We derived cost estimates in a similar fashion to the probability estimates, but additionally queried local and national hospital and insurance data (Table 3). Of note, to calculate the cost of a preterm delivery, we used a weighted average based on the distribution of timing of preterm delivery in the United States. We adjusted all costs to reflect 2014 U.S. dollars. The costs accounted for in the model included the cost of testing, maternal follow-up and treatment, penicillin desensitization, maternal delivery, and neonatal care. In this model, we did not include long-term care costs.
Data regarding syphilis seroconversion during pregnancy are limited to two studies. These studies reported incidence of syphilis seroconversion between 0% and 0.17%.9,109,10 Shiber et al identified an average incidence over a 17-year period (1993–2009) of 0.029%. In the first year of the study, there was an incidence of seroconversion of 0.17%. However, in the 16,423 women who delivered from 2005 to 2009, no seroconversions were noted. In the only other study evaluating seroconversion in pregnancy, there were no seroconversions in the 1,377 women rescreened in the third trimester between 2004 and 2006.9 To assign a base case, we used the final 9-year period evaluated by Shiber et al (2000–2009), which demonstrated an incidence of syphilis seroconversion in pregnancy of 0.012%.10 This is likely higher than is found in the U.S. population in general, but by using this base case, we intentionally biased our decision analysis toward a best case scenario for screening.
Our other major assumptions included all women are screened for syphilis early in pregnancy, therefore assuming 100% compliance with the screening recommendations. All women who rescreened positive received adequate and timely treatment (more than 4 weeks before delivery) with one dose of 2.4 million units of benzathine penicillin G.11,1211,12 Because all syphilis infections detected in women by this antenatal syphilis screening program received adequate and timely treatment, their neonates required only serologic follow-up.13 All women were identified as having syphilis in the third trimester, and therefore had a fetus beyond the threshold of viability and were treated and then monitored in the hospital to assess for Jarisch-Herxheimer reaction, but not necessarily admitted. In practice, treatment failure is more likely when ultrasound abnormalities are present, but the exact risk is imprecise.11,1411,14 Therefore, all treatments were assumed to be successful, but in sensitivity analysis, this was allowed to vary to 50% treatment failure in the setting of an abnormal ultrasound finding. A Jarisch-Herxheimer reaction was assumed to result in a 24-hour maternal hospitalization but to have no effect on preterm delivery or intrauterine fetal demise.11,1411,14 The effect of a Jarisch-Herxheimer reaction on preterm delivery and intrauterine fetal demise is also unclear, so this risk was also initially assumed to be negligible but varied in a sensitivity analysis to up to doubling the probabilities of both preterm delivery and intrauterine fetal demise. We assumed that all women with an anaphylactic allergy to penicillin were appropriately identified. We did not account for a risk for anaphylaxis given its rarity (1/5,000–10,000 courses of penicillin therapy)15–1815–1815–1815–18 but accounted for the cost of penicillin desensitization in the women with a reported allergy to penicillin who received penicillin for treatment of syphilis.
The primary outcome was the cost to prevent one case of congenital syphilis. Secondary outcomes included the cost to prevent one intrauterine fetal demise or neonatal death, the number of women who would need to be rescreened in the third trimester to prevent one of these adverse outcomes, and the likelihood that screening would be cost-effective under the plausible probabilities included in the model. We assumed that screening was cost-effective if it cost less than $285,000 to prevent one case of congenital syphilis. We chose this cost-effectiveness threshold because the average lifetime direct medical cost (accounting for 5% discounting) to care for an individual with a severe disability, like congenital syphilis, is approximately $285,000,19 thus making the cost to prevent the disease equivalent to the expenditures to care for a child born with the disease. Because we did not directly include long-term care costs in the model, this allowed us to relate the outcome of the model to overall health care costs. In sensitivity analysis, we allowed the cost-effectiveness threshold to vary up to $1,000,000 because this may better account for the direct and indirect costs as well as differences in discounting of long-term care of a severe disability.19,2019,20
In addition to the base case analysis, we performed one-way and two-way sensitivity analyses. In particular, this allowed for an investigation into how the incidence of seroconversion affected the outcomes of interest. Monte Carlo simulation (a computational algorithm that relies on repeated random sampling of all variables across their CIs based on their distributions) was also used. The distributions used in the Monte Carlo simulation for the probabilities were triangular or normal as appropriate based on available data. Gamma distributions were used for all cost variables. Given the plausible variation in all of the probabilities and cost estimates included in the model, no variable was excluded from the Monte Carlo analysis. Ten thousand simulations were run to estimate the percentage of time that universal third-trimester rescreening would be cost-effective as compared with no rescreening. This number of simulations was chosen because it is our routine practice when performing these analysis. We performed all analyses using TreeAge Pro 2013 Suite. The study did not involve human participants and was exempt from institutional review board approval.
The results of the base case model are presented in Table 4. The base case model demonstrated that at a rate of syphilis seroconversion of 0.012%, the cost to prevent one case of congenital syphilis with universal third-trimester syphilis rescreening was $419,842. Prevention of one intrauterine fetal demise cost $3,621,144 and one neonatal death cost $6,052,534. Under the base case assumptions, by rescreening 4,000,000 pregnant women (the annual pregnant population in the United States) in the third trimester, 60 cases of congenital syphilis, seven intrauterine fetal demises, and four neonatal deaths would be prevented. Finally, to prevent one case of congenital syphilis, 65,790 women would need to be rescreened. Similarly, 568,182 would need to be rescreened to prevent one intrauterine fetal demise and 952,381 to prevent one neonatal death resulting from maternal syphilis.
In one-way sensitivity analyses, the model was most sensitive to the incidence of syphilis seroconversion (Table 5). If the incidence of seroconversion was greater than 0.017%, universal third-trimester screening became cost-effective. If the incidence of seroconversion was greater than 0.2%, universal third-trimester screening became cost-saving. However, if the incidence was low (ie, 0.001%), it would cost more than $5 million to prevent one case of congenital syphilis. Additionally, the model was highly sensitive to the neonatal cost of preterm delivery and the probability of congenital syphilis given untreated maternal syphilis. If the cost of preterm delivery reached $75,000 on average (or approximately the cost of a 31- to 32-week neonate21), universal third-trimester rescreening became cost-effective. However, when delivered near term, at a cost of $3,328, the overall cost to prevent one case of congenital syphilis became $443,077. Similarly, at the highest probability of congenital syphilis without treatment, 29%, prevention of one case of congenital syphilis would cost $218,528 and would therefore be cost-effective. However, at any probability less than 22.5%, rescreening would not be cost-effective. The model was relatively refractory to changes in other variables. For example, at a relative risk of preterm delivery given untreated maternal syphilis of 9.09, the cost to prevent one case of congenital syphilis would be $391,135.
Two-way sensitivity analysis was then performed to evaluate the interaction between the two most influential variables, the incidence of syphilis seroconversion and the neonatal cost of a preterm delivery (Fig. 1). As the incidence of seroconversion approached 0.017%, universal screening became cost-effective at a lower overall preterm delivery cost (practically speaking, a greater gestational age at birth).
Using Monte Carlo Simulation, rescreening is cost-saving in only 3% of simulations (Fig. 2). However, in an additional 84.7% of simulations, rescreening costs less than $285,000 to prevent one case of congenital syphilis. At a minimum cost-effectiveness threshold of $100,000 to prevent one case of congenital syphilis, only 7.6% of simulations would be cost-effective or cost-saving. However, at a maximum threshold of $1,000,000, which at least partially accounts for the indirect costs of caring for a child with multiple comorbidities, 99.9% of simulations would be cost-effective or cost-saving (Appendix 2, available online at http://links.lww.com/AOG/A669).
This model shows that in the United States, universal syphilis rescreening in the third trimester in women who screen negative at their initial prenatal visit requires a large number of women be screened at a high health care cost to prevent one adverse outcome from maternal syphilis. Unfortunately, data regarding syphilis seroconversion during pregnancy are limited to the two aforementioned studies.
An alternative to using the limited information regarding seroconversion during pregnancy is using the overall rate of primary and secondary syphilis in women in the United States as reported by the CDC. This incidence ranged from zero to 5.7 cases per 100,000 by state or area (0–0.0057%) in 2013.1 Notably, the national prevalence for women was reported to be 0.9 per 100,000 in 2013 (Table 6).1 If policy were then to be based on state-level prevalence estimates, even the highest risk states or areas would require a very large number of women be screened at a high health care cost (Table 6). If we assume that most communities fall near the national average of primary and secondary syphilis rates in women (0.0009%), this is 19 times lower than the incidence of seroconversion that would make universal third-trimester screening cost-effective, costing more than $5,000,000 per case of congenital syphilis prevented. Therefore, given that the majority of the country has an incidence at or below 0.0009% (because it is reasonable to assume that pregnant women represent only a small portion of these prevalence estimates, and therefore would have a lower incidence), universal third-trimester screening should not be recommended. However, it remains reasonable for individual high-risk cities and communities to evaluate the incidence in their population on an ongoing basis and implement a program of routine syphilis rescreening with a sufficiently high incidence of seroconversion.
Strengths of this study are the ability to evaluate large ranges of costs and probabilities. Additionally, a unique characteristic of this study is that we were able to evaluate the CDC recommendation for third-trimester screening from a cost perspective.
The major limitation of all cost analyses is that the evaluation is only as good as the available data. In this model in particular, we made use of data from outside of the United States and extrapolated them to our population. Additionally, there is currently clinical uncertainty regarding the optimal treatment regimen for syphilis in pregnancy12 and the actual relative risks of different clinical scenarios on both the probability of preterm delivery and the probability of a neonate having congenital syphilis. We did vary these relative risks in our sensitivity analyses to account for this clinical uncertainty. We did not directly account for lifetime costs within the model. However, we set the cost-effectiveness threshold to lifetime costs so that screening would be a net-neutral health care expenditure. Finally, we did not account for indirect costs such as loss of work.
Even with a recommendation for universal syphilis screening in the first and possibly third trimesters in the United States, only 52% of mothers of neonates with congenital syphilis reported having at least one prenatal visit.22 Our analysis therefore assumes a best-case scenario regarding compliance with initial screening, and, although universal syphilis rescreening would prevent a small number of cases of congenital syphilis, interventions to improve access to health care may have a greater effect on reduction of the rate of congenital syphilis in the United States.
1. Patton ME, Su JR, Nelson R, Weinstock H; Centers for Disease Control and Prevention (CDC). Primary and secondary syphilis—United States, 2005–2013. MMWR Morb Mortal Wkly Rep 2014;63:402–6.
2. Alexander JM, Sheffield JS, Sanchez PJ, Mayfield J, Wendel GD Jr. Efficacy of treatment for syphilis in pregnancy. Obstet Gynecol 1999;93:5–8.
3. Webster LA, Rolfs RT. Surveillance for primary and secondary syphilis—United States, 1991. MMWR CDC Surveill Summ 1993;42:13–9.
4. Stray-Pedersen B. Economic evaluation of maternal screening to prevent congenital syphilis. Sex Transm Dis 1983;10:167–72.
5. Abyad A. Cost-effectiveness of antenatal screening for syphilis. Health Care Women Int 1995;16:323–8.
6. Connor N, Roberts J, Nicoll A. Strategic options for antenatal screening for syphilis in the United Kingdom: a cost effectiveness analysis. J Med Screen 2000;7:7–13.
7. Wolff T, Shelton E, Sessions C, Miller T. Screening for syphilis infection in pregnant women: evidence for the U.S. Preventive Services Task Force reaffirmation recommendation statement. Ann Intern Med 2009;150:710–6.
8. Kahn JG, Jiwani A, Gomez GB, Hawkes SJ, Chesson HW, Broutet N, et al.. The cost and cost-effectiveness of scaling up screening and treatment of syphilis in pregnancy: a model. PLoS One 2014;9:e87510.
9. Edwards RK, Bennett M, Langstraat C, Greene D. Assessment of the value of rescreening for syphilis in the third trimester of pregnancy. Infect Dis Obstet Gynecol 2006;2006:56504.
10. Shiber L, Todia WJ. Cost and clinical utility of repeated syphilis screening in the third trimester in a high-risk population. Am J Obstet Gynecol 2014;210:267.e1–5.
11. Wendel GD Jr, Sheffield JS, Hollier LM, Hill JB, Ramsey PS, Sánchez PJ. Treatment of syphilis in pregnancy and prevention of congenital syphilis. Clin Infect Dis 2002;35(suppl 2):S200–9.
12. Walker GJ. Antibiotics for syphilis diagnosed during pregnancy. The Cochrane Database of Systematic Reviews 2001, Issue 3. Art. No.: CD001143. DOI: 10.1002/14651858.CD001143.
13. Centers for Disease Control and Prevention (CDC). Congenital syphilis—United States, 2000. MMWR Morb Mortal Wkly Rep 2001;50:573–7.
14. Sheffield JS, Sánchez PJ, Morris G, Maberry M, Zeray F, McIntire DD, et al.. Congenital syphilis after maternal treatment for syphilis during pregnancy. Am J Obstet Gynecol 2002;186:569–73.
15. Johannes CB, Ziyadeh N, Seeger JD, Tucker E, Reiter C, Faich G. Incidence of allergic reactions associated with antibacterial use in a large, managed care organisation. Drug Saf 2007;30:705–13.
16. Galvao TF, Silva MT, Serruya SJ, Newman LM, Klausner JD, Pereira MG, et al.. Safety of benzathine penicillin for preventing congenital syphilis: a systematic review. PLoS One 2013;8:e56463.
17. Gruchalla RS, Pirmohamed M. Clinical practice. Antibiotic allergy. N Engl J Med 2006;354:601–9.
18. Rudolph AH, Price EV. Penicillin reactions among patients in venereal disease clinics. A national survey. JAMA 1973;223:499–501.
19. Waitzman NJ, Romano PS, Scheffler RM. Estimates of the economic costs of birth defects. Inquiry 1994;31:188–205.
20. Centers for Disease Control and Prevention (CDC). Economic costs associated with mental retardation, cerebral palsy, hearing loss, and vision impairment—United States, 2003. MMWR Morb Mortal Wkly Rep 2004;53:57–9.
21. Phibbs CS, Schmitt SK. Estimates of the cost and length of stay changes that can be attributed to one-week increases in gestational age for premature infants. Early Hum Dev 2006;82:85–95.
22. Genç M, Ledger WJ. Syphilis in pregnancy. Sex Transm Infect 2000;76:73–9.
23. Larsen SA, Steiner BM, Rudolph AH. Laboratory diagnosis and interpretation of tests for syphilis. Clin Microbiol Rev 1995;8:1–21.
24. Hollier LM, Harstad TW, Sanchez PJ, Twickler DM, Wendel GD Jr. Fetal syphilis: clinical and laboratory characteristics. Obstet Gynecol 2001;97:947–53.
25. Myles TD, Elam G, Park-Hwang E, Nguyen T. The Jarisch-Herxheimer reaction and fetal monitoring changes in pregnant women treated for syphilis. Obstet Gynecol 1998;92:859–64.
26. Klein VR, Cox SM, Mitchell MD, Wendel GD Jr. The Jarisch-Herxheimer reaction complicating syphilotherapy in pregnancy. Obstet Gynecol 1990;75:375–80.
27. Apter AJ, Kinman JL, Bilker WB, Herlim M, Margolis DJ, Lautenbach E, et al.. Represcription of penicillin after allergic-like events. J Allergy Clin Immunol 2004;113:764–70.
28. Chang C, Mahmood MM, Teuber SS, Gershwin ME. Overview of penicillin allergy. Clin Rev Allergy Immunol 2012;43:84–97.
29. Blencowe H, Cousens S, Kamb M, Berman S, Lawn JE. Lives Saved Tool supplement detection and treatment of syphilis in pregnancy to reduce syphilis related stillbirths and neonatal mortality. BMC Public Health 2011;11(suppl 3):S9.
30. Gomez GB, Kamb ML, Newman LM, Mark J, Broutet N, Hawkes SJ. Untreated maternal syphilis and adverse outcomes of pregnancy: a systematic review and meta-analysis. Bull World Health Organ 2013;91:217–26.
31. Martin JA, Hamilton BE, Osterman MJ, Curtin SC, Matthews TJ. Births: final data for 2013. Natl Vital Stat Rep 2015;64:1–65.
32. Owusu-Edusei K Jr, Gift TL, Ballard RC. Cost-effectiveness of a dual non-treponemal/treponemal syphilis point-of-care test to prevent adverse pregnancy outcomes in sub-Saharan Africa. Sex Transm Dis 2011;38:997–1003.
33. Owusu-Edusei K Jr, Peterman TA, Ballard RC. Serologic testing for syphilis in the United States: a cost-effectiveness analysis of two screening algorithms. Sex Transm Dis 2011;38:1–7.
34. Hawk AF, Chang EY, Shields SM, Simpson KN. Costs and clinical outcomes of noninvasive fetal RhD typing for targeted prophylaxis. Obstet Gynecol 2013;122:579–85.
35. Bivins HA Jr, McCallum OJ, Roegge DD. The mother's house: a new concept in antepartum care. Am J Obstet Gynecol 2007;197:201.e1–3.
36. Blandford JM, Gift TL. The cost-effectiveness of single-dose azithromycin for treatment of incubating syphilis. Sex Transm Dis 2003;30:502–8.
37. Chisholm CA, Katz VL, McDonald TL, Bowes WA Jr. Penicillin desensitization in the treatment of syphilis during pregnancy. Am J Perinatol 1997;14:553–4.
38. Owusu-Edusei K Jr, Chesson HW, Gift TL, Tao G, Mahajan R, Ocfemia MC, et al.. The estimated direct medical cost of selected sexually transmitted infections in the United States, 2008. Sex Transm Dis 2013;40:197–201.
39. Chesson HW, Gift TL, Pulver AL. The economic value of reductions in gonorrhea and syphilis incidence in the United States, 1990–2003. Prev Med 2006;43:411–5.
40. Gilbert WM, Nesbitt TS, Danielsen B. The cost of prematurity: quantification by gestational age and birth weight. Obstet Gynecol 2003;102:488–92.
41. Gold KJ, Sen A, Xu X. Hospital costs associated with stillbirth delivery. Matern Child Health J 2013;17:1835–41.
Supplemental Digital Content
© 2015 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.