Objective
The purpose of this study was to assess obstetrician-gynecologists’ regarding their beliefs about when pregnancy begins and to measure characteristics that are associated with believing that pregnancy begins at implantation rather than at conception.
Study Design
We mailed a questionnaire to a stratified, random sample of 1800 practicing obstetrician-gynecologists in the United States. The outcome of interest was obstetrician-gynecologists’ views of when pregnancy begins. Response options were (1) at conception, (2) at implantation of the embryo, and (3) not sure. Primary predictors were religious affiliation, the importance of religion, and a moral objection to abortion.
Results
The response rate was 66% (1154/1760 physicians). One-half of US obstetrician-gynecologists (57%) believe pregnancy begins at conception. Fewer (28%) believe it begins at implantation, and 16% are not sure. In multivariable analysis, the consideration that religion is the most important thing in one’s life (odds ratio, 0.5; 95% confidence interval, 0.2–0.9) and an objection to abortion (odds ratio, 0.4; 95% confidence interval, 0.2–0.9) were associated independently and inversely with believing that pregnancy begins at implantation.
Conclusion
Obstetrician-gynecologists’ beliefs about when pregnancy begins appear to be shaped significantly by whether they object to abortion and by the importance of religion in their lives.
Two views on when pregnancy begins are prominent in the medical and lay literature. Since 1965, the American College of Obstetricians and Gynecologists (ACOG) has defined pregnancy as beginning with implantation of the embryo in the uterine wall. This definition is used also by the Guttmacher Institute, Planned Parenthood, and some textbooks. Defining pregnancy as beginning with implantation fits with certain facts: women who have regular periods generally do not consider themselves pregnant, regardless of whether ova were fertilized; human chorionic gonadotropin, which prevents menses and is the basis of the pregnancy test, is not produced before implantation ; and in vitro fertilization allows for fertilization without pregnancy. However, other books, organizations, physicians, and politicians define pregnancy as beginning with fertilization of the ovum by the sperm (often called conception ).
For Editors’ Commentary, see Table of Contents
This question of whether pregnancy begins with fertilization or with implantation has often been raised in public debates about whether some contraceptive technologies should be considered abortifacients. Intrauterine devices act in part by preventing implantation ; although hormonal contraceptives’ primary mechanism of action is to prevent ovulation, there is still debate about whether and how often they act secondarily by preventing implantation. Even the potential of preventing implantation concerns those who oppose abortion and consider that action abortifacient. Some investigators have suggested that the beginning of pregnancy was redefined to make the intrauterine device more acceptable to patients who are concerned that it could act as an abortifacient. Other experts have countered that, even if a technology such as the postcoital contraceptive or the intrauterine device did prevent implantation, this would not make it an abortifacient because it acts before pregnancy (defined as beginning with implantation). These disputes have fresh relevance now that the Food and Drug Administration has approved the “5-day pill” (ulipristal acetate), which is a prescription-only contraceptive that is effective when taken within 120 hours of unprotected sexual intercourse.
Despite this controversy, little is known about what practicing obstetrician-gynecologist (Ob/Gyn) physicians believe about when pregnancy begins. The Guttmacher Institute has stated, “On the question of when a woman is considered pregnant, the medical and scientific communities have long been clear: Pregnancy is established only when a fertilized egg has been implanted in the wall of a woman’s uterus.” To our knowledge, no studies have examined whether this issue has been settled in the minds of obstetrician-gynecologists, who are the physicians for whom the issue is most relevant. Only 1 previous study surveyed Ob/Gyn physicians, and it was limited to members of the Louisville Ob/Gyn society (n = 96). A more recent survey in 2 South Carolina family medicine clinics included patients (n = 178), but no physicians, and was geographically limited. We surveyed a national representative sample of Ob/Gyn physicians to assess their beliefs about when pregnancy begins. Because previous studies have found that physicians who are more religious are more likely to oppose abortion, we tested the hypothesis that believing pregnancy begins with implantation would be associated inversely with being religious and with objecting to abortion.
Materials and Methods
Study population
The methods for this study have been reported previously. Between October 2008 and January 2009, we mailed a confidential, self-administered, 12-page questionnaire to a national, stratified random sample of 1800 Ob/Gyn physicians who were ≤65 years old. The sample size was selected so that a 60% response would yield a margin of error of approximately 3%. The sample was drawn from the American Medical Association’s Physician Masterfile, a database intended to include all practicing US physicians. To increase minority representation (especially less-represented religions), we used validated ethnic surname lists to create 4 strata. We sampled 180 physicians with typical South Asian surnames, 225 physicians with typical Arabic surnames, 180 physicians with typical Jewish surnames, and 1215 other physicians (from all those whose surnames were not on 1 of these ethnic lists). Physicians received up to 3 separate mailings of the questionnaire. The first included a $20 bill, and the third offered an additional $30 for participation. Physicians also received an advance letter and a postcard reminder after the first questionnaire mailing. All data were double-keyed, cross-compared, and corrected against the original questionnaires. This study was approved by the University of Chicago institutional review board.
Survey instrument
The primary criterion measure asked physicians, “Which of the following statements comes closest to your beliefs about when pregnancy begins?” Response options were (1) at conception, (2) at implantation of the embryo, and (3) not sure.
Primary predictors were 2 religious measures and a measure of objection to abortion. Religious affiliation was categorized as none/no affiliation, Hindu, Jewish, Muslim, Catholic (included Roman Catholic [n = 237] and Eastern Orthodox [n = 25]), Evangelical Protestant, non-Evangelical Protestant, and other religion. The importance of religion was assessed with the question, “How important would you say your religion is in your own life?” Response options were (1) not at all important, (2) fairly important, (3) very important, and (4) the most important part of my life. As an indicator of objection to abortion, we asked physicians whether they morally object to abortion for a 22-year-old single woman who was 6 weeks’ pregnant after failed hormonal contraception. Physicians’ demographic characteristics (sex, race/ethnicity, age, and geographic region) were included as controls.
We did post-hoc analyses to examine whether beliefs about pregnancy were associated with 3 additional areas of controversy. We asked physicians whether they have a moral/ethical objection to abortion for a 36-year-old woman in the first trimester of pregnancy who needs radiation and chemotherapy for newly diagnosed breast cancer. We asked whether physicians have a moral/ethical objection to intrauterine devices. We asked physicians about their practices regarding emergency hormonal contraception and whether they offer it “to all women they believe are at risk for unplanned pregnancy” or do not offer it to all women (offer it only to women who say they have had unprotected intercourse, only to victims of sexual assault, or to nobody under any circumstances).
Data management and analysis
Case weights were incorporated into our statistical analyses to account for the oversampling strategy (the design weight) and to correct for differences in response rates among the surname categories and between US and foreign medical school graduates (the poststratification adjustment weight). Weights were the inverse probability of a person with the relevant characteristic being in the final dataset. The final weight for each case/respondent was the product of the design weight and the poststratification adjustment weight. This method of case weighting, which is used widely in population-based research, enabled us to adjust for sample stratification and variable response rates to generate estimates for the population of US Ob/Gyn physicians. We used the chi-square test to examine associations between each background variable and physicians’ beliefs about when pregnancy begins. We then conducted multivariable logistic regression to examine the relationship between the belief that pregnancy begins at implantation (rather than at conception, after excluding 195 respondents who indicated they were “not sure”) and (model 1) religious affiliation, (model 2) the importance of religion, (model 3) objections to abortion, and (model 4) demographic characteristics. We estimated the 4 models beginning with model 1 as predictors and adding models 2, 3, and 4 in sequence, observing changes in coefficients across models. The sample size was kept at 900 for all of the models. This is the size of the sample for model 4, which was the model with the most variables and consequently with the most cases dropped because of missing values. All analyses were performed with the survey-design–adjusted commands of Stata SE statistical software (version 11.0; Stata Corporation, College Station, TX).
Results
The response rate was 66% (1154/1760 questionnaires) after the exclusion of 40 potential respondents who were retired or who could not be located after 2 attempts to obtain a valid address. The response rate varied by stratum: 54% among those with Arabic surnames (120/221), 61% among those with South Asian surnames (107/175), 68% among those with Jewish surnames (120/176), and 68% in the primary sample (807/1188). Graduates of foreign medical schools were less likely to respond than graduates of US medical schools (58% compared with 68%; P = .001). The response rate did not differ significantly by age, sex, region, or board certification. Respondents’ demographic characteristics are reported in Table 1 . The sampling method increased the number of Muslim, Jewish, and Hindu respondents. However, the application of case weights to the analyses adjusted for this increase. Thus, our estimates of the percentages of US Ob/Gyn physicians in each racial category and in each religious affiliation were similar to our previous national survey that did not oversample by ethnic surname.
