Investigation of a transient increase in omphalocele prevalence in a birth cohort of TRICARE beneficiaries

BACKGROUND

The Department of Defense Birth and Infant Health Research (BIHR) program leverages medical encounter data to conduct birth defect surveillance among infants born to military families. Omphalocele is a major abdominal wall defect with an annual prevalence of ~2 per 10,000 births in BIHR data, but an unexpected increase was observed during 2017-2019, reaching 6.4 per 10,000 births in 2018. To investigate this transient increase in prevalence, this study aimed to validate the omphalocele case algorithm among infants born 2016-2021.

METHODS

Omphalocele cases were identified by ICD-10 code Q79.2 (exomphalos) on one inpatient or two outpatient infant encounter records and validated using parental and infant electronic health records. Characteristics of true and false positive cases were assessed using bivariate analyses and compared over time.

RESULTS

Of 638,905 live births from 2016 to 2021, 230 met the ICD-10 case definition for omphalocele; 138 (60.0%) cases were eligible for validation, of which 68 (49.3%) were true positives. The geometric mean time from birth to first ICD-10 omphalocele diagnosis was 1.1 (standard error [SE] 0.1) days for true positives and 11.9 (SE 3.1) days for false positives. Among the 70 false positives, 36 (51.4%) were cases of confirmed umbilical hernia; rates of umbilical hernia and delayed omphalocele diagnoses (>30 days after birth) were elevated among false positives during 2017-2019.

CONCLUSIONS

Higher misuse of ICD-10 code Q79.2 during 2017-2019 likely influenced the associated increase in omphalocele prevalence. Timing of diagnosis should be considered for omphalocele case definitions using medical encounter data.

Severe Maternal Morbidity among Women in the U. S. Military, 2003-2015

OBJECTIVE

 This study aimed to assess trends and correlates of severe maternal morbidity at delivery among active duty women in the U.S. military, all of whom are guaranteed health care and full employment.

STUDY DESIGN

 Linked military personnel and medical encounter data from the Department of Defense Birth and Infant Health Research program were used to identify a cohort of delivery hospitalizations among active duty military women from January 2003 through August 2015. Cases of severe maternal morbidity were identified by applying 21- and 20-condition algorithms (with and without blood transfusion) developed by the Centers for Disease Control and Prevention. Rates (per 10,000 delivery hospitalizations) were reported overall and by specific condition. Multivariable Poisson regression models estimated associations with demographic, clinical, and military characteristics.

RESULTS

 Overall, 187,063 hospitalizations for live births were included for analyses. The overall 21- and 20-condition severe maternal morbidity rates were 111.7 (n = 2089) and 37.4 (n = 699) per 10,000 delivery hospitalizations, respectively. The 21-condition rate increased by 184% from 2003 to 2015; the 20-condition rate increased by 40%. Compared with non-Hispanic White women, the adjusted 21-condition rate of severe maternal morbidity was higher for Hispanic (adjusted rate ratio [aRR] = 1.28, 95% confidence interval [CI]: 1.13-1.46), non-Hispanic Black (aRR = 1.34, 95% CI: 1.21-1.49), Asian/Pacific Islander (aRR = 1.35, 95% CI: 1.13-1.61), and American Indian/Alaska Native (aRR = 1.39, 95% CI: 1.06-1.82) women. Rates also varied by age, clinical factors, and deployment history.

CONCLUSION

 Active duty U.S. military women experienced an increase in severe maternal morbidity from 2003 to 2015 that followed national trends, despite protective factors such as stable employment and universal health care. Similar to other populations, military women of color were at higher risk for severe maternal morbidity relative to non-Hispanic White military women. Continued surveillance and further investigation into maternal health outcomes are critical for identifying areas of improvement in the Military Health System.

KEY POINTS

· Cesarean delivery and multiple birth were the strongest correlates of severe maternal morbidity in this population.. · Racial disparities persisted across indicators of severe maternal morbidity.. · Rates of disseminated intravascular coagulation were higher than those reported nationally..

Validation of ICD-9-CM codes for major genitourinary birth defects in Military Health System administrative data, 2006-2014

BACKGROUND

The Department of Defense Birth and Infant Health Research program is dedicated to birth defects research and surveillance among military families. Here, we assess and refine the validity of International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for selected genitourinary birth defects in the Military Health System (MHS). We additionally outline methods for the calculation of positive predictive value (PPV) and negative predictive value (NPV), sensitivity, and specificity using a stratified sampling design.

METHODS

Among military infants born from 2006 through 2014, a random sample of ICD-9-CM screen-positive cases (for six genitourinary birth defects) and screen-negative cases were selected for chart review. PPV, NPV, sensitivity, and specificity were calculated for individual defects and any included defect (i.e., overall); measures were weighted by the inverse probability of being sampled.

RESULTS

Of 461,557 infants, 686 were sampled for chart review. Bladder exstrophy was accurately reported (PPV: >90%), while the accuracy of renal dysplasia, renal agenesis/hypoplasia, and hypospadias was moderate (PPVs: 66%-68%) and congenital hydronephrosis was low (PPV: 20%). Specificity and NPVs always exceeded 98%. The overall PPV was 50%; however, excluding congenital hydronephrosis screen-positive cases and requiring at least two inpatient or outpatient diagnostic codes resulted in a PPV of 85%.

CONCLUSIONS

The validity of major genitourinary birth defect codes varied in MHS administrative data. The accuracy of an overall defect measure improved by omitting congenital hydronephrosis and requiring at least two diagnostic codes. Although PPV is generally useful for research, additional calculation of NPV, sensitivity, and specificity better informs the identification of appropriate selection criteria across surveillance and research settings.

Validation of diagnosis codes for chorioamnionitis in medical encounter data, 2013-2018

PURPOSE

Chorioamnionitis refers to intrauterine infection/inflammation that can be diagnosed clinically or from laboratory testing. This study aimed to validate chorioamnionitis International Classification of Diseases (ICD) codes using reference standards for clinical and histologic cases.

METHODS

Department of Defense Birth and Infant Health Research program data identified a cohort of live deliveries at two United States military hospitals from 2013 to 2018. Deliveries were screened for chorioamnionitis using ICD codes from maternal delivery records; a sample of screen positive and negative deliveries was selected for chart review. Primary analyses validated deliveries using a reference standard for clinical chorioamnionitis; secondary analyses employed a reference standard that also included histologic cases, but were limited by temporal differences in availability of laboratory data. Sensitivity, specificity, and positive (PPV) and negative (NPV) predictive values were calculated with 95% confidence intervals (CIs).

RESULTS

Overall, 1857 deliveries (465 screen positive, 1392 screen negative) were eligible for analysis and 336 met the reference standard for clinical chorioamnionitis, yielding a PPV of 0.68 (95% CI 0.63, 0.72) and sensitivity of 0.76 (95% CI 0.72, 0.81). In secondary analyses, 390 deliveries met the reference standard for clinical or histologic chorioamnionitis, resulting in an overall PPV of 0.75 (95% CI 0.71, 0.79); in 2018, when more laboratory results were available, the PPV was 0.91 (95% CI 0.84, 0.97). NPV and specificity were ≥0.97 across reference standards.

CONCLUSIONS

Chorioamnionitis ICD codes exhibited moderate correlation with clinical disease, suggesting challenges in using medical encounter data to isolate clinical cases from those only identified through laboratory testing.

The association of regional perinatal risk factors and neonatal intensive care capacity for Military Health System-insured newborns

OBJECTIVE

To characterize hospitals where military-insured newborns received care and test the association of regional perinatal risk with neonatal intensive care unit (NICU) capacity.

STUDY DESIGN

We identified birth hospitals for live newborns October 2015-December 2018 (n = 296,568) and assigned newborns to health service areas (HSAs). Perinatal risk factors and the number of neonatal special care beds and neonatologists were calculated at HSA levels. Cross-sectional correlation analyses assessed perinatal risk factors and capacity across HSAs.

RESULTS

27.0% (n = 10) of military birth hospitals had special care beds (intermediate and intensive) compared with 44.3% of civilian hospitals (n = 1224; p < 0.05). The number of special care beds and neonatologists per newborn varied more than twofold across regions and were only weakly associated with the proportion of higher risk newborns (R2 < 0.05).

CONCLUSIONS

The lack of meaningful association of regional perinatal risk with NICU capacity poses challenges for effective specialized care among military-associated newborns.

Health outcomes among offspring of US Coast Guard responders to the Deepwater Horizon oil spill, 2010-2011

OBJECTIVE

To evaluate the potential for adverse health outcomes among infants born to US Coast Guard (USCG) responders to the Deepwater Horizon (DWH) oil spill disaster.

METHODS

Department of Defense Birth and Infant Health Research programme data identified a cohort of singleton infants born 2010-2011 to USCG personnel in the DWH Oil Spill Coast Guard Cohort study. Infants were included if their military parent ('sponsor') responded to the oil spill during a selected reproductive exposure window (ie, 3 months preconception for male sponsors and periconception through pregnancy for female sponsors), or if their sponsor was a non-responder. χ2 tests and multivariable log-binomial regression were used to compare the demographic and health characteristics of infants born to spill responders and non-responders.

RESULTS

Overall, 1974 infants with a male sponsor (n=182 responder, n=1792 non-responder) and 628 infants with a female sponsor (n=35 responder, n=593 non-responder) in the DWH Oil Spill Coast Guard Cohort were identified. Health outcomes were similar among the offspring of male responders and non-responders. The frequency of any poor live birth outcome (ie, low birth weight, preterm birth or birth defect) was higher among infants born to female responders (17.1%, n=6) than non-responders (8.9%, n=53); the maternal age-adjusted association was suggestively elevated (risk ratio 1.93, 95% CI 0.89 to 4.16).

CONCLUSION

Infant health outcomes were comparable between the offspring of male USCG oil spill responders and non-responders. Findings were limited by the small number of infants identified, particularly among female responders, and should be interpreted with caution.

Pregnancy and posttraumatic stress disorder: associations with infant outcomes and prenatal care utilization

BACKGROUND

Posttraumatic stress disorder (PTSD) affects 3.6-9.7% of women, and has been associated with adverse outcomes in pregnancy; however, associations with prenatal care (PNC) utilization are not clear.

OBJECTIVE

To evaluate associations of PTSD in pregnancy with PNC utilization and adverse infant outcomes in an active-duty military population (a population with universal health insurance).

METHODS

This was a retrospective cohort study of pregnant active-duty service members in Department of Defense Birth and Infant Health Research program data from 2007 to 2014. Administrative medical encounter data were used to define PTSD cases and outcomes of interest. Descriptive statistics and multivariable log-binomial regression compared PNC utilization and adverse infant outcomes (preterm birth, small for gestational age [SGA], major birth defects) among service members with current PTSD (defined as PTSD in the year prior to pregnancy or during pregnancy) to those without current PTSD.

RESULTS

Of the 103,221 singleton live births identified, 1657 (1.6%) were born to active-duty service members diagnosed with current PTSD. Service members with PTSD were more likely to initiate PNC in the first trimester (93.5% vs. 90.2%) and score adequate plus on the Adequacy of Prenatal Care Utilization Index (63.2% vs. 40.0%) compared to service members without PTSD. PTSD case status was not associated with preterm birth, SGA, or major birth defects, regardless of the adjustment set used (fully adjusted RR 0.96, 95% CI 0.82-1.13; RR 1.08, 95% CI 0.79-1.48; and RR 1.03, 95% CI 0.79-1.34, respectively).

CONCLUSION

For pregnant service members with current PTSD, no associations with adverse infant outcomes were noted, and these patients initiated care earlier and had higher PNC utilization scores compared to pregnant service members without current PTSD. Universal health care coverage and utilization of PNC in this population may mitigate adverse pregnancy outcomes observed in civilian populations of patients with PTSD.

History of pandemic H1N1-containing influenza vaccination and risk for spontaneous abortion and birth defects

BACKGROUND

One recent study suggested an association between receipt of pandemic H1N1 (pH1N1)-containing vaccines in consecutive influenza seasons and spontaneous abortion, but corroborating scientific evidence is limited. In the present study, we leveraged a population of vaccine-compliant pregnant military women to examine history of pH1N1-containing influenza vaccination and adverse pregnancy outcomes. Because seasonal influenza vaccination is compulsory for military service, safety concerns regarding repeat vaccination are particularly relevant in this population.

METHODS

Pregnancies and live births from Department of Defense Birth and Infant Health Research program data were linked with military personnel immunization records to identify women vaccinated with a pH1N1-containing vaccine in pregnancy prior to 21 6/7 weeks' gestation, October 2009-April 2015. Cox and modified Poisson regression models estimated associations between vaccination with pH1N1- versus non-pH1N1-containing influenza vaccine in the season prior to the index pregnancy, and spontaneous abortion and birth defects, respectively. Cox models were calculated for two periods of follow-up: through (1) 21 6/7 weeks' gestation and (2) 28 days postvaccination.

RESULTS

Of 26,264 pregnancies, 21,736 (82.8%) were among women who received a dose of pH1N1-containing vaccine in the prior influenza season and 4,528 (17.2%) were among women who received non-pH1N1-containing vaccine in the prior influenza season. Among 23,121 infants, 19,365 (83.8%) and 3,756 (16.2%) had mothers exposed and unexposed to pH1N1-containing vaccine in the prior influenza season, respectively. The adjusted hazard ratio (aHR) for spontaneous abortion approximated 1.0 across the complete follow-up period (95% confidence interval [CI]: 0.89-1.13) and was slightly elevated when censored at 28 days postvaccination, though the CI was imprecise (aHR: 1.19; 95% CI: 0.97-1.46). No associations with birth defects were observed.

CONCLUSION

This work lends additional safety evidence and support for vaccination against pH1N1 in pregnancy, regardless of the vaccine received in the prior influenza season.

Posttraumatic Stress Disorder in a Cohort of Pregnant Active Duty U.S. Military Servicewomen

The present study aimed to describe the demographic and occupational characteristics, comorbidities, and psychotropic medication receipt associated with posttraumatic stress disorder (PTSD) diagnosis during pregnancy among a sample of active duty U.S. military servicewomen. Data from the U.S. Department of Defense Birth and Infant Health Research program were used to identify pregnancies in active duty servicewomen from 2007 through 2014. Demographic and occupational data were linked with electronic medical and pharmacy records to capture mental health diagnoses and medication receipt dates. Cases of PTSD were identified by the presence of ICD-9-CM Diagnostic Code 309.81 on maternal records from 1 year before the date of the last menstrual period through the end of pregnancy. Of 134,244 identified pregnancies among active duty servicewomen, 2,240 (1.7%) met the case criteria for PTSD. Women with a PTSD diagnosis compared to those without a PTSD diagnosis were more likely to be White non-Hispanic (51.3% vs. 47.4%), unmarried (33.3% vs. 28.2%), in the Army (49.6% vs. 35.8%) or Marine Corps (10.9% vs. 8.0%), in a service and supply occupation (18.2% vs. 13.6%), and to have a junior enlisted rank (56.3% vs. 50.1%) and have been previously deployed (51.2% vs. 39.6%), RRs = 1.15-1.75. Among PTSD cases, the most common mental health comorbidities were depressive disorder (60.9%), adjustment disorder (43.4%), and anxiety disorder (39.3%). During pregnancy, 44.2% of PTSD cases and 7.2% of noncases received psychotropic medications. Demographic and occupational characteristics, comorbidities, and psychotropic medication use differed substantially among PTSD cases and noncases in this large records-based study.

Postpartum Fitness and Body Mass Index Changes in Active Duty Navy Women

INTRODUCTION

Active duty Navy women participate in biannual Physical Fitness Assessments (PFAs), which include height and weight measurements and a Physical Readiness Test (PRT). PFAs are waived during pregnancy and resume the cycle after 6 months following maternity leave. The purpose of this study was to compare changes in PFA results over time between women who had or did not have a live birth during the follow-up period, and identify characteristics of women with lower PFA results postpartum.

MATERIALS AND METHODS

This longitudinal study included 14,142 active duty Navy women, aged 19-40 years, with PFA results during July 2011-June 2015. Multivariable logistic regression, Stuart-Maxwell tests, and mixed effects modeling were used to examine changes in PRT scores and body mass index over time between women with and without a live birth during follow-up. All data were analyzed in 2017. This study was approved by the institutional review boards at the Uniformed Services University Office of Research and the Naval Health Research Center, and informed consent was waived in accordance with 32 CFR § 219.116(d).

RESULTS

Postpartum women had increased odds of PRT failures (AOR = 3.88, 95% CI: 1.44-10.40) and lower PRT scores (AOR = 1.47, 95% CI: 1.12-1.92) up to 2.5 years postpartum, versus women without a live birth. Being enlisted, obese/overweight prepregnancy, and younger were risk factors for suboptimal PFA outcomes. Mean core strength and cardiovascular endurance, but not upper body strength, scores were significantly lower in postpartum women at 1 year postpartum versus women without a live birth.

CONCLUSIONS

Our findings show that additional interventions may be needed to assist women in returning to prepregnancy fitness up to 1 year postpartum. Future studies should examine additional factors that may improve postpartum fitness in addition to enhancing maternity leave policies.

Safety of yellow fever vaccination in pregnancy: findings from a cohort of active duty US military women

BACKGROUND

Yellow fever (YF) is a rare viral disease that can be prevented through receipt of a live attenuated vaccine. In the US military, service members must receive the YF vaccine before assignment to endemic areas, putting active duty service women at heightened risk for inadvertent exposure during preconception or pregnancy. Few studies have investigated the safety of YF vaccination in pregnancy to date, and none in a military population.

METHODS

Department of Defense Birth and Infant Health Research program data were used to identify pregnancies and infants among active duty US military women, 2003-2014. Multivariable regression models estimated associations with YF vaccine exposure during preconception/pregnancy and adverse outcomes (e.g. spontaneous abortion, birth defects). Sensitivity analyses were performed that excluded pregnancies exposed to other live vaccines. For analyses of birth defects only, a secondary sensitivity analysis was performed that excluded infants diagnosed with chromosomal anomalies.

RESULTS

Of the 196 802 pregnancies and 160 706 singleton infants identified, 1347 (0.7%) and 1132 (0.7%), respectively, were exposed to the YF vaccine. No increased risks for adverse pregnancy or infant outcomes were observed in the main analysis. In sensitivity analyses that excluded pregnancies exposed to other live vaccines, preconception YF vaccine exposure was associated with birth defects [adjusted risk ratio (aRR) = 1.71, 95% confidence interval (CI) = 1.08-2.73]; this association was attenuated when further excluding infants with chromosomal anomalies (aRR = 1.59, 95% CI = 0.97-2.62).

CONCLUSIONS

Overall, YF vaccine exposure did not appear to be associated with most adverse outcomes among this population of pregnant military women. A tenuous association between preconception YF vaccine exposure and birth defects was observed in sensitivity analyses, which may warrant further investigation.

Maternal and infant outcomes following exposure to quadrivalent human papillomavirus vaccine during pregnancy

INTRODUCTION

The Department of Defense encourages service members ≤26 years of age to receive the human papillomavirus (HPV) vaccine. Although this vaccine is not recommended in pregnancy, inadvertent vaccination may occur. The objective of this study was to assess whether active duty US military women who received the quadrivalent HPV vaccine (4vHPV) during pregnancy were at increased risk for adverse maternal or infant outcomes.

METHODS

The study population included active duty US military women aged 17-28 years with at least one pregnancy between 2007 and 2014, and the infants resulting from those pregnancies. Pregnancies, live births, and outcomes were identified using medical codes in administrative medical records. Exposure to 4vHPV during pregnancy was ascertained from personnel immunization records. Multivariable regression models were used to calculate risk estimates and 95% confidence intervals for the maternal outcomes of spontaneous abortion, preeclampsia/eclampsia and preterm labor, and the infant outcomes of preterm birth, birth defects, growth problems in infancy or in utero, and infant sex.

RESULTS

Overall, 90,600 pregnancies and 75,670 singleton infants were identified. Approximately 2% of pregnancies and infants were exposed to 4vHPV during pregnancy. After adjustments, no positive associations were detected between inadvertent exposure to 4vHPV during pregnancy and any adverse pregnancy or infant outcomes.

DISCUSSION

Our findings add to an established body of literature demonstrating the safety of 4vHPV when inadvertently administered during pregnancy. Although 4vHPV is no longer administered in the US, its use continues overseas; therefore, safety studies remain important. Furthermore, such studies can provide reassurance to women inadvertently exposed to nonavalent HPV vaccine (9vHPV) in pregnancy, which protects against four of the same antigens as 4vHPV, since safety of 9vHPV has not yet been established in pregnant women.

Japanese encephalitis vaccination in pregnancy among U.S. active duty military women

INTRODUCTION

Japanese encephalitis (JE) vaccine is an inactivated vaccine that has shown no risks in pregnancy in animal models, but epidemiologic studies are lacking. U.S. military service members located in JE endemic regions are required to be vaccinated; understanding the potential adverse events (AEs), including AEs that may occur in pregnancy, is needed. Here, we assessed pregnancy and infant health outcomes in association with JE vaccination in pregnancy.

METHODS

The study population consisted of 192,570 pregnancies to active duty women (2003-2014), captured in the Department of Defense Birth and Infant Health Research program. JE vaccine in pregnancy, vaccine count, formulation, trimester, and whether first career dose coincided with pregnancy were compared with unexposed pregnancies to assess risk of pregnancy and infant health outcomes. Adjusted risk estimates and 95% confidence intervals (CIs) were calculated by multivariable models.

RESULTS

Of the 192,570 identifed pregnancies, 513 were exposed to the JE vaccine; 474 exposures occurred in the first trimester. For all outcomes, elevated risk estimates ranging from 1.53 to 1.70, were observed with receipt of >1 JE vaccine in pregnancy, though 95% CIs were wide and encompassed the null. First dose of JE vaccination in pregnancy was associated with a 1.87 (95% CI: 1.12-3.13) times increased risk of low birthweight (LBW) when excluding pregnancies exposed to other non-routinely recommended vaccinations in pregnancy. All other associations were null in both main and subset analyses.

CONCLUSIONS

The overall results of these analyses provide reassuring findings for the safety of JE vaccination in pregnancy. Higher counts of JE vaccine received in pregnancy yielded large yet non-statistically significant risk estimates for all outcomes, though likely driven by lack of pregnancy awareness. An association was observed with LBW in subset analyses, but it was limited to women receiving their first JE vaccine and not observed in the larger main analyses.

Infant health outcomes among offspring of male U.S. military divers

While there are suggestions that the extreme hyperbaric conditions encountered during deep saturation diving may impact male reproductive function, few studies have investigated whether paternal occupational diving influences offspring health outcomes. To examine this, Department of Defense Birth and Infant Health Research program data were used to identify the offspring of male active duty divers and non-divers in the U.S. military, 2001-2016. Log-binomial regression models estimated associations with infant outcomes (e.g., major structural birth defects, low birthweight). Among 1,148,252 identified singleton infants, 3,843 were considered the offspring of male divers; paternal occupational diving was not positively associated with any adverse infant outcome under study. These findings corroborate existing literature and further suggest that male divers in the U.S. military are not occupationally exposed to reproductive hazards that adversely influence offspring infant health outcomes.

Atovaquone-proguanil exposure in pregnancy and risk for adverse fetal and infant outcomes: A retrospective analysis

BACKGROUND

Malaria in pregnancy can cause severe maternal and fetal complications. Chloroquine (CQ) and mefloquine (MQ) are recommended for chemoprophylaxis in pregnancy, but are not always suitable. Atovaquone-proguanil (AP) might be a viable option for malaria prevention in pregnancy, but more safety data are needed.

METHODS

Data for pregnancies and live births among active duty military women, 2003-2014, from the Department of Defense Birth and Infant Health Research program were linked with pharmacy data to determine antimalarial exposure. Multivariable Cox and logistic regression models were used to assess the relationship of antimalarial exposure with fetal and infant outcomes, respectively.

RESULTS

Among 198,164 pregnancies, 50 were exposed to AP, 156 to MQ, and 131 to CQ. Overall, 17.6% of unexposed pregnancies and 28.0%, 16.0%, and 6.1% of pregnancies exposed to AP, MQ, and CQ, respectively, ended in fetal loss (spontaneous abortion or stillbirth) (adjusted hazard ratios [aHR] = 1.46, 95% confidence interval [CI] 0.87-2.46; aHR = 1.06, 95% CI 0.72-1.57; and aHR = 0.47, 95% CI 0.24-0.94, respectively).

CONCLUSIONS

The small number of AP exposed pregnancies highlights the difficulty in assessing safety. While definitive conclusions are not possible, these data suggest further research of AP exposure in pregnancy and fetal loss is warranted.

TWITTER LINE

More research on fetal loss following atovaquone-proguanil exposure in pregnancy is warranted.

Offspring sex ratio of male active duty U.S. Navy submariners, 2001-2015

The natural human sex ratio at birth (male:female) slightly favors males, and altered sex ratios might be indicative of exposure to reproductive hazards. In the U.S. Navy submarine community, there is a widespread belief that submariners are more likely to father females, but corroborating scientific evidence is limited. To assess this, Department of Defense Birth and Infant Health Research program data were used to identify 7,087 singleton infants whose fathers were considered submariners. Chi-square tests and unconditional logistic regression models were used to compare the offspring sex ratio of male submariners with 2 other active duty populations and the U.S. population. The offspring sex ratio of male submariners was 1.048, which did not substantially differ from the sex ratio of each comparison population. Furthermore, this study found no meaningful variation in offspring sex ratio by length of submarine or military service or by rating.

Department of Defense Birth and Infant Health Registry: select reproductive health outcomes, 2003-2014

Established following a 1998 directive, the Department of Defense Birth and Infant Health Registry (Registry) team conducts surveillance of select reproductive health outcomes among military families. Data are compiled from the Military Health System Data Repository and Defense Manpower Data Center to define the Registry cohort and outcomes of interest. Outcomes are defined using ICD-9/ICD-10 and Current Procedural Terminology codes, and include: pregnancy outcomes (e.g., live births, losses), birth defects, preterm births, and male:female infant sex ratio. This report includes data from 2003-2014 on 1,304,406 infants among military families and 258,332 pregnancies among active duty women. Rates of common adverse infant and pregnancy outcomes were comparable to or lower than those in the general US population. These observations, along with prior Registry analyses, provide reassurance that military service is not independently associated with increased risks for select adverse reproductive health outcomes. The Registry's diverse research portfolio demonstrates its unique capabilities to answer a wide range of questions related to reproductive health. These data provide the military community with information to identify successes and areas for improvement in prevention and care.

Safety of inadvertent anthrax vaccination during pregnancy: An analysis of birth defects in the U.S. military population, 2003-2010

BACKGROUND

Anthrax vaccine adsorbed (AVA) vaccination is compulsory for United States military servicemembers with operational indicators. As the number of female military servicemembers has increased, so has the chance of inadvertent AVA vaccination during pregnancy. Building upon past analyses assessing AVA vaccination during pregnancy and birth defects risk, this study sought to determine if inadvertent AVA vaccination during pregnancy is significantly associated with risk of birth defects after adjusting for other potential risk factors.

METHODS

The study population included 126,839 liveborn infants in the Department of Defense Birth and Infant Health Registry (2003-2010). Mothers were categorized by AVA vaccination exposure timing in relation to pregnancy. Infant medical records were assessed for birth defect diagnoses within the first year of life. Multivariable logistic regression was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs).

RESULTS

Infants of first trimester AVA vaccinated mothers versus receipt at any other time point (OR, 1.10; 95% CI, 0.93-1.29) were not at higher odds of birth defects in adjusted models. Infants of mothers vaccinated prepregnancy versus postpregnancy had a 1.11 (95% CI, 1.01-1.22) higher odds of having a birth defect. Vaccination postpregnancy versus never vaccinated revealed a 10% lower odds of birth defects (OR, 0.90; 95% CI, 0.83-0.99).

CONCLUSIONS

No strong associations between inadvertent AVA vaccination during pregnancy and birth defects risk were observed. Marginal associations between prepregnancy vaccination or never vaccinated women and birth defects risk was observed when compared to postpregnancy vaccination. These findings may be due to self-selection and/or reverse causation bias when assessing comparisons with postpregnancy vaccination, and a "healthy worker" effect when assessing comparisons with women never vaccinated.

Analysis of pregnancy and infant health outcomes among women in the National Smallpox Vaccine in Pregnancy Registry who received Anthrax Vaccine Adsorbed

The National Smallpox Vaccine in Pregnancy Registry (NSVIPR) actively follows women inadvertently vaccinated with smallpox vaccine during or shortly before pregnancy to evaluate their reproductive health outcomes. Approximately 65% of NSVIPR participants also inadvertently received Anthrax Vaccine Adsorbed (AVA) while pregnant, providing a ready opportunity to evaluate pregnancy and infant health outcomes among these women. AVA-exposed pregnancies were ascertained using NSVIPR and electronic healthcare data. Rates of pregnancy loss and infant health outcomes, including major birth defects, were compared between AVA-exposed and AVA-unexposed pregnancies. Analyses included AVA-exposed and AVA-unexposed pregnant women who also received smallpox vaccine 28 days prior to or during pregnancy. Rates of adverse outcomes among the AVA-exposed group were similar to or lower than expected when compared with published reference rates and the AVA-unexposed population. The findings provide reassurance of the safety of AVA when inadvertently received by a relatively young and healthy population during pregnancy.

Postpartum depression and timing of spousal military deployment relative to pregnancy and delivery

OBJECTIVE

This study aimed to determine the relationship between spousal deployment and postpartum depression diagnosis among U.S. military wives, accounting for the timing of deployment with respect to pregnancy and delivery.

METHODS

A retrospective cohort study was conducted to evaluate the association between spousal deployment and postpartum depression among pregnant wives of active-duty service members. Electronic medical records for 161,454 births occurring between 2004 and 2009 were used to define postpartum depression. Three non-mutually exclusive exposure variables were created to categorize deployments as occurring before, during, or after the infant's delivery. A multivariable logistic regression model mutually adjusted for these exposure variables was fitted, producing an odds ratio for each of the three timing categories.

RESULTS

A modest significant association was detected only in those whose husbands deployed in pregnancy and returned after delivery (i.e., deployed during delivery) [odds ratio (OR) 1.10, 95 % confidence interval (CI) 1.04-1.15]. An interactive effect between preexisting depression or anxiety and deployment during delivery was also detected in the data (OR 1.13, 95 % CI 1.07-1.20 for those without a preexisting diagnosis; OR 0.87, 95 % CI 0.80-0.95 for those with a preexisting diagnosis).

CONCLUSION

Health care providers should continue to be aware of spousal deployment as a military-unique stressor in this population and rigorously screen for potential symptoms of postpartum depression, especially among those whose husbands are absent at delivery.

Birth defects in infants born in 1998-2004 to men and women serving in the U.S. military during the 1990-1991 Gulf War era

BACKGROUND

Concerns about reproductive health persist among U.S. military members who served in the 1990-1991 Gulf War. This study explores the long-term impact of 1990-1991 Gulf War deployment on the prevalence of birth defects among infants of Gulf War veterans.

METHODS

Health care data from the Department of Defense Birth and Infant Health Registry and demographic and deployment information from the Defense Manpower Data Center were used to identify infants born between 1998 and 2004 to both male and female 1990-1991 Gulf War veterans. Multivariable logistic regression models estimated the adjusted odds of any birth defect and eight specific birth defects among infants of deployers versus non-deployers. In addition, birth defects were evaluated among infants born to 1990-1991 Gulf War veterans with deployment-specific exposures.

RESULTS

Among 178,766 infants identified for these analyses, 3.4% were diagnosed with a birth defect in the first year of life. Compared to infants of non-deployers, infants of deployers were not at increased odds of being diagnosed with a birth defect, or any of eight specific birth defects, in the first year of life. A slightly increased prevalence of birth defects was observed among infants born to men who deployed to the 1990-1991 Gulf War for 153 to 200 days compared to those who deployed for 1 to 92 days. No other deployment-specific exposures were associated with birth defects in these infants.

CONCLUSIONS

The 1990-1991 Gulf War deployers, including those with specific exposures of concern, were not found to be at increased risk for having infants with birth defects 7 to 14 years after deployment.

Racial differences in prostate cancer risk remain among US servicemen with equal access to care

BACKGROUND

Prostate cancer is the most common cancer among US men, however, the etiology remains unclear. Yet, one consistency is that black non-Hispanic men are at increased risk for prostate cancer compared to white, non-Hispanic men. The goal of this study was to assess relations between demographic and other potential prostate cancer risk factors in the context of the US military healthcare system, which provides equal access to all US servicemen.

METHODS

Military healthcare and demographic data were used to describe risk factors for prostate cancer in the US military from September 1993 to September 2003. Cox's proportional hazards regression was employed to model the time to prostate cancer hospitalization.

RESULTS

Four hundred eight first prostate cancer hospitalizations were identified among 2,761,559 servicemen. The adjusted rate per 100,000 persons rose from 1.41 to 3.62 for white non-Hispanic men and 1.43 to 6.08 for black non-Hispanic men by the end of the study. The increasing incidence over time for combined race/ethnic groups was similar to trends reported in the Surveillance, Epidemiology, and End Results Program for the US civilian population. No association was observed between occupation and prostate cancer hospitalization. However, black non-Hispanic men were at increased risk compared with white non-Hispanic men (hazard ratio = 2.72, 95% confidence interval: 2.12, 3.49).

CONCLUSIONS

No association was observed between occupation and prostate cancer hospitalization. In this relatively young cohort, black non-Hispanic race/ethnicity was found to be predictive of prostate cancer, and this association existed regardless of access to care and socioeconomic status.

Haemangiomas and associated congenital malformations in a large population-based sample of infants

Haemangiomas are common, benign, vascular tumours, observed in 4-12% of infants during the first year of life. Most cases progress without complication, yet a small proportion experience life-threatening complications. Concomitant congenital malformations have been reported in a small but significant proportion of haemangioma patients. This study aimed to describe haemangioma cases and to identify patterns of congenital malformations associated with these diagnoses in a large population. Diagnoses of haemangiomas and 21 congenital malformations were extracted from electronic medical records of 467 295 singleton infants born to US military families from 1998 to 2003. Cluster analysis was used to group cases according to these diagnoses. Multivariable logistic regression was used to further explore the associations of the 21 congenital malformations with the diagnosis of haemangioma and to assess the adjusted relationships between a number of characteristics of interest and diagnosis of haemangioma. Clusters found to be associated with haemangioma were characterised by anomalies of the cervix, vagina, and external female genitalia, anophthalmia or microphthalmia, hydrocephalus without spina bifida, and reduction deformities of the brain. Logistic regression identified three congenital malformations significantly associated with haemangioma diagnosis: spina bifida without anencephalus, hydrocephalus without spina bifida, and anomalies of the cervix, vagina and external female genitalia. Characteristics significantly associated with haemangioma included female gender, preterm birth, white non-Hispanic race/ethnicity and increasing maternal age. This exploratory study identified a number of important associations between haemangiomas and congenital malformations that may provide insight into the pathogenesis of these disorders and have possible implications for clinical care.