Cross-sectional data from the 2016 Navy Pregnancy and Parenthood Survey were used to calculate pregnancy and unintended pregnancy rates, the percentage of pregnancies that were unintended, and the percentage of unintended pregnancies that resulted in birth and abortion in the prior fiscal year among active-duty US Navy members aged 44 years or younger reporting female gender. The survey was conducted by the Naval Air Warfare Center Training Systems Division (NAWCTSD) Air Branch 4635 (Manpower and Personnel Studies).20 The data used for this analysis were obtained via Freedom of Information Act request from the Naval Air Warfare Center.
Survey Design, Sampling, and Data Collection
Data were collected from August to November 2016. The target population included a stratified random sample of 22,924 Navy women (38% of all active-duty Navy women) and 9,665 Navy men (4% of all active-duty Navy men) in pay grades E2-E9 and O1-O5. Participants received a survey invitation letter and two reminder letters at their command address and three reminder emails. The survey was conducted online. Overall, 4,802 participants reporting male or female gender completed the survey, with a 20% return rate for females after correcting the sample size for return-to-sender errors.20
The survey asked a common core set of questions to all participants on retention influencers, parenthood, family planning, sabbaticals, attitudes toward birth control and health care providers, and adoption leave. To maintain anonymity of respondents, the survey manager removed all login information from the data before analysis20 and this information was not included in our dataset.
We restricted our sample to participants of reproductive age, aged 44 years or younger, reporting female gender. Gender was assessed by the question, “What is your gender?” from which respondents could select “male” or “female.”
Our outcomes of interest were unintended pregnancy and unintended pregnancy that resulted in abortion in the prior fiscal year. Pregnancy in the prior fiscal year was measured by answering “yes” to the question, “Did you become pregnant between 1 October 2014 and 30 September 2015? (Do NOT count pregnancies that began before 1 October 2014 even though you were pregnant on that date.).” One additional participant was coded as having a pregnancy during the prior fiscal year as they reported a pregnancy in that time period in response to a question about how long ago their most recent pregnancy during Navy service occurred.
Unintended pregnancy in the prior fiscal year was measured by answering “yes” to the question, “Did you have an UNPLANNED pregnancy between 1 October 2014 and 30 September 2015? Note: For this survey, a planned pregnancy is one that you wanted at that time (i.e., you intentionally became pregnant.).”
Participants were also asked a series of questions about their most recent pregnancy since entering the Navy, including on the pregnancy outcome, whether they were deployed when they became pregnant, and whether they were using birth control at the time of the pregnancy and if not, why not. We restricted analyses of these variables to participants who reported an unintended pregnancy in the prior fiscal year and assumed that reporting on most recent pregnancy among these individuals referred to the pregnancy in the prior fiscal year.
The outcomes of participants’ unintended pregnancies in the prior fiscal year (abortion or birth) were assessed by the question, “What was the outcome of [your most recent] pregnancy?” We excluded participants who reported their unintended pregnancy outcome as currently pregnant, having had a miscarriage, or having had an ectopic pregnancy so as to include pregnancy outcomes determined by the participant. Participants were coded as having had an abortion if they reported their most recent pregnancy outcome as “abortion.” Participants were coded as having had a birth if they reported their most recent pregnancy outcome as “live birth (delivery after 36th week of gestation),” “premature birth (delivery between the 20th through 36th week of gestation),” or “stillbirth.”
Participants were asked demographic questions on their current age, marital status, and pay grade, as well as their deployment status when they became pregnant; race and ethnicity questions were not asked in the survey. Covariates were selected based on availability in the dataset, associations with unintended pregnancy found in prior research with servicemembers,8,21,22 and a priori interest in the impacts of deployment status on pregnancy outcomes. We coded participants’ pay grades as enlisted (E1-9) or officer (W2-5, O1-O7 or above); two respondents did not answer this question; however, a variable for Navy rate that was included in the dataset indicated they were in the enlisted pay grade and we coded them accordingly. Deployment at the time of unintended pregnancy was measured by answering “deployed” to the question, “Where was your [Navy] unit in the operational cycle when you [most recently] became pregnant?”
Descriptive statistics were calculated with Stata Statistical Software version 15.1 (StataCorp, College Station, TX). Missing data (nine participants who did not respond to the questions on prior pregnancy; one missing entry for two covariates; one missing entry for birth control use at time of most recent pregnancy) were excluded and the sample size for our outcomes of interest varied based on item response and survey skip patterns. The STROBE guidelines were used in reporting our study findings.23 This study was approved by the Allendale Investigational Review Board.
For the prior fiscal year overall and by background characteristics, we calculated the total number of pregnancies, the number of pregnancies that were unintended, and the percentage of pregnancies that were unintended. We divided the number of pregnancies and unintended pregnancies by the total sample population to obtain self-reported pregnancy and unintended pregnancy rates per 1,000 participants, by participants overall and by available background characteristics, including age group (18–24, 25–29, 30–44), relationship status (single, never married; divorced, widowed, separated; married), and pay grade (enlisted, officer); 95% confidence intervals (CIs) were also calculated for unintended pregnancy rates overall and by background characteristics.
We also calculated adjusted unintended pregnancy rates to account for underreporting of abortion. To do this, we employed methodology previously used to measure unintended pregnancy in the civilian population and the US military.22 Using this methodology, we increased the self-reported number of pregnancies reported by 11.9% (to mirror household survey findings that nearly half of abortions are underreported, reflecting 11.9% of the total number of reported pregnancies); 95% of these additional pregnancies were assumed to be unintended, as they represented induced abortions.22 We added these additional unintended pregnancies to the unintended pregnancies that were self-reported to calculate the adjusted unintended pregnancy rate.
Finally, we calculated the percentage of self-reported unintended pregnancies in the prior fiscal year that resulted in birth and abortion, along with 95% CIs for these overall and by background characteristics, including age group, relationship status, pay grade, and deployment status at the time of the pregnancy.