisted primarily of clinical trials, economic studies, observational studies, and epidemiologic databases. Details are provided below and summarized in 

. With respect to treatment, the base-case estimates for therapy duration, offset, and discontinuation were derived from advice from clinical experts.

Our analysis focuses on treatment of postmenopausal women at high risk for fractures; therefore, fracture rates in the age-matched general population were adjusted to reflect the increased risk of new fractures in patients with low BMD and a previous vertebral fracture using the relative risk values noted in Table 1.

 This adjustment has been described fully  elsewhere.

Age-specific hip fracture incidence rates for the total US female population aged 65 to 100 years (all races) were obtained from a retrospective study on the Nationwide Inpatient Sample 2001 hospital discharge database. The analysis included only closed fractures that did not result from severe trauma, and were defined as inpatient hospital cases with 1 of the following 

International Classification of Diseases, Ninth Revision

 codes as the primary diagnosis: 820.0x (transcervical), 820.2x (pertrochanteric), 820.8x (neck of femur).

Age-specific incidence rates for clinically ascertained vertebral fracture were taken from an analysis of the Rochester  Epidemiological Project database, which captured patients treated on both an inpatient and outpatient basis (only 24% of fracture patients are hospitalized). 

Mortality in the year after hip fracture was modeled based on analysis of Medicare claims data from 1999 to 2000 with age-specific mortality rates per 10,000 estimated at 130.29 for age 65 to 69 years, 138.07 for age 70 to 74 years, 166.80 for age 75 to 79 years, 203.38 for age 80 to 84 years, 298.74 for age 85 to 89 years, and 298.74 for age 90 years and older. No excess mortality was modeled following vertebral fracture. 

The bone-specific osteoporosis treatments evaluated in the model include risedronate, alendronate, ibandronate, and PTH. These reflect current practice patterns, capturing those treatments that are used most frequently. The model did not consider possible concomitant calcium or vitamin D treatment with these therapies, as the efficacy estimates for the comparator therapies were derived from comparisons with placebo patients who received clinically appropriate calcium and/or vitamin D supplementation. To estimate fracture rates in the treated cohort, therapy-specific efficacy rates were multiplied by adjusted age-specific fracture incidence. Therapy-specific efficacy values (percent risk reduction for hip and vertebral fracture) were obtained from randomized controlled trials with a patient population similar to the base-case risk group.

 For PTH and ibandronate, where there was no evidence (statistically significant data from a randomized controlled trial) to support the effect of a treatment on the incidence of hip fracture, only the effect on the incidence of vertebral fracture was modeled. Although we modeled clinical vertebral fracture incidence, treatment effectiveness was estimated based on clinical trial data, which relies on radiographically evident vertebral fracture.

To calculate the QALYs, utility weights were applied to each health state. Utilities reflect how quality of life in a health state is valued on a scale from 0 (death) to 1 (perfect health). This analysis assumed an age-specific utility weight of 0.833 for women aged 65 to 69 years and 0.792 for women aged 75 to 79 years in the general population.28 Age-specific utilities were reduced following fracture based on published evidence (Table 1).

To test the robustness of model results, model parameters were varied in multiple 1-way sensitivity analyses for women aged 65 years with low BMD and a previous vertebral fracture. To characterize the impact of uncertainty on treatment efficacy in the cost-effectiveness analyses, analyses that utilized either upper or lower bounds of the 95% confidence intervals for treatment efficacy for each agent were undertaken. Analyses that varied the analytic time horizon from 3 years to lifetime (base case = 10 years) and modified the therapy offset to 5 years (base case = immediate offset) also were undertaken. Therapy offset, or percentage of maximum effect after cessation of therapy, was assumed to decline linearly over time at 90%, 70%, 50%, 30%, and 10% for years 1, 2, 3, 4, and 5, respectively.

 For alendronate, however, based on data from the Fracture Intervention Trial Long-term Extension (FLEX) study,

a 50% slower rate of efficacy decline was used (ie, 90%, 80%, 70%, 60%, 50%). The impact of health utilities on the base-case estimates was evaluated first by varying the utilities by &#177;25%, and then by limiting utility decrements due to fracture only to the year of the fracture.

 Additional analyses that addressed therapy discontinuation and fracture costs also were completed based on results from the Persistence Study of Ibandronate versus Alendronate (PERSIST) trial.

 We used a cumulative therapy discontinuation rate of 76%,

 except for ibandronate, for which persistence was assumed to be 50% better than it was for weekly bisphosphonate dosing.

Given that treatment efficacy is a key factor driving the results of the cost-effectiveness analysis, we explored a broad  range of comparisons for this parameter value. The trials were not able to show any statistically significant difference in effects because they may have been underpowered for hip fractures; therefore, we explored a 1-way sensitivity analysis scenario where a 90% hip fracture efficacy rate was applied to PTH. We also undertook a probabilistic sensitivity analysis, applying triangular distributions to the efficacy variables and using upper and lower 95% confidence intervals as minimum and maximum values. 

Over the 10-year period, women aged 65 years with a previous vertebral fracture who were treated with a bisphosphonate (risedronate, alendronate, or ibandronate) experienced fewer hip and vertebral fractures (a total of 441-464 fractures) compared with those receiving no therapy (550 fractures) or treatment with PTH (501 fractures) (

). Likewise, those treated with a bisphosphonate (risedronate, alendronate, or ibandronate) had more QALYs (6.646, 6.647, and 6.624 QALYs, respectively) compared with those receiving no therapy (6.580 QALYs) or treatment with PTH (6.608 QALYs). The most hip fractures (n = 137) and vertebral fractures (n = 413) were experienced by those who received no therapy. Women treated with risedronate experienced the fewest hip fractures (n = 105), whereas patients treated with ibandronate experienced the fewest vertebral fractures (n = 327). Except for those who received no therapy, women treated with PTH experienced the most vertebral fractures (n = 364), whereas those treated with ibandronate or PTH experienced the most hip fractures (n = 137). A similar pattern was observed among treatments across each risk group.

Total cost was lowest for the untreated cohort, followed by risedronate, alendronate, ibandronate, and PTH in all risk groups except patients aged 75 years with previous fracture. In that group the total costs were lowest for the risedronate cohort, followed by alendronate, no therapy, ibandronate, and PTH.

The cost-effectiveness results changed qualitatively with changes in the assumptions about treatment efficacy and analytical time horizon (

). That is, from the decision makers&#8217; point of view, when $50,000 per QALY was considered to be a decision threshold, the decision to adopt a treatment strategy changed. When the low estimates for efficacy were modeled, the cost-effectiveness for risedronate compared with no therapy changed from $22,068 per QALY gained to $114,694 per QALY gained. When a 3-year time horizon was assumed, the cost-effectiveness estimates for risedronate compared with no therapy changed from $22,068 per QALY gained to $85,391 per QALY gained, and the cost-effectiveness estimates for alendronate compared with risedronate changed from $362,845 to $258,803 per QALY gained. Although the cost-effectiveness estimates for alendronate compared with risedronate changed substantially in the sensitivity analyses, all the values were much greater than $50,000 (ie, dominated), with the only exception being the 5-year therapy offset scenario, where alendronate&#8217;s cost per QALY gained was similar to risedronate&#8217;s, so this would not likely impact the adoption decision. A comparable pattern emerged for the results reported as cost per hip fracture averted. Both ibandronate and PTH were dominated in all of the sensitivity analyses measured as both cost per QALY gained and cost per hip fracture averted, except for the therapy discontinuation analysis, where ibandronate was on the cost-effectiveness frontier for cost per QALY gained. 

Aggregating the 4 risk groups yielded an estimated 8.23 million women in the United States aged 65 to 84 years with low BMD. Currently, 27% of these women receive bisphosphonates and a further 2.9% receive PTH or calcitonin. The rest remain  untreated. Based on our estimates of cost-effectiveness of the individual therapies over a 3-year time horizon, a decision to treat the approximately 5.7 million women not receiving active therapy with bisphosphonates would cost an additional $5563 million (21% increase in total cost) and would result in an additional 83,159 QALYs (0.45% increase) and 390,049 fewer fractures (35% decrease) (

In the highest risk group (women aged 75 years with previous fracture), the additional cost of therapy was offset by the savings in inpatient, outpatient, and long-term care for a net savings of $18 million (0.2% decrease in total cost), accompanied by an additional 18,232 QALYs (0.5% increase) and 96,786 fewer fractures (28% decrease). 

Currently 48.5% of women with previous fracture (aged 65-84 years with low BMD) are treated with bisphosphonates and an additional 5.1% receive PTH or calcitonin. A policy decision to treat with bisphosphonates the approximately 1.5 million untreated women in this group would result over a 3-year period in 32,558 QALYs gained (0.46% increase) and 155,038 fractures avoided (28% decrease), at an additional societal cost of $783 million (6.2% increase) (data not shown).

2. Looker AC, Orwoll ES, Johnston CC Jr, et al.

 Prevalence of low femoral bone density in older U.S. adults from NHANES III. 

4. Bone Health and Osteoporosis: A Report of the Surgeon General.

 Rockville, MD: US Dept of Health and Human Services, Public Health Service, Office of the Surgeon General; 2004. 

http://www.surgeongeneral.gov/library/bonehealth/content.html

6. Pfister AK, Welch CA, Lester MD, Emmett MK, Saville PD, Duerring SA.

 Cost-effectiveness strategies to treat osteoporosis in elderly women. 

8. Gold MR, Siegal JE, Russell LB, Weinstein MC.

Cost-Effectiveness in Health and Medicine

. New York: Oxford University Press; 1996.

 Hip fracture rates (820XX codes, closed only). 2001. The Healthcare Cost and Utilization Project. Sponsored by the Agency for Healthcare Research and Quality. 

http://www.hcup-us.ahrq.gov/db/nation/nis/NIS_Introduction_2001.jsp

12. McClung MR, Geusens P, Miller PD, et al.

 Effect of risedronate on the risk of hip fracture in elderly women. Hip Intervention Program Study Group. 

14. Black DM, Cummings SR, Karpf DB, et al.

 Randomised trial of effect of alendronate on risk of fracture in women with existing vertebral fractures. Fracture Intervention Trial Research Group. 

 Boniva product monograph. Nutley, NJ: Roche Laboratories Inc; 2003. 

18. Agency for Healthcare Research and Quality.

 Nationwide Inpatient Sample, Release 10. Healthcare Cost and Utilization Project. 2001. 

 Osteoporosis: review of the evidence for prevention, diagnosis and treatment and cost effectiveness analysis. 

23. Chrischilles E, Shireman T, Wallace R.

 Costs and health effects of osteoporotic fractures. 

25. Tosteson AN, Gabriel SE, Grove MR, Moncur MM, Kneeland TS, Melton LJ III.

 Impact of hip and vertebral fractures on quality-adjusted life years. 

27. US Department of Labor, Bureau of Labor Statistics.

 Consumer Price Index, U.S. City Averageâ€”Medical Care. 2006. 

29. Black DM, Schwartz AV, Ensrud KE, et al.

 Effects of continuing or stopping alendronate after 5 years of treatment: the Fracture Intervention Trial Long-term Extension (FLEX): a randomized trial. 

 National Prescription Audit. Plymouth Meeting, PA: IMS Health; 2005.

 A cost effectiveness analysis of calcium and vitamin D supplementation, etidronate, and alendronate in the prevention of vertebral fractures in women treated with glucocorticoids.

36. Schousboe JT, Ensrud KE, Nyman JA, Kane RL, Melton LJ III.

 Potential cost-effective use of spine radiographs to detect vertebral deformity and select osteopenic post-menopausal women for aminobisphosphonate therapy. 

38. Borgstrom F, Carlsson A, Sintonen H, et al.

 The cost-effectiveness of risedronate in the treatment of osteoporosis: an international perspective. 

40. Chesnut CH III, Skag A, Christiansen C, et al

. Effects of oral ibandronate administered daily or intermittently on fracture risk in postmenopausal osteoporosis. 

42. Barrett-Connor E, Grady D, Sashegyi A, et al.

 Raloxifene and cardiovascular events in osteoporotic postmenopausal women: four-year results from the MORE (Multiple Outcomes of Raloxifene Evaluation) randomized trial. 

44. Cauley JA, Norton L, Lippman ME, et al.

 Continued breast cancer risk reduction in postmenopausal women treated with raloxifene: 4-year results from the MORE trial. Multiple outcomes of raloxifene evaluation. 

46. Tosteson AN, Melton LJ III, Dawson-Hughes B, et al.

 Cost-effective osteoporosis treatment thresholds: the United States perspective. 

Articles

METHODS

Osteoporosis treatment for high-risk women is cost-effective, with bisphosphonates providing the most benefit at the lowest cost.

Therapies for Treatment of Osteoporosis in US Women: Cost-effectiveness and Budget Impact Considerations

: To evaluate whether lifetime future net tax revenues from an in vitro fertilization (IVF)-conceived child are substantial enough to warrant public subsidy relative to the mean IVF treatment costs required to obtain 1 live birth.

: Mathematical generational accounting model.

: The model estimates direct financial interactions between the IVF-conceived child and the government during the childâ€™s projected lifetime. In the model, we accrue IVF costs required to conceive the child to the government, and then we estimate future net tax revenue to the federal and state governments from this individual, offset by direct financial transfers from the government (eg, child allowances, education, Medicare, and Social Security). We discount lifetime costs and gross tax payments at Treasury Department rates to establish the present value of investing in IVF. We applied US Congressional Budget Office projected changes in tax rates over the course of the model. 

: An IVF-conceived child, average in every respect (eg, future earnings, healthcare consumption, and life expectancy), represents a net positive return to the government. Based on an average employed individual born in 2005, the projected net lifetime tax contribution is US $606,200. Taking into consideration IVF costs and all direct financial interactions, the net present value is US $155,870.

: Lifetime net taxes paid from a child relative to the childâ€™s initial IVF investment represent a 700% net return to the government in discounted US dollars from fully employed individuals. This suggests that removing barriers to IVF would have positive tax benefits for the government, notwithstanding its beneficial effect on overall economic growth.

Financial and legislative barriers to fertility treatments prevent many couples from achieving their desired family size, resulting in fewer children being born. 

The costs attributed to in vitro fertilization (IVF) treatment are insignificant in light of the lifetime net tax contributions of IVF-conceived children. 

Minimizing barriers to fertility treatments is likely to have long-term economic benefits

Based on the US GA model developed by Kotlikoff,

 a basic mathematical model was developed taking the perspective of the US government to estimate the discounted lifetime net tax contribution derived from a single individual. The model describes the financial position between the child and the government during the child&#8217;s projected lifetime. For comparison, the model estimates lifetime net tax contributions for a naturally conceived versus an IVF-conceived child, where the major cost difference is assumed to be IVF treatment costs and any extra costs related to the child&#8217;s care. In this model, we assign IVF costs to the government to assess the merits of funding such a policy. All direct government expenditures and tax contributions were discounted using Treasury Department rates.

Conceptually, there are 3 broad stages in lifetime financial interactions, each with differing components of the financial exchange, as follows: (1) early life, when the government primarily contributes resources to individuals through child tax credits, healthcare, and educational expenses; (2) employment, when individuals begin returning resources to the government through federal, state, and local taxes; and (3) retirement, when the government expends additional resources on Social Security and old-age programs. Two general models are estimated. The first model assumes that individuals graduate from high school and then follow the average higher education, employment, and unemployment trends (hereafter referred to as average employment). The second model assumes full-time education from ages 6 to 19 years, with full-time employment from age 20 years until  retirement at age 65 years (hereafter referred to as full employment). The models assume that successful IVF treatment results in a single live birth (with a mean life expectancy of 79 years) and that the child is identical to a naturally conceived individual.

 In all scenarios, the model includes hospital delivery costs, taking into consideration additional costs frequently accrued to IVF-conceived children attributed to low birth weight.

 Age-graded government expenses and tax contributions were assessed across a hypothetical individual&#8217;s lifetime to derive discounted lifetime net tax contributions using net present value (NPV) calculations and undiscounted lifetime net tax contributions. Following similar GA calculations used to assess US immigration policy, we consider various costs generated and taxes paid.

Revenues collected by the government include federal and state income tax (the national mean rate in this model), corporate tax, excise tax, Federal Insurance Contributions Act tax, Supplemental Medical Insurance contributions, federal retirement tax, property tax, and sales tax. To calculate the accounting models, age profiles of each expenditure and revenue component were identified from existing data sources. Because the models describe financial interactions across an individual&#8217;s lifetime, these age profiles are adjusted to account for depreciation of money over time through the application of a discount rate. The US Congressional Budget Office 2007 projections were used as the basis for estimates of inflation, individual earnings increases, tax rate increases, increases in Supplemental Medical Insurance revenue, and Medicare and Medicaid expenditures.

 Increases in expenditures on schooling are based on historical rates of increase.

 Beyond the period for which these long-term forecasts are available, we assume that particular components grow to keep pace with demographic and productivity growth. A discount rate of 4% was applied to lifetime tax revenue and transfer payments. The discount rate was compounded continuously.

. The mean IVF treatment costs to produce a live birth are considered herein as a further expense unique to an individual conceived using IVF. The national mean cost per IVF treatment cycle in 2003 is US $12,400.

 Cost per live birth is calculated as the mean cost per cycle divided by the age-adjusted probability of a live birth, where the treatment efficacy is known to vary primarily by the age of the mother (ie, lower success rates with older age) and by other factors.

 The age-adjusted cost per live birth is given in 

. Analyses are based primarily on 2 waves of the annual March Current Population Survey.

 The Current Population Survey is a monthly survey of about 50,000 households conducted by the Bureau of the Census for the Bureau of Labor Statistics and is the primary source of information on the labor force characteristics of the US population. The sample is representative of the civilian noninstitutional population. Estimated sales tax revenue is obtained from the Consumer Expenditure Survey, also conducted by the Bureau of Labor Statistics. Government expenditures on congestible goods are obtained from the US Statistical Abstract, following prior work.

 Educational expenditures are taken from the Digest of Education Statistics.

. Given the baseline assumptions for the age profiles of expenditures and revenues, the net financial exchange of an individual at any age is derived in the accounting models simply as the discounted sum of all the economic components up to that age. Specifically, lifetime individual NPV is the discounted sum of all revenues to the government at all ages minus expenditures at all ages as follows:

The projected lifetime net tax contribution trajectories for an average employment naturally conceived child and for an IVF-conceived child are shown in the 

. There is a net increase in government revenue by age 37 years for naturally conceived children versus by age 40 years for an IVF-conceived child. The additional costs attributed to conceiving an IVF child are shown as an increased cost at birth. In all simulations, the financial position between the child and the government changes as the child enters the workforce and again at the point of leaving full-time employment, with a net profit to the taxing authority. In vitro fertilization coverage represents a minor component of the net cost for creation of new taxpayers. 

The projected combined returns to the federal and state governments in lifetime net tax contributions from an IVFconceived child and from a naturally conceived child are given in Table 1. For a naturally conceived child, the mean discounted lifetime tax contribution amounts to US $292,285 for full employment; the projected net undiscounted lifetime tax contribution for a full employment individual born in 2005 is US $1,103,000 (

). For IVF-conceived children, the net undiscounted lifetime tax contribution is similar to that for naturally conceived children even after IVF investment costs are factored into the analysis.

The age at which the financial position between a naturally conceived child and an IVF-conceived child changes with respect to the taxing authority is given in Table 1. We have labeled this as the breakeven age to highlight the age at which the government has  recouped all direct financial transfers (discounted) in the early years to achieve working-age participation. Assuming full employment, the breakeven ages with net profit to the government are 34 years for naturally conceived children and 36 to 38 years for IVF children conceived to mothers younger than 41 years. The difference in the breakeven age for IVF-conceived children compared with naturally conceived children differs by 2 to 5 years depending on variations in increasing IVF investment costs attributed to older mothers.

The principal aim of this research was to assess the net tax contributions of an IVF-conceived child, with the intention of informing future policy directives that may influence access to fertility treatments. In reality, there are many good reasons to justify improved IVF access, including medical need, equity, and respect for an individual&#8217;s human right to a family, and reproductive rights. However, in many countries, including the United States, access to fertility care is limited, and many couples are unable to afford treatment. Based on the results shown herein, one could easily argue on the basis of economics that financial or legislative barriers to IVF treatments for infertile couples should be removed.

: We are grateful for comments and guidance provided by Dr Craig Currie from Cardiff Medical School.

: From Ferring International Center (MPC), St Prex, Switzerland; the RAND Corporation (MSP), Santa Monica, CA; the RAND Corporation (SH), Cambridge, England; the Fertility Centers of Illinois (BR K), Chicago; the Boston IVF and the Department of Obstetrics, Gynecology and Reproductive Biology, Harvard Medical School (SP O), Boston, MA; and the Infertility Center of St Louis at St Lukeâ€™s Hospital (SJS), St Louis, MO. 

: This study was funded by the not-for-profit RAND Corporation and by Ferring International Center.

: Mr Connolly is an employee of Ferring International Center, a funder of this study. Dr Pollard and Mr Hoorens report receiving grant funding from Ferring International Center. Dr Oskowitz is a board member of Columbia Labs, a manufacturer of hormones used in IVF treatment, and reports owning stock in that company. Drs Kaplan and Silber report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article. 

: Concept and design (MPC, MSP, SH, SJS); acquisition of data (MSP); analysis and interpretation of data (MPC, MSP, SPO, SJS); drafting of the manuscript (MPC, MSP, SH, BRK, SPO, SJS); critical revision of the manuscript for important intellectual content (MPC, MSP, SH, BRK, SPO, SJS); statistical analysis (MSP); obtaining funding (MPC, MSP, SH); administrative, technical, or logistic support (MPC); and supervision (MPC). 

: Sherman J. Silber, MD, Infertility Center of St Louis at St Lukeâ€™s Hospital, 224 S Woods Mill Rd, St Louis, MO 63017. E-mail: 

1. Andersen AN, Gianaroli L, Felberbaum R, de Mouzon J, Nygren KG; European IVF-Monitoring Programme (EIM) for the European Society of Human Reproduction and Embryology (ESHRE).

 Assisted reproductive technology in Europe, 2002: results generated from European registers by ESHRE. 

2. Andersen AN, Goossens V, Gianaroli L, Felberbaum R, de Mouzon J, Nygren KG.

 Assisted reproductive technology in Europe, 2003: results generated from European registers by ESHRE. 

4. National Institute for Health and Clinical Excellence.

 Clinical Guideline 11. Fertility: assessment and treatment for people with fertility problems. February 2004. 

http://www.nice.org.uk/Guidance/CG77/Guidance/pdf/English

 Current practices and controversies in assisted reproduction: report of a meeting on â€œMedical, Ethical and Social Aspects of Assisted Reproductionâ€ held at WHO headquarters in Geneva, Switzerland, 17-21 September 2001. 

http://www.who.int/reproductive-health/infertility/report.pdf

The Demographic Dividend: A New Perspective on the Economic Consequences of Population Change.

 Europeâ€™s declining population and the contribution of ART. 

. An international survey of the health economics of IVF and ICSI. 

. Utilization of infertility services: how much does money matter? 

16. Finkelstein BS, Silvers JB, Marrero U, Neuhauser D, Cuttler L.

 Insurance coverage, physician recommendations, and access to emerging treatments: growth hormone therapy for childhood short stature. 

18. Soumerai SB, McLaughlin TJ, Ross-Degnan D, Casteris CS, Bollini P.

 Effects of limiting Medicaid drug-reimbursement benefits on the use of psychotropic agents and acute mental health services by patients with schizophrenia. 

20. Cardarelli R, Sefton J, Kotlikoff LJ.

22. Bonduelle M, Liebaers I, Deketelaere V, et al.

 Neonatal data on a cohort of 2889 infants born after ICSI (1991-1999) and of 2995 infants born after IVF (1983-1999). 

. The current fiscal impact of immigrants and their descendants: beyond the immigrant household. In: Smith JP, Edmonston B, eds. 

The Immigration Debate: Studies on the Economic, Demographic, and Fiscal Effects of Immigration

. Washington, DC: National Academy Press; 1998:183-205.

28. MaCurdy T, Nechyba T, Bhattacharya J.

 An economic framework for assessing the fiscal impacts of immigration. In: Smith JP, Edmonston B, eds. 

. Washington, DC: National Academy Press; 1998:13-65.

 Digest of Education Statistics 2005 (NCES 2006-030). Washington, DC: National Center for Education Statistics, US Dept Education; 2006.

32. Centers for Disease Control and Prevention.

 Assisted Reproductive Technology Success Rates: National Summary and Fertility Clinic Report (2004). Atlanta, GA: Centers for Disease Control and Prevention, US Dept Health and Human Services; 2006.

34. Callahan TL, Hall JE, Ettner SL, Christiansen CL, Greene ME, Crowley WF.

 The economic impact of multiple-gestation pregnancies and the contribution of assisted-reproduction techniques to their incidence. 

 Diagnosis and management of female infertility. 

 Access to assisted reproduction services and infertility treatment in Belgium in the context of the European countries. 

. Explaining disparities in treatment seeking: the case of infertility. 

. Use of infertility services in the United States: 1995. 

44. Leunens L, Celestin-Westreich S, Bonduelle M, Liebaers I, Ponjaert-Kristoffersen I.

 Cognitive and motor development of 8-yearold children born after ICSI compared to spontaneously conceived children. 

 Effects of infertility insurance mandates on fertility. 

. Cheaper than a newcomer: on the social production of IVF policy in Israel. 

POLICY

We estimate the future net tax contributions from an IVF-conceived child to highlight that removing barriers to fertility treatments can have long-term economic benefits.

Long-term Economic Benefits Attributed to IVF-conceived Children: A Lifetime Tax Calculation

: Investigators from the Framingham Offspring Study (FOS) recently proposed a new simple point score for estimating 8-year diabetes mellitus (DM) risk.

: To validate prediction models and to compare DM risk estimated by the point score with observed DM incidence.

: Longitudinal observational cohort study.

: We identified 20,644 members of Kaiser Permanente Northwest (KPNW) who had no prior evidence of DM and who had all data elements necessary to estimate the models recorded between July 1, 1999, and June 30, 2000. Patients were followed up through June 30, 2007. We reestimated the FOS logistic regression models in the total KPNW sample (cumulative DM incidence, 16.5%) and in a randomly selected subsample with incidence identical to that in the FOS (5.1%). We also compared DM risk predicted by the FOS point score with actual 8-year DM incidence observed in the KPNW samples.

: The prediction models performed similarly in the FOS and KPNW samples, with almost identical odds ratios for independent variables and areas under the receiver operating characteristic curves for the models. The FOS point score substantially underestimated actual DM incidence in the total sample. In the subsample with DM incidence identical to that in the FOS sample, the point score was extremely accurate.

: The accuracy of the point score requires that the underlying incidence of the individuals to whom it is applied be similar to the population from which it was derived. Accurate adaptation requires recalculating DM incidence at each point level.

Diabetes mellitus (DM) prediction equations developed by the Framingham Offspring Study were validated, but the point score underestimated DM risk in a health maintenance organization population.

Diabetes mellitus incidence can be accurately predicted from a small set of known demographic and clinical risk factors that are easily measured in a clinical setting. 

Using the point score to estimate DM risk of individual patients must be done cautiously.

Although the prevalence and incidence of type 2 diabetes mellitus (DM) continue to increase at an alarming rate,

 many at-risk patients do not develop DM.

 Accurate prediction of DM onset may help target interventions to prevent DM and facilitate the efficient allocation of healthcare resources. A few DM prediction models have been developed,

 but none are widely used in clinical practice. Investigators from the Framingham Offspring Study (FOS)

 recently proposed a new simpler algorithm for estimating 8-year DM risk. The algorithm uses a point score derived from a small set of  variables easily obtainable at a clinic visit. The authors&#8217; methods followed those of the original Framingham Heart Study, which created a point score for predicting coronary heart disease (CHD).

 The now widespread acceptance and use of the FOS CHD point score coupled with the simplicity of the FOS DM risk score may lead to rapid adoption of the FOS DM risk score in clinical practice. However, as the FOS investigators recommend, the DM prediction models and the corresponding point score must first be validated in other population samples. We sought to replicate the FOS models using electronic medical records from a large integrated health plan and to compare the DM risk estimated by the point score with the observed DM incidence.

The baseline characteristics of 20,644 KPNW subjects are given in 

, along with the characteristics of 3140 persons in the FOS sample. The KPNW sample seemed to be somewhat older (mean age, 57.4 vs 54.0 years) and more obese (mean BMI [calculated as weight in kilograms divided by height in meters squared], 30.3 vs 27.1). Greater proportions of the KPNW sample were hypertensive (67.8% vs 44.2%) and had elevated triglyceride levels (41.8% vs 31.8%).

The FOS and KPNW personal models for DM prediction using only categories of age, sex, family history of DM, and BMI are given in 

. Age younger than 50 years and BM I less than 25.0 were the referent categories. Although ages 50 to 64 years and 65 years or older significantly increased the risk of DM in both samples, the odds ratios were much greater for these variables in the FOS sample. Male sex produced a similar odds ratio in both samples but was not statistically significant in the FOS sample. In both samples, BM I of 25.0 to 29.9 approximately doubled the risk of DM, while BM I of 30.0 or higher increased the risk by about 6 times. The FOS model seemed to yield somewhat better discrimination than the KPNW model (area under the ROC curve, 0.724 vs 0.676).

The calibration analysis of the simple clinical model using categorical variables is given in 

. Applying the FOS function directly to the full KPNW sample produced a high &#967;2 statistic, indicating poor calibration. Recalibrating the model substantially lowered the &#967;2 statistic, but the value was still highly significant. Although still lower, the &#967;2 statistic for the reestimated equation was also statistically significant. In all 3 applications, calibration was better among the KPNW subsample.

The FOS algorithm to estimate DM risk assigns 10 points for a fasting glucose level of 100 to 125 mg/dL, 5 points for an HDL-C level of less than 40 mg/dL in men or less than 50 mg/dL in women, 2 points for a BMI of 25.0 to 29.9, 5 points for a BM I of 30.0 or higher, 3 points for parental history of DM, 3 points for a triglyceride level of 150 mg/dL or higher, and 2 points for blood pressure of 130/85 mm Hg or higher or receiving treatment. The 8-year risk of type 2 DM as projected by the sum of the points as developed by the FOS is plotted in the 

, ranging from less than 3% among patients who score 10 or fewer points to greater than 35% among patients who score 25 or more points. Actual DM incidence for the total KPNW sample at each point level is plotted according to the points assigned from baseline values, ranging from 5% (<10 points) to almost 60% (>25 points). Actual DM incidence at each point level among the subset of the KPNW sample randomly selected to match the FOS proportion developing DM is also plotted, closely following the estimated risk of the FOS point score. 

In this observational study of the electronic medical records of 20,644 health maintenance organization members, we connfirmed the DM prediction equations developed by the FOS. In the simple clinical models, the odds ratios of the predictor variables were remarkably similar to those reported by the FOS, whether using categorical or continuous measures. The overall discrimination as measured by the area under the ROC curve was also similar. The validation of the FOS equations supports the conclusion that complex models are not needed to determine risk of type 2 DM. A set of clinical variables that is easy to collect (namely, BMI, the presence of hypertension, family history of DM, and HDL-C, triglyceride, and fasting glucose levels) produced an area under the ROC curve of 0.84.

However, we did not confirm the FOS point score predicting 8-year DM risk. Although we found that more points equated to greater risk, the accuracy of the point score risk  estimate seemed to be contingent on the overall incidence of DM in the population being similar to the 5.1% found in the FOS data. In our study sample, which had a much higher DM incidence of 16.5%, the point score dramatically underestimated individual risk. When we randomly selected a subset of patients who developed DM at the same 5.1% rate found in the FOS data, the expected risk predicted by the point score and the actual proportion developing DM at each point total were almost identical. This finding has important ramifications for the applicability of the simple point scores to higher risk populations. Applying the FOS point score to an individual patient could lead to severe overestimation or underestimation of the patient&#8217;s risk if the underlying mean incident  rate is substantially different from that in the FOS sample. However, because actual DM incidence increased with each point at a rate almost identical to that suggested by the FOS score, the number of points accurately represents increasing DM risk. Therefore, if managed care plans with data resources similar to those of KPNW plotted their actual DM incidence at each point level, the point score could be accurately adapted to their specific populations.

Race/ethnicity mix may explain some of the difference in the proportions of the FOS and KPNW samples that developed DM. The FOS sample was 99% white compared with approximately 90% of the KPNW sample. Selection bias likely explains much more of the difference. Although the FOS sample was population based, it was composed of volunteers, who may have been healthier than nonvolunteers. Because the KPNW sample was observational, we were limited to analyzing approximately 20% of KPNW patients for whom all the necessary clinical data were available. Among the total KPNW population aged 26 to 82 years (those with and without all the clinical measures), the 7-year incidence of DM was 6.5%. That 16.5% of our study sample experienced DM during the same 7 years suggests that these patients had the necessary measures because their clinicians may have been concerned about their DM risk. However, these are precisely the type of patients to which a clinician would likely apply the FOS point score. Although the FOS point score might accurately assess DM risk on a populationwide basis, it seems to substantially underestimate that risk among clinical attendees at all point levels.

In addition to the KPNW sample being composed of healthcare seekers, there were some specific differences between the KPNW and FOS samples that likely affected the underlying incidence of DM. In particular, the mean BM I of the KPNW sample was 30.3 compared with 27.1 in the FOS sample. Because obesity is one of the strongest risk factors for DM, it is not surprising that DM incidence was much greater in the KPNW sample. Nonetheless, if the predictive power of BM I is continuous, the FOS risk score will underestimate the effect of very high BMI because patients with a BMI of 30.0 receive the same 5 points as patients with a BM I of 35.0. This is an important consideration given the continuing trend of rising obesity prevalence.

 In addition, the KPNW sample had a much higher proportion of subjects with hypertension and high triglyceride levels, both well-known risk factors for DM incidence.

To our knowledge, this is the first study to examine the applicability of the FOS prediction equations of incident DM. However, it is not the first to find Framingham equations to be inaccurate. Previous investigations of the Framingham CHD risk equations have shown that underestimation of risk occurs in patients with DM.

 This is due to an underlying risk of CHD in DM that is greater than that in the general population on which the Framingham equations were based. When such systematic underestimation occurs, a process known as recalibration can be used, which replaces the Framingham mean predictor values and mean incident rate with the mean values and mean incident rate from the new cohort. Recalibration has been shown to correct overestimation or underestimation in ethnic groups with different underlying CHD risk but does not produce a better model fit than reestimating the equations with each site&#8217;s data.

 We similarly found that recalibration greatly improved the fit of the FOS function but not more so than reestimating risk in our data using identical variables.

The FOS is not the first study to attempt DM risk prediction. Most of those efforts have focused on predicting prevalent but undiagnosed DM.

 A simple tool that relied on questionnaire-based data was developed in Finland, but that risk score predicted DM medication initiation, not incident DM per se.

 Another longitudinal risk score using 4 variables (age, sex, and triglyceride and fasting glucose levels) has been proposed, but the area under the ROC curve for that score was 0.71, considerably lower than that in the FOS model. The San Antonio Heart Study investigators developed an incident DM prediction model similar to the FOS model in the variables used and in performance but did not propose a risk score.

 Given the simplicity of the FOS risk score and the widespread use of a similar score for predicting CHD risk, the FOS tool might be rapidly adopted into clinical practice. Although the FOS equations seem valid for predicting DM incidence, the point score did not accurately  predict incident DM in a population at much greater risk. Therefore, the accuracy of a point score requires that the underlying incidence of the individuals to whom it is applied be similar to the population from which it was derived. Because the clinician is unlikely to know whether this is indeed true, using the point score to estimate DM risk of individual patients must be done cautiously. Health plans that can recalculate the risk associated with each point level can accurately adapt the point score to predict DM risk in their populations.

: Kaiser Permanente Center for Health Research, Portland, OR (GAN, JBB).

: The authors (GAN, JBB) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

: Concept and design (GAN, JBB); acquisition of data (GAN); analysis and interpretation of data (GAN, JBB); drafting of the manuscript (GAN, JBB); critical revision of the manuscript for important intellectual content (JBB); statistical analysis.

Diabetes mellitus (DM) prediction equations developed by the Framingham Offspring Study were validated, but the point score underestimated DM risk in an HMO population.

September 2008

Therapies for Treatment of Osteoporosis in US Women: Cost-effectiveness and Budget Impact Considerations

Long-term Economic Benefits Attributed to IVF-conceived Children: A Lifetime Tax Calculation

Validating the Framingham Offspring Study Equations for Predicting Incident Diabetes Mellitus

Medicine Reimbursement Recommendations in Canada, Australia, and Scotland

Return on Investment for Bariatric Surgery

Metastatic Non–Small Cell Lung Cancer: Costs Associated With Disease Progression

A Study on the Economic Impact of Bariatric Surgery

Complementary and Alternative Medicine in Cancer Care

Utilization Outcomes and Patient Cost Sharing