*BRFSS, Behavior Risk Surveillance System; CDC, Centers for Disease Control and Prevention; HLY; healthy life years; IUD, intrauterine device; LARC, long-acting reversible contraceptive; NSFG, National Survey of Family Growth; ONDIEH, Office of Noncommunicable Diseases, Injury and Environmental Health; ZAM, Zika virus–associated microcephaly.
†Authors’ calculation based on 2015 Puerto Rico total population size, 2015 birth rate, and adjusted contraception usage in 2002 BRFSS in Puerto Rico (online Technical Appendix Table, http://wwwnc.cdc.gov/EID/article/23/1/16-1322-Techapp1.pdf).
‡Less-effective methods include condoms, spermicides, fertility awareness methods, withdrawal, sponge, and diaphragm. Moderately effective methods include oral contraceptive pills, patches, vaginal rings, and injectable contraceptives. Highly effective method includes IUDs and subdermal implants. Dual use refers to the combination of moderately effective method use and male condoms. The contraception use distribution at baseline (without intervention) is based on adjusting the distribution reported in the 2002 BRFSS in Puerto Rico excluding women using permanent contraception by assuming 5 percentage points fewer women using no contraception than in 2002, which is based on a 36% reduction of birth rate among women 15–44 years of age in Puerto Rico from 2002 to 2015 (National Vital Statistics Report for 2002 and 2015 unpublished birth data from Puerto Rico), and the reported reasons for US teen pregnancy reduction (http://www.cdc.gov/nchhstp/newsroom/docs/factsheets/yrbs-fact-sheet-final-508.pdf); the decline in teen pregnancies was the fastest of any age group in Puerto Rico during 2010–2014, as reported in National Vital Statistics Reports. Data from Title X clinics in Puerto Rico also show that the percentage of women of reproductive age served in Title X clinics increased from 2.2% in 2006 to 10% in 2015. We also assume that new contraception users have the same contraception method distribution as contraception users as reported in the 2002 BRFSS survey (8). We assume 22% dual use among moderately effective method users at baseline (based on NSFG 2011–2013 data and unpublished analyses supplied by Karen Pazol, ONDIEH, CDC; we accounted for the effectiveness of dual-method use in preventing pregnancy but not for preventing sexual transmission of Zika virus). We calculated method-specific annual pregnancy rates by multiplying the failure rates of contraception methods under typical use by a calculated correction factor of 0.88 to adjust for the model over-predicting the number of unintended births in 2015 using the typical failure rates only. Multiplying method-specific contraceptive failure rates by numbers of women in each method category typically results in more predicted pregnancies and births than are actually observed, in part because of heterogeneity in sexual behaviors (19).
§Assuming 50% of no contraception users, 60% of less-effective contraception users, and 100% of moderately effective contraception users will visit a healthcare provider for contraception services under intervention. Based on the NSFG 2011–2013, among women of reproductive age who did not intend to become pregnant and not using permanent contraceptive methods, 21% of no contraception users, 33% less-effective method users, 97% moderately effective method users had at least 1 visit for contraception services in the last 12 months (personal communication, Karen Pazol, ONDIEH, CDC, 2016). Contraceptive services include receiving a birth control method or a prescription, receiving a checkup for birth control, receiving counseling about birth control, receiving a sterilizing operation, receiving counseling about a sterilizing operation, receiving emergency contraception, or receiving counseling about emergency contraception. We assume the intervention will increase the percentage of women visiting their provider for contraception counseling.
¶In the Contraceptive CHOICE project (9), 67% of participants who wished to avoid pregnancy chose to use highly effective methods, and 33% chose to use moderately effective methods. For the main scenario, we applied those estimates to 40% of the target population, assuming that 40% of unintended pregnancies are unwanted, and assumed that 20% of the remaining 60% of switchers would choose highly effective methods.
#Ellington et al. (5) estimated 180 cases (interquartile range 100–270 cases) of congenital microcephaly associated with Zika virus in Puerto Rico among an estimated 31,272 births (unpublished birth data, Puerto Rico Department of Health, 2015). On the basis of those estimates, we estimated the prevalence of congenital microcephaly: 58/10,000 births (interquartile range 32/10,000–86/10,000).
**The pregnancy loss rate in the US Zika Pregnancy Registry as of July 21, 2016, was 35% (14). The termination rate is calculated as the overall pregnancy loss rate minus the assumed stillbirth rate.
††A range of the main value ±20% was used to create upper and lower bounds used in sensitivity analyses.
‡‡Includes cost for devices and costs for 1 year of injections, pills, patches, rings, condoms, and the cost for related services in the first year.
§§Weighted average cost for hormonal IUD (50%, $659), copper IUD (25%, $598), and implant (25%, $659), based on Contraception CHOICE study (9) and commonly used devices in Puerto Rico.
¶¶Weighted average cost for generic contraceptive pill (78%, $370/y), injectable (14%, $240/y plus $130 for consultation for the year), and ring or patch (8%, $964/y), with the distribution of moderately effective methods based on NSFG, 2011–2013.
##Weighted average cost of moderately effective method plus cost of less-effective method.
***Male condom.
†††Including the cost for checking the placement of IUD in the first year of insertion.
‡‡‡Weighted average for vaginal and C-section delivery, including prenatal care, delivery, postpartum, and neonatal cost at delivery and in the first 3 months.
§§§Assumes additional costs related to repeated Zika virus testing by IgM (1–2 tests) for all pregnant women, 4 extra detailed ultrasound examinations, and 25% of women getting amniocentesis during all Zika-positive pregnancies (25% infection rate at baseline, range 10%–70%) (5), based on Oduyebo et al. (20).
¶¶¶Assumes 2 Zika virus tests (IgM and PCR) for serum and placenta, cranial ultrasound, and eye examination for all infants born to Zika virus–positive mothers based on Fleming-Dutra et al. (21).
###Three PCR tests for Zika virus using placenta, cord, and brain tissues of the fetus based on Martines et al. (22).
****Assuming all prenatal care cost, including extra cost of Zika virus–associated testing and monitoring during pregnancy and extra cost for Zika virus–associated testing for fetus.
††††Assuming half of the prenatal care cost, including extra cost of Zika virus–associated testing and monitoring during pregnancy and extra cost for Zika virus–associated testing for fetus.
‡‡‡‡Cost of prenatal care and delivery and extra cost of Zika virus–associated testing and monitoring during pregnancy and testing of infants for Zika virus.
§§§§Present value of cumulative medical and supportive care cost for infant with ZAM, discounted at 3% annually and taking into account mortality. Expenditures by employer-sponsored health plans for privately insured children with combined diagnoses of microcephaly and congenital cytomegalovirus enrolled during the first 4 years of life were used to project medical costs for cases of ZAM.