Initial expenditures for contraceptive methods may belie their cost-effectiveness over time

The costs associated with unintended pregnancy make contraceptive use highly cost-effective, and the initial costs associated with using a particular method may not fully reflect the method's cost-effectiveness over time, according to a study of the clinical and economic impact of 15 contraceptive methods.(1) Voluntary sterilization, despite the large initial expenditure it requires, is extremely cost-effective over time, as are long-term temporary methods. Postcoital contraception, used in emergency situations (such as after unprotected intercourse or method failure), also yields considerable savings. For all methods, however, achieving maximal cost-effectiveness depends on proper use.

Study Design

For each of the methods studied (female and male sterilization, oral contraceptives, the implant, the injectable, hormonal and Copper T IUDs, the diaphragm, male and female condoms, the sponge, spermicides, the cervical cap, withdrawal and periodic abstinence), the investigators calculated the number of pregnancies that would be avoided if sexually active women of reproductive age used the method for up to five years.

To allow direct comparison of method costs, the researchers assumed that women would continue use throughout a given period, regardless of their experience with side effects or unintended pregnancy; for short-term methods, the investigators adjusted costs to reflect nonuse during an unintended pregnancy and resumption of use following the resolution of that pregnancy. For barrier methods, they assumed that users would engage in 83 acts of intercourse each year; they further assumed that diaphragms and cervical caps would have to be replaced after three years.

The first-year failure rates used in the analysis reflect typical (both perfect and imperfect) use. For most methods, the model applied the same failure rate for each of the five years; for the implant, the Copper T IUD, and male and female sterilization, however, it used a variety of failure rates for years 2-5, because the effectiveness of these methods changes over time.

The researchers estimated that pregnancy outcomes would be distributed as follows: Depending on the method, 24-47% of pregnancies would end in induced abortion, 6-12% would end in spontaneous abortion, 1-50% would be ectopic (50% among women who had been sterilized and 1-16% among users of other methods) and 20-40% would go to term. For all pregnancies that were not ectopic, the outcome proportions would be 48% for induced abortion, 12% for spontaneous abortion and 40% for term births.

In this analysis, a method's total cost comprised three types of direct medical costs--those related to acquiring and using the method (office visits, surgical procedures, drugs and devices, but not routine gynecologic care), to side effects and to unintended pregnancies (from conception through delivery and newborn hospitalization). The researchers evaluated these from the perspectives of both private and public payers. Although actual private and public payment mechanisms provide different coverage for contraception, the investigators assumed that both cover the full costs of all methods; this assumption, they explain, reflects societal costs and permits comparison of plans.

Data for the private payer model derive mainly from the MEDSTAT Systems MarketScan database, which includes information on insurers' claims payments from 45 major metropolitan areas; those for the public payer model are taken from MediCal, the California Medicaid program.

Direct Medical Costs

In the private payer model, the costs of acquiring contraceptives begin at just a few dollars for nonprescription methods. Costs approach $100 for the diaphragm and cervical cap, which need to be fitted by a clinician, and are several hundred dollars for methods that require insertion or administration by a physician. The cost of obtaining a female sterilization is close to $2,500. These costs are generally about 1.5-2.0 times those in the public payer model; for the Copper T and the male and female condoms, the differential is larger (2.5-3.0).

In general, the incidence of side effects associated with contraceptive use is less than 1%. However, some conditions associated with female sterilization, the implant, the injectable and IUDs occur among 1-3% of users, and the incidence of urinary tract infection and vaginitis reaches 15% among users of the diaphragm, sponge, spermicides and cervical cap.

The side effects with the least costly treatments are iron-deficiency anemia, irregular bleeding and various infections (less than $150 in both payment systems); cardiovascular conditions resulting from pill use entail the highest costs ($15,555-18,384 in the private payer model, and $3,898-8,374 in the public payer model). Several methods actually produce savings by preventing adverse health conditions. The researchers assumed that among pill users, the savings produced by the decreased risk of ovarian and endometrial cancers would offset the cost produced by the increased risk of cervical, breast and liver cancers.

Unintended pregnancy costs are substantial, ranging from $416 for an induced abortion to $8,619 for a term pregnancy in the private payer model and from $345 to $3,623 in the public payer model.

Relative Costs

For private payers, total first-year costs of contraception range from $285 for the injectable to $2,554 for female sterilization; public payers' costs vary from $188 for the Copper T to $1,238 for female sterilization (see Table 1). (Table 1 omitted) The investigators observe, however, that the components of these costs differ widely among methods, with the proportion of total costs attributable to unintended pregnancy declining as method effectiveness increases. For example, in the private payer model, the process of obtaining a female sterilization contributes 96% of the method's total five-year costs, whereas the resolution of unintended pregnancy represents 85% of the five-year costs of the male condom.

In the private payer model, the least costly methods after one year of use are the injectable, the pill and the hormonal IUD. After 2-3 years, however, the Copper T IUD, male sterilization and the implant become the least costly; these methods remain the least costly after five years. In the public payer model, the injectable and both IUDs are the least costly methods at one year; vasectomy, the implant and the Copper T cost the least at five years.

Over five years, a woman using no method will experience about four unintended pregnancies; the associated costs will total an estimated $14,663 in the private payer model and $6,490 in the public payer model. Use of any of the most highly effective methods (male and female sterilization, both IUDs, the implant, the injectable and the pill), meanwhile, will prevent about four pregnancies and save roughly $12,000-14,000 per woman in the private payer model. Even use of the least effective methods will produce substantial savings (for example, about $9,000 for the sponge or cervical cap, which the investigators project will prevent close to three pregnancies).

Results with Alternate Assumptions

In additional analyses performed for the private payer model, the researchers determined that varying the assumptions produced changes in the results. When they assumed that method users would engage in perfect use, cost savings associated with several less-effective methods grew; the increases in savings ranged from $2,415 for periodic abstinence to $2,661 for the female condom. Withdrawal became the second most cost-saving method, followed by male sterilization, the implant, the male condom and periodic abstinence.

When the researchers assumed that first-year failure rates would reflect typical use (except for permanent methods, the implant and the IUDs) and that rates in years 2-5 would reflect perfect use, the estimated savings for all methods increased. The Copper T, male sterilization and the implant remained the most cost-effective for the long term, however.

The use of the male condom in combination with any other method increases overall contraceptive effectiveness, but also adds to method acquisition costs. The impact of such combined method use was greatest for the sponge and the cervical cap, increasing cost savings by about one-half; for the most effective methods, cost savings declined because too few additional pregnancies were prevented to offset the cost of the condoms. Periodic abstinence and withdrawal became the least costly methods at five years.

Reliance on postcoital pills in the event of either imperfect use of nonhormonal methods or condom failure substantially reduces the risk of pregnancy; according: to the analysis, routine use of postcoital contraception would increase cost savings associated with the use of barrier methods by $1,123-2,470. Additionally, the assumption that the pill would be available through public health programs at a discounted price of $5 per cycle substantially reduced the five-year costs of oral contraceptives in the public payer model; the pill remained the fourth most cost-effective method, behind the Copper T, vasectomy and the implant.

Finally, when the investigators included costs of sexually transmitted diseases (STDs) in their analysis, the five-year costs associated with nonuse of contraceptives rose by $407 to a total of $15,070; at five years, the Copper T, the implant, male sterilization and the injectable were the least costly methods ($1,051-1,592). Barrier methods produced modest savings over nonuse ($183-283), and IUDs raised the costs (by $103-175). The researchers note that the cost-effectiveness of contraceptive methods with regard to STDs would have appeared greater if the analysis had been restricted to younger women, rather than including all women of reproductive age, because younger women are at a relatively high risk of STDs.

Conclusion

In discussing their findings, the investigators point out that since third-party payers typically cover pregnancy-related costs, they benefit from the savings generated by contraceptive use even though they often do not cover the costs of contraception. Therefore, if broader coverage improves access to birth control and increases the effective use of contraceptives, investment in such coverage will be to the advantage of third-party payers.

As the investigators note, a number of limitations of the study design affect interpretation of the results. First, by defining pregnancy-related costs as ending with newborn hospitalization, the investigators underestimated the societal costs of unintended pregnancies and thus the value of contraceptive use. Meanwhile, they overestimated the societal costs of unintended pregnancies ending in births by assuming that these births were unwanted rather than mistimed. Also, the assumption that nonectopic pregnancies are equally likely to end in births regardless of method used inflated the relative costs of less-effective methods.

According to the researchers, their analysis demonstrates that the initial costs of contraceptives inaccurately predict the methods' ultimate economic value because highly effective long-term methods require large up-front expenditures. Furthermore, they conclude, the substantial economic and social benefits generated by the use of any contraceptive justify investments by third-party payers, or by employers who pay insurance premiums, in broad coverage.

Reference

1. J. Trussell et al., "The Economic Value of Contraception: A Comparison of 15 Methods," American Journal of Public Health, 85:494-503, 1995.

Copyright The Alan Guttmacher Institute Jul 1995
Provided by ProQuest Information and Learning Company. All rights Reserved