Correspondence

Insurance Coverage and Outcomes of in Vitro Fertilization

N Engl J Med 2003; 348:958-959March 6, 2003

Article

To the Editor:

The analysis by Jain et al. (Aug. 29 issue)1 of insurance coverage for in vitro fertilization, based on data published by the Centers for Disease Control and Prevention (CDC),2 was limited by the use of aggregate-level data and the inability to control for confounders. We completed a similar analysis using the full data set, organized with each in vitro fertilization procedure as an observation.3 Like Jain et al., we focused on the same states with mandated coverage (Illinois, Massachusetts, and Rhode Island); we compared them with three states without insurance mandates (Indiana, Michigan, and New Jersey). The states without mandated insurance were chosen on the basis of geographic proximity to the states with mandated insurance coverage, the completeness of data on residency, the proportion of procedures performed in state residents, and the overall number of procedures performed. Differences between the states with coverage and the states without coverage were evaluated by chi-square tests and stratified, multivariable, logistic-regression models in which the following variables were controlled for: age, infertility diagnosis, previous in vitro fertilization procedures, previous live births, cryopreservation of nontransferred embryos (a proxy for the availability of embryos), and two procedural techniques for in vitro fertilization — use of intracytoplasmic sperm injection and assisted hatching.

The pattern of results was not consistent among the three states with coverage. Therefore, these states were compared separately with the three states without coverage, which were grouped for the comparisons. To enhance the comparability of results among states, we restricted analyses to patients 35 years of age or younger.

Although Jain et al. reported declines in the mean number of embryos transferred and the rate of gestations of triplets or higher-order multiples among states with coverage, our state-specific analysis revealed notable differences. We observed significant reductions in the number of embryos transferred, as measured by the proportion of procedures involving the transfer of three or more embryos, among procedures performed in Massachusetts (66.4 percent) and Rhode Island (73.5 percent) but found no reduction in Illinois (83.6 percent) as compared with the states without coverage (81.6 percent). Furthermore, although the rate of gestations of triplets or higher-order multiples was significantly lower in Massachusetts (8.7 percent), the rates in Illinois (12.9 percent) and Rhode Island (13.8 percent) were similar to those in the states without coverage (13.4 percent). After control for confounders, the pregnancy rate in Massachusetts was not significantly lower than the rates in the states without insurance (adjusted odds ratio, 0.9 [95 percent confidence interval, 0.8 to 1.0]).

Our findings indicate that mandated insurance coverage may affect embryo-transfer practices, but the heterogeneity of the findings in different states suggests that we must be cautious in speculating about whether and how these patterns in practices translate into health outcomes such as multiple births. As Guzick states in his editorial,4 “More work is needed to sharpen these statewide estimates and to quantify the benefits and costs of mandated coverage.”

Meredith A. Reynolds, Ph.D.
Laura A. Schieve, Ph.D.
Centers for Disease Control and Prevention, Atlanta, GA 30341-3717
mtr6@cdc.gov

Herbert B. Peterson, M.D., M.P.H.
World Health Organization, Geneva, Switzerland

4 References

1. 1

   Jain T, Harlow BL, Hornstein MD. Insurance coverage and outcomes of in vitro fertilization. N Engl J Med 2002;347:661-666
   Full Text | Web of Science | Medline

2. 2

   American Society for Reproductive Medicine, Society for Assisted Reproductive Technology, RESOLVE. 1998 Assisted reproductive technology success rates: national summary and fertility clinical reports. Atlanta: Centers for Disease Control and Prevention, December 2000.
   

3. 3

   Reynolds MA, Schieve LA, Jeng G, Peterson HB. Does insurance coverage decrease the risk for multiple births associated with assisted reproductive technology? Fertil Steril (in press).
   

4. 4

   Guzick DS. Should insurance coverage for in vitro fertilization be mandated? N Engl J Med 2002;347:686-688
   Full Text | Web of Science | Medline

Author/Editor Response

Dr. Reynolds and colleagues have reported results of their own analysis, which was based not only on the publicly disclosed data that we used, but also on more detailed information that was not publicly available and that allowed them to address potential confounders. However, rather than replicating our analysis of all 71,708 cycles of in vitro fertilization performed in 1998, with more complete control for confounding, Reynolds et al. selected a sample of 3 of the 39 states without mandated insurance coverage and limited their analysis to patients 35 years of age or younger — which unfortunately means that our results and theirs are not comparable. It is not clear how similar the 3 selected states without coverage are to the other 36 states without coverage. In addition, since the authors selected only a sample of such states, their analyses could well have been underpowered to detect differences in the number of embryos transferred and in the pregnancy rates. We agree that there could very well be heterogeneity in birth outcomes of in vitro fertilization according to state, irrespective of insurance status, but we believe that more information is needed to address this issue.

Finally, when we undertook this study, the CDC denied our request for access to the expanded Society for Assisted Reproductive Technology (SART)–CDC data base; now, Reynolds et al. suggest that our study was limited by the lack of the very data contained in that data base. We urge the CDC to make the expanded data base readily available to qualified researchers from SART, recalling that SART members not only collect and report these data to the CDC, but also finance their collection. We believe this step would facilitate epidemiologic research that could ultimately benefit infertile patients.

Tarun Jain, M.D.
Bernard L. Harlow, Ph.D.
Mark D. Hornstein, M.D.
Brigham and Women's Hospital, Boston, MA 02115
mhornstein@partners.org

Citing Articles (1)

Citing Articles

1. 1

   J. Ryan Martin, Jason G. Bromer, Denny Sakkas, Pasquale Patrizio. (2011) Insurance coverage and in vitro fertilization outcomes: a U.S. perspective. Fertility and Sterility 95:3, 964-969
   CrossRef