Breech Presentation at Delivery: a Marker for Congenital Anomaly?

ORIGINAL ARTICLE Breech presentation at delivery: a marker for congenital anomaly?

D Mostello1, JJ Chang2, F Bai2, J Wang3, C Guild4, K Stamps2 and TL Leet2,{

OBJECTIVE: To determine whether congenital anomalies are associated with breech presentation at the time of birth. STUDY DESIGN: A population-based, retrospective cohort study was conducted among 460 147 women with singleton live births using the Missouri Birth Defects Registry, which includes all defects diagnosed during the ﬁrst year of life. Maternal and obstetric characteristics and outcomes between breech and cephalic presentation groups were compared using w2-square statistic and Student’s t-test. Multivariable binary logistic regression analysis was used to estimate adjusted odds ratios (aORs) and 95% conﬁdence intervals (CIs). RESULT: At least one congenital anomaly was more likely present among infants breech at birth (11.7%) than in those with cephalic presentation (5.1%), whether full-term (9.4 vs 4.6%) or preterm (20.1 vs 11.6%). The relationship between breech presentation and congenital anomaly was stronger among full-term births (aOR 2.09, CI 1.96, 2.23, term vs 1.40, CI 1.26, 1.55, preterm), but not in all categories of anomalies. CONCLUSION: Breech presentation at delivery is a marker for the presence of congenital anomaly. Infants delivered breech deserve special scrutiny for the presence of malformation.

Journal of Perinatology (2014) 34, 11–15; doi:10.1038/jp.2013.132; published online 24 October 2013 Keywords: congenital abnormalities; labor presentation; pregnancy outcome; registries

INTRODUCTION been an obstetric maxim that congenital anomalies are associated 5 In about 4% of pregnancies, the fetus is in breech presentation at with breech presentation, the data supporting this assertion are 2,5,6,9–11 the time of delivery.1 Although the immediate impact of this from studies limited by their small sample sizes (especially finding for the obstetrician is typically on mode of delivery, the problematic when reporting rates of uncommon anomalies), lack 5,11 reason why the fetus failed to assume the cephalic presentation of stated method of ascertainment of anomalies, incomplete 2,10 warrants consideration. Unfortunately, the mechanism controlling statistical analysis, findings limited to those reported on birth 7 5 fetal alignment for delivery is not clear. Hypotheses to explain why certificates, and experience reported from a single institution. the fetus fails to assume the cephalic presentation include primary If breech presentation is indeed a marker for underlying fetal disorders, such as fetal neuromuscular dysfunction or abnormality, the presence of breech presentation near the time neurologic deficits2,3 or spatial considerations, such as uterine of delivery becomes an important trigger for the obstetrician and anomalies,1 site of placental implantation4 or amount of amniotic pediatrician to think beyond scheduling the Cesarean section and fluid,5,6 which may interfere with the process of achieving cephalic checking for hip clicks to an enhanced scrutiny of the fetal/ presentation. These theories revolve around primarily mechanical newborn anatomy and function. factors due to the lack of proper fetal movement or tone,7 a bulky We sought to address the question of whether congenital hydrocephalic head or lack of proper uterine architecture due to anomaly in the newborn is associated with breech presentation at over-distention or a lax maternal abdominal wall. the time of live birth, using population-based data with a reliable Although these explanations seem logical, existing data do not source of anomaly ascertainment, and multivariable analysis to support these hypotheses. For example, although breech presenta- identify and control for relevant confounding factors. tion occurs in almost one-third of infants with dysautonomia, no greater hypotonia is noted in those who present breech vs cephalic.8 In addition, when controlled for gestational age at birth, breech METHODS presentation was not associated with polyhydramnios or placenta The Missouri Birth Defects Registry is a passive data collection system that previa in a large population-based study.7 Indeed, recognition of a identifies live-born infants diagnosed with birth defects during the first potential mechanical factor to explain presentation is found in only year of life. This registry receives information from multiple sources, 15% of breech deliveries systematically assessed for them.6 including birth and death certificates, abstracts of medical records of 2 newborn and pediatric patients from both inpatient and outpatient Braun et al. proposed the concept that breech presentation encounters, and a special healthcare needs database. may constitute an important indicator of a ‘deeper underlying Using this resource, a population-based, retrospective cohort study was problem in fetal morphogenesis and/or function’ of which the conducted among women who gave birth at 24 weeks gestation or greater breech presentation is only one manifestation. Although it has in Missouri between January 1, 1993 and December 31, 1999, and the

1Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Women’s Health, School of Medicine, Saint Louis University, St Louis, MO, USA; 2Department of Epidemiology, School of Public Health, Saint Louis University, St Louis, MO, USA; 3Department of Biostatistics, School of Public Health, Saint Louis University, St Louis, MO, USA and 4Department of Pediatrics and the Center for Outcomes Research, School of Medicine, Saint Louis University, St Louis, MO, USA. Correspondence: Dr D Mostello, Division of Maternal-Fetal Medicine, Department of Obstetrics, Gynecology and Women’s Health, School of Medicine, Saint Louis University, 6420 Clayton Road, St Louis, MO 63117, USA. E-mail: [email protected] Presented at the 32nd Annual Meeting of the Society for Maternal-Fetal Medicine, February 6–11, 2012, Dallas, TX, USA. {Dr Leet is deceased. Received 1 May 2013; revised 1 September 2013; accepted 13 September 2013; published online 24 October 2013 Breech presentation and congenital anomaly D Mostello et al 12 Table 1. Demographic, obstetric and medical characteristics by fetal presentation at delivery stratiﬁed by term status

Characteristics X37 weeks P-value o37 weeks P-value

Cephalic Breech Cephalic Breech n ¼ 415 273 (%) n ¼ 11 939 (%) n ¼ 30 063 (%) n ¼ 2872 (%)

Maternal age (years), Mean (s.d.) 26.3 (5.9) 27.3 (6.0) o0.0001 25.8 (6.3) 26.7 (6.5) o0.0001 12–18 9 7.5 o0.0001 12.7 9.5 o0.0001 19–27 49.2 43.8 49.3 46.9 28–35 34.7 38.9 30.5 33 435 7.1 9.8 7.6 10.7 Maternal race o0.0001 o0.0001 Non-Hispanic whites 82 89 71.3 81 Non-Hispanic blacks 14.7 7.7 25.6 16.3 Others 3.4 3.3 3.1 2.7 Maternal education o0.0001 o0.0001 High school diploma or lower 53.7 48.9 62.3 58.2 Some college or higher 46.3 51.1 37.7 41.8 Maternal low SES 47.6 40.8 o0.0001 57.3 51.5 o0.0001 Parity o0.0001 o0.0001 0 40.8 54.5 45.7 49.4 1 33.6 27.3 27.9 28.4 X2 25.6 18.1 26.4 22.2 Smoking usage 19.5 21.2 o0.0001 25.7 27.5 0.04 Alcohol usage 1.3 1.5 0.12 2.4 2.1 0.3 Hydramnios/oligohydramnios 1.5 3.2 o0.0001 4.4 10.7 o0.0001 Diabetes 2.4 3.2 o0.0001 4.3 4.8 0.23 Renal disease 0.24 0.22 0.70 0.77 0.9 0.43 Hypertension, preeclampsia 4.5 4.6 0.72 14.8 15.1 0.64 Cardiac disease 0.63 0.75 0.12 0.83 0.87 0.80 Late or no prenatal care 14.5 12 o0.0001 20.9 16.7 o0.0001 Female child 54.7 48.8 o0.0001 50.0 46.9 0.002 Child’s birth weight (g), Mean (s.d.) 3442 (476) 3334 (505) o0.0001 2446 (652) 2083 (749) o0.0001 Gestational age (weeks), Mean (s.d.) 39.4 (1.1) 39.0 (1.1) o0.0001 34.4 (2.3) 32.7 (3.3) o0.0001 SGA 9.5 13.8 o0.0001 15.9 20.2 o0.0001 Abbreviations: SES, socioeconomic status; SGA, small for gestational age. Data are percent of n unless speciﬁed as mean þ / À standard deviation (s.d.).

resulting offspring from these births. The study population consisted of all a categorical variable to account for a potential nonlinear relationship and Missouri residents who delivered singleton pregnancies. Eligible women categorized as less than 18 years, 19 to 27 years, 28 to 35 years, and 36 were identified by using birth certificate and birth defects registry data. years or older. Race was categorized as non-Hispanic white, non-Hispanic Only singleton live births were included to avoid the confounding effects black and other. Low socioeconomic status was defined as enrollment in of multiple gestation on breech presentation and other adverse outcomes, any of the programs including Medicaid, Women, Infants and Children, or such as shorter pregnancy duration and low birth weight. We excluded the Food Stamp Program. Education was divided into two groups: lower pregnancies complicated by placenta previa as a possible confounder for than or equal to a high school degree and some college or more. Cigarette breech presentation.1,4 The study cohort included women with a fetus in smoking and alcohol consumption were treated as dichotomous variables. either breech or cephalic presentation at the time of delivery. Other Parity was defined as prior live births and categorized into three groups, 0, presentations were excluded. 1 and X2. Prenatal care was categorized as care beginning during the first Breech presentation was considered present at the time of delivery if it was trimester, and care beginning after the first trimester (late) or no care checked on the birth certificate. Congenital anomalies were identified through through the pregnancy. Small for gestational age was also used as an the Missouri Birth Defects Registry. Congenital anomalies were analyzed in indicator of relative infant birth weight, which was defined as birth weight toto as well as in specific systems, including chromosomal anomalies, central below the tenth percentile for the clinical estimate of gestational age nervous system anomalies, cardiovascular anomalies, oral clefts, gastrointest- based on a national standard.12 inal anomalies, urogenital anomalies and musculoskeletal anomalies. We In the descriptive analysis, the w2-square statistic was used for also created a category for ‘multi-system anomalies’, defined as anomalies categorical data and Student’s t-test for continuous variables to study identified in more than one system, as a proxy for complex anomalies and and compare the characteristics between breech and cephalic presenta- congenital syndromes that may involve multiple organ systems. tion groups. Because the rate of breech presentation at birth is inversely Maternal and fetal factors previously found to be associated with either proportional to gestational age,13 we stratified our bivariable and breech presentation or congenital anomalies were evaluated as potential multivariable analyses by term status categorized as preterm birth (o37 confounders (listed in Table 1): Maternal demographic, lifestyle and weeks) or full-term birth (X37 weeks). The descriptive analysis, therefore, pregnancy characteristics included maternal age, race, low socioeconomic also was stratified to allow for easier reference and comparison. We status, education, cigarette smoking, alcohol consumption, parity and late conducted bivariable analysis between outcomes of interest and or no prenatal care. Obstetric and maternal medical risk factors included presentation using the w2-square statistic. Multivariable binary logistic hydramnios or oligohydramnios, diabetes of any type, renal disease, regression analysis was used to estimate odds ratios and 95% confidence chronic hypertension or preeclampsia, and cardiovascular disease. Infant intervals (CIs). Potential confounders were evaluated using the 10% characteristics, such as birth weight, gender, and gestational age were also change-in-point-estimate rule in the multivariable analysis.14 Primary assessed in the analysis. Some risk factors and potential confounders are outcomes were congenital anomaly of any kind, congenital anomaly not available fields in the database, including maternal congenital affecting specific organ systems, or multi-system anomalies. All statistical anomalies, maternal uterine anomalies, and the performance of an analyses were conducted using SAS, version 9.1 software (Cary, NC, USA). external cephalic version prior to delivery. Maternal age was analyzed as All tests were two-tailed, and Po0.05 was considered significant.

Journal of Perinatology (2014), 11 – 15 & 2014 Nature America, Inc. Breech presentation and congenital anomaly D Mostello et al 13 Table 2. Anomalies according to presentation at birth, stratiﬁed by term status

Congenital anomaly X37 weeks P-value o37 weeks P-value

Breech Cephalic Breech Cephalic n ¼ 11 939 n (%) n ¼ 415 273 n (%) n ¼ 2872 n (%) n ¼ 30 063 n (%)

Any anomalies 1126 (9.4) 19234 (4.6) o0.0001 600 (20.1) 3474 (11.6) o0.0001 Central nervous system 77 (0.6) 915 (0.2) o0.0001 83 (2.9) 462 (1.5) o0.0001 Cardiovascular 226 (1.9) 5219 (1.3) o0.0001 238 (8.3) 1363 (4.5) o0.0001 Oral clefts 33 (0.3) 707 (0.2) 0.006 20 (0.7) 84 (0.3) 0.0001 Gastrointestinal 83 (0.7) 2308 (0.6) 0.04 64 (2.2) 484 (1.6) 0.01 Urogenital 200 (1.7) 4968 (1.2) o0.0001 100 (3.5) 600 (2) o0.0001 Musculoskeletal 587 (4.9) 4463 (1.1) o0.0001 161 (5.6) 718 (2.4) o0.0001 Chromosomal 56 (0.5) 677 (0.2) 0.0001 38 (1.3) 205 (0.7) 0.0001 Multi-system 137 (1.2) 1591 (0.4) o0.0001 123 (4.3) 592 (2) o0.0001

This study was considered by Saint Louis University Institutional Review with breech presentation; therefore, we controlled for gestational Board as ‘non-human subject research’, owing to the use of de-identified age (by week) in the multivariable analyses. data, and no further oversight was needed. Owing to the significantly different distributions of risk by fetal presentation, we evaluated maternal age, race, education, socioeconomic status, parity, smoking, alcohol usage, amniotic fluid RESULTS abnormalities, maternal medical conditions during pregnancy, A total of 519 504 births in Missouri occurred over the study insufficient prenatal care and infant’s small for gestational age period, among which 475 551 met the inclusion criteria. In total, status, sex, as well as gestational age as potential confounding 460 147 subjects were included in the final analysis, as 15 404 variables. Of these covariates, only gestational age (by week) women had at least one missing value for the outcome or other changed the point estimates by 10% or more. After adjustment for variables included in the analysis. The study group included gestational age and some other clinically relevant variables, 14 811 (3.2%) women with breech presentation and 445 336 having at least one congenital anomaly occurred with greater (96.8%) women with cephalic presentation. There were 32 935 likelihood in infants delivered with breech presentation than in (7.2%) preterm births among all subjects in the study population. those with cephalic presentation (for full-term birth: aOR (adjusted The incidence of breech presentation was 2.8% among full-term odds ratio) ¼ 2.09, 95% CI: 1.96, 2.23, and for preterm birth: births, and 9.5% among preterm births. aOR ¼ 1.40, 95% CI: 1.26, 1.55, Table 3). The odds of having a Sample characteristics by fetal presentation, stratified by term congenital anomaly in a specific organ system when born breech status, are shown in Table 1. The preterm birth group and full-term compared with cephalic are shown in Table 3 for both preterm birth group shared similar distributions of most characteristics, and term births. Though infants born preterm were more likely to although many were statistically different. Women giving birth have a congenital anomaly, a stronger association between with breech presentation were significantly older, more likely to congenital anomaly and breech presentation was seen for full- have higher levels of education and socioeconomic status, to be term birth, and specifically for central nervous system, muscu- non-Hispanic white and nulliparous. They also were more likely to loskeletal system, and chromosomal anomalies. Oral clefts showed have an abnormality of amniotic fluid volume. Infants born breech a stronger relationship with breech presentation in preterm rather were more likely to be small for gestational age. In those women than full-term births. delivering at term, diabetes was more prevalent in those with breech presentation. Of those women who delivered preterm, those whose infants were in breech presentation had infants of DISCUSSION lower gestational age and birth weight. Our finding of a greater likelihood of congenital anomaly in At least one congenital anomaly was more likely to be present in infants with breech presentation supports the concept of an infants in the breech presentation (11.7%) than those in cephalic underlying problem in fetal morphogenesis or function2 of which presentation (5.1%) at birth. Table 2 shows the rates of congenital the breech presentation is only one sign. Although in no way do anomaly for breech vs cephalic presentation for both preterm we consider the presence of breech presentation in the causal (20.1% vs 11.6%) and term deliveries (9.4% vs 4.6%). The pattern pathway of most congenital disorders, breech presentation is an was similar for chromosomal anomalies and in specific organ easily identifiable factor at the time of birth that may signal the systems including central nervous, cardiovascular, oral clefts, presence of an underlying abnormality. The association is present urogenital, gastrointestinal and musculoskeletal. Those who deliv- for many categories of anomalies and not limited only to ered preterm were more likely to have congenital anomalies chromosomal and central nervous system defects. The high regardless of presentation. Among the systems we analyzed, percentage of births affected by congenital anomalies in our study cardiovascular, musculoskeletal and urogenital anomalies were is about twice the rates reported in other studies5,7,11,15 some of the most common for both term and preterm groups. Multi-system which used only birth certificate data or other short-term anomalies were more likely in infants born from the breech ascertainment of cases. Our overall rates are in a similar range presentation than in cephalic-presenting infants in both term (1.2% as those referred to by Cruikshank1 in his review of breech vs 0.4%) and preterm groups (4.3% vs 2%). We detected a presentation and by Kauppila16 in his comprehensive assessment significant interaction effect between breech presentation and term of perinatal mortality in breech deliveries, but categories of status for anomalies in the following systems: central nervous anomalies were not specified in these reports. (Po0.01), cardiovascular (Po0.02) and musculoskeletal (Po0.0001). Despite the common practice of a midtrimester, sonographic This confirmed our decision, based on clinical observation, to anomaly screen, prenatal detection of abnormalities was esti- stratify by term status. Within each stratum, gestational age was an mated at 64% in the best of hands and at 13% in non-tertiary important confounder for the association of congenital anomalies centers.17 The largest study addressing prenatal detection rates18

& 2014 Nature America, Inc. Journal of Perinatology (2014), 11 – 15 Breech presentation and congenital anomaly D Mostello et al 14 Table 3. Multivariable analyses: odds of having a congenital anomaly with breech compared to cephalic presentation at delivery, stratiﬁed by term status

Congenital anomalies X37 weeks (n ¼ 427 212) o37 weeks (n ¼ 32 935)

cOR (95% CI) aOR (95% CI)a cOR (95% CI) aOR (95% CI)a

Any anomalies 2.15 (2.01, 2.29) 2.09 (1.96, 2.23) 2.02 (1.84, 2.23) 1.40 (1.26, 1.55) Central nervous 2.94 (2.33, 3.71) 2.61 (2.06, 3.30) 1.91 (1.51, 2.41) 1.14 (0.89, 1.46) Cardiovascular 1.52 (1.33, 1.73) 1.40 (1.22, 1.60) 1.90 (1.65, 2.20) 1.24 (1.04, 1.48) Oral clefts 1.63 (1.15, 2.31) 1.51 (1.06, 2.14) 2.50 (1.54, 4.08) 1.91 (1.13, 3.21) Gastrointestinal 1.26 (1.01, 1.57) 1.21 (0.97, 1.50) 1.39 (1.07, 1.81) 1.04 (0.79, 1.36) Urogenital 1.41 (1.22, 1.62) 1.46 (1.26, 1.69) 1.77 (1.43, 2.20) 1.43 (1.14, 1.79) Musculoskeletal 4.76 (4.36, 5.20) 4.47 (4.08, 4.89) 2.43 (2.04, 2.89) 2.00 (1.66, 2.40) Chromosomal 2.89 (2.20, 3.79) 1.91 (1.44, 2.54) 1.95 (1.38, 2.77) 1.52 (1.05, 2.21) Multi-system 3.15 (2.64, 3.76) 2.18 (2.05, 2.33) 2.39 (1.96, 2.92) 1.40 (1.26, 1.56) Abbreviations: aOR, adjusted odds ratio; CI, conﬁdence interval; cOR, crude odds ratio. aAdjusted for maternal age, race, education, low socioeconomic status, smoking, alcohol usage, diabetes, hydramnios/ oligohydramnios, child’s sex, small for gestational age and gestational age (by week).

used infant status at birth and at 6 days as the standard for the presentation,1 so the cohort of preterm births in which breech presence of anomalies, and only calculated detection rates based presentation is associated with anomaly may be diluted by breech on sonographically detectable abnormalities, so the percentage of presentation due to other factors. Brenner et al.5 found that most abnormalities identified prenatally is likely lower than reported. of the infants with congenital abnormalities were delivered after Therefore, the presence of breech presentation at birth can be an 35 weeks gestation and Mazor et al.11 reported that 90% of all important marker to raise the index of suspicion in evaluating congenital anomalies occurred in neonates weighing 2000 g or newborns for the presence of anomalies. Although ultrasound more at birth. Indeed, our findings show that 80% of infants with technology has improved over time, making it possible that a anomalies are delivered at term. greater number of anomalies would be detected prenatally now Our findings substantiate birth certificate7 and Collaborative when compared with when these studies were done, the increase Project10 findings of association of breech presentation with in maternal obesity and continued limited access to high-quality older maternal age,7,10 primiparity,7 low birth weight,7,10 shorter ultrasound services may counteract the advances in technology. duration of gestation,6,7,10 maternal diabetes, smoking and In the Collaborative Project, infants who were delivered breech Caucasian race. No explanations for these associations have been fared worse on individual components of the neurological established to date. High parity is considered a risk factor for breech examination at 1 year compared with children of cephalic presentation, but an association of breech presentation with a deliveries.10 This greater neurologic morbidity in those delivered parity of two or more was not evident in our study. The association breech has been attributed to the mode of delivery,10 but the observed in prior studies seems to apply only at parity X510 or 6,3 presumption of birth trauma in vaginal breech deliveries may and sample size for that degree of parity is small in our database. have precluded detection of an association between neurologic The major strengths of our study include the large numbers of function and the presentation itself. Braun et al.2 was prescient in subjects from the population-based design and the comprehen- proposing that the deficient function may be of prenatal onset sive ascertainment of anomalies through the first year of life by and the reason behind the breech presentation as well. the Birth Defects Registry. Not all birth defects are obvious at birth. Our findings confirm those from other studies which show Indeed, ascertainment limited to defects observed at birth or congenital anomalies are significantly associated with breech shortly thereafter would likely capture most neural tube defects, presentation.2,5,11 Brenner et al.5 also found overall rates for for example, but would miss a significant percentage of heart and congenital abnormalities of every organ system were higher in other defects.17,19,20 When assessed over a longer period, about breech deliveries, but owing to smaller numbers, found significant 19% of cases of birth defects are diagnosed prenatally, 48% differences in only a few systems after correction for confounding between birth and one month of age, and 12% between one factors. Of the congenital anomalies (116 in total) found by Mazor month and a year of age.20 Another 12% are found later in et al.11 among breech deliveries at a single institution in Israel, the childhood.20 Registries of birth defects, such as the one we used, organ system most involved was musculoskeletal, with dislocation have improved case ascertainment with inclusion of data from of the hip accounting for 83% of the cases. In our study, infant death certificates21 and hospital discharges.22 They capture congenital dislocation of the hip was present in about 25% of 89% of the birth defects ascertained by a more intensive breech infants with anomalies at term and about 6% of breech surveillance system.23 infants with anomalies born preterm (data not shown). Braun Our large sample size provided sufficient statistical power to et al.2 reported on specific genetic disorders of infants and evaluate for the presence of congenital anomaly overall as well as associations with breech presentation at delivery. Unfortunately, by various organ systems and to stratify by term status. The nature our database did not specify diagnoses to the same level of detail. of our population-based data, however, does impose some Term status modifies the association between risk of congenital limitations: Some women in the cohort may have delivered more anomaly and presentation at birth. The stronger association of than once in the period selected; the data was de-identified and congenital anomaly with breech presentation at term compared not maternally linked, so we are unable to control for that with preterm is not surprising. As the rate of breech presentation possibility. In addition, the reliance on birth certificate data for at birth is inversely proportional to gestational age,12 some of the some of the demographic data may introduce misclassification breech presentations in the premature cohort may be explained errors with regard to gestational age, the presence of breech by early gestational age alone, with labor or other reason presentation as well as medical risk factors.24 However, the for delivery occurring before the fetus would have converted misclassification of some measurements is likely non-differential, to a cephalic presentation on her own. Similarly, factors that which would bias our point estimates toward the null value. In a predispose to preterm labor may also predispose to breech comparison with information from hospital records, birth

Journal of Perinatology (2014), 11 – 15 & 2014 Nature America, Inc. Breech presentation and congenital anomaly D Mostello et al 15 certificates correctly reported breech presentation 91% to 99% of 2 Braun FHT, Jones KL, Smith DW. Breech presentation as an indicator of fetal the time.24,25 Residual confounding may be present from poten- abnormality. J Pediatr 1975; 86: 419–421. tially relevant variables not available in our dataset. Factors which 3 Axelrod FB, Leistner HL, Porges RF. Breech presentation among infants with may be associated with breech presentation, such as maternal familial dysautonomia. J Pediatr 1974; 84: 107–109. uterine anomalies and uterine leiomyomas, were not available, 4 Fianu S, Vaclavinkova V. The site of placental attachment as a factor in the hence, were not controlled for in the analysis. In addition, we may aetiology of breech presentation. Acta Obstet Gynecol Scand 1978; 57: 371–372. 5 Brenner WE, Bruce RD, Hendricks CH. The characteristics and perils of breech have underestimated the association between breech presen- presentation. Am J Obstet Gynecol 1974; 118: 700–712. tation and anomalies by failing to account for infants born from 6 Luterkort M, Persson P, Weldner B. Maternal and fetal factors in breech the cephalic presentation after a successful external cephalic presentation. Obstet Gynecol 1984; 64: 55–59. version from breech prior to delivery. Another possible limitation 7 Rayl J, Gibson PJ, Hickok DE. A population-based case-control study of risk factors is diagnostic bias. Some clinicians may be already armed with a for breech presentation. Am J Obstet Gynecol 1996; 174: 28–32. belief that infants delivered from the breech presentation are 8 Axelrod FB, Porges RF, Sein ME. Neonatal recognition of familial dysautonomia. more likely to have anomalies, and therefore these infants receive J Pediatr 1987; 110: 946–948. more detailed evaluation in the first year of life, resulting in 9 Gimovsky ML, Paul RH. Singleton breech presentation in labor: experience in greater reporting of anomalies in this group. 1980. Am J Obstet Gynecol 1982; 143: 733–739. 10 Berendes HW, Weiss W, Deutschberger J, Jackson E. Factors associated with According to our findings, the presence of breech presentation breech delivery. Am J Public Health Health 1965; 55: 708–719. at the time of delivery, especially at term, is a marker for the 11 Mazor M, Hagay ZJ, Leiberman JR, Biale Y, Insler V. Fetal malformations associated presence of a congenital anomaly. As a significant risk factor for an with breech delivery. Implications for obstetric management. J Reprod Med 1985; anomaly, the breech presentation at delivery should be noted and 30: 884–886. conveyed, as any risk factor, in documentation beyond the 12 Alexander GR, Kogan MD, Himes JH. 1994–1996 US singleton birth weight per- delivery summary. Finding breech presentation close to term centiles for gestational age by race, Hispanic origin, and gender. Matern Child prenatally, especially in someone previously unevaluated or Health J 1999; 3: 225–231. incompletely evaluated sonographically, may warrant targeted 13 Sorensen T, Hasch E, Lange AP. Fetal presentation during pregnancy [Letter]. ultrasound evaluation.11 Careful evaluation in the nursery for Lancet 1979; 2: 477. 14 Mickey RM, Greenland S. The impact of confounder selection criteria on effect anomalies as well as genetic evaluation may be indicated for those estimation. Am J Epidemiol 1989; 129: 125–137. delivered from the breech presentation. Screening, such as with 15 Hall JE, Kohl SG, O’Brien F, Ginsberg M. Breech presentation and perinatal mor- pulse oximetry for critical congenital heart disease, if not done tality. Am J Obstet Gynecol 1965; 91: 665–683. universally, may be appropriately targeted to infants born breech. 16 Kauppila O. The perinatal mortality in breech deliveries and observations on Because the breech presentation may only be a marker for some affecting factors: a retrospective study of 2227 cases. Acta Obstet Gynecol Scand underlying disorder, appreciation of the association may spur (Suppl) 1975; 39: 1–79. earlier recognition of problems in the weeks to years following 17 Goldberg JD. Routine screening for fetal anomalies: expectations. Obstet Gynecol discharge from the nursery. Further research using genetic Clin N Am 2004; 31: 35–50. microarray and follow-up of neurologic development may help 18 Grandjean H, Larroque D, Levi S. Sensitivity of routine ultrasound screening of pregnancies in the Eurofetus database. The Eurofetus Team. Ann N Y Acad Sci in determining the nature of the underlying disorders. 1998; 847: 118–124. 19 Wren C, Reinhardt Z, Khawaja K. Twenty-year trends in diagnosis of life-threa- tening neonatal cardiovascular malformations. Arch Dis Child Fetal Neonatal Ed CONFLICT OF INTEREST 2008; 93: F33–F35. The authors declare no conflict of interest. 20 Bower C, Rudy E, Callaghan A, Quick J, Nassar N. Age at diagnosis of birth defects. Birth Defects Res A Clin Mol Teratol 2010; 88: 251–255. 21 Tanner JP, Salemi JL, Hauser KW, Correia JA, Watkins SM, Kirby RS. Birth defects ACKNOWLEDGEMENTS surveillance in Florida: infant death certificates as a case ascertainment source. Birth Defects Res A Clin Mol Teratol 2010; 88: 1017–1022. Many thanks go to Gilad Gross, MD for his helpful advice on improving this 22 Wang Y, Sharpe-Stimac M, Cross PK, Druschel CM, Hwang S. Improving case manuscript. The authors also acknowledge the Missouri Department of Health and ascertainment of a population-based birth defects registry in New York state using Senior Services, Section of Epidemiology for Public Health Practice as the original hospital discharge data. Birth Defects Res A Clin Mol Teratol 2005; 73: 663–668. source of the data. The analysis, interpretations and conclusions in the present study 23 Salemi JL, Tanner JP, Kennedy S, Block S, Bailey M, Correia JA et al. A comparison are those of the authors and not the Missouri Department of Health and Senior of two surveillance strategies for selected birth defects in Florida. Public Health Services, Section of Epidemiology for Public Health Practice. Rep 2012; 127: 391–400. 24 Frost F, Starzyk P, George S, McLaughlin JF. Birth complication reporting: the effect of birth certificate design. Am J Public Health 1984; 74: 505–506. REFERENCES 25 Zollinger TW, Przybylski MJ, Gamache RE. Reliability of Indiana birth certificate 1 Cruikshank DP. Breech presentation. Clin Obstet Gynecol 1986; 29: 255–263. data compared to medical records. Ann Epidemiol 2006; 16: 1–10.

& 2014 Nature America, Inc. Journal of Perinatology (2014), 11 – 15