A pregnancy is defined as post-term when it reaches 42 weeks (294 days) from the first day of the last normal menstrual period, or 14 days beyond the best obstetric estimate of the date of delivery. Post-term pregnancies have higher risks of perinatal mortality and neonatal morbidity, and labor is therefore often induced between 41 and 43 weeks of gestation.1–3 In Sweden, induction of labor is generally performed after 42 completed weeks.
Ultrasound biometry in the first half of pregnancy is recognized to be more reliable than the last menstrual period (LMP) as a predictor of the date of delivery.4–7 A randomized controlled trial has shown that the use of ultrasound as the method for estimating the date of delivery is associated with reduced rate of post-term pregnancies.8 These findings have also been confirmed by observational studies in Sweden9 and other countries.4,10–13
In Sweden, before 1980, fewer than 5% of hospitals practiced ultrasound scanning. Since 1990, ultrasound scanning has been offered to all pregnant women and 95% of them accept.9,14 Ultrasound data have been the exclusive basis for pregnancy dating in Sweden from the early 1990s, and this was further supported by a national recommendation from the Swedish Council on Technology Assessment in Health Care in 1998.15 The routine scanning is typically performed in early second trimester, and aims mainly at pregnancy dating. All clinical decision-making is based on the ultrasound estimation of the gestational age.
Assessment of gestational length by ultrasound in the second trimester is based on the assumption that fetuses of the same gestational age have equal size including biparietal diameter.16,17 Studies have, however, shown detectable differences in biparietal diameter already at the end of the first trimester with respect to fetal sex,18–20 maternal smoking,21,22 and early growth restriction.23 Thus, it is plausible that second-trimester ultrasound dating may lead to systematic errors in estimation of gestational age.
There has been no evaluation of possible discrepancies between female and male fetuses in risks of adverse birth outcomes when adjusting gestational age according to ultrasound.24,25 Given that ultrasound-dated female fetuses are more likely than male fetuses to be considered younger than actual gestational age (due to their smaller fetal size), we hypothesize that ultrasound dating may not only lead to differences in rates of post-term pregnancies with male and female fetuses,5,26,27 but also to sex differences in morbidity and mortality in pregnancies assessed as post-term. This would be because girls who are classified as post-term based on ultrasound may be more severely post-term than boys assigned the same gestational age. We used data from the Swedish Medical Birth Register to compare the perinatal outcomes of male and female infants born before and after the introduction of the ultrasound dating method.
The Swedish Medical Birth Register contains prospectively-collected information on more than 99% of all births in Sweden since 1973.28 All information is provided through antenatal, obstetrical, and neonatal records that are filled in by midwives and physicians. The register includes data about maternal sociodemographic characteristics, and information recorded prospectively during pregnancy, delivery, and the neonatal period. Diagnoses are classified and recorded by the treating physician according to the International Classification of Diseases (ICD) for 1973–1978 (ICD-8) and for 1995–2007 (ICD-9 and ICD-10).
Births were included if they were singletons born to mothers aged between 15 and 45 years, with valid birth dates for both mother and newborn, valid hospital codes, and gestational age from 39 weeks + 0 days to 43 weeks + 6 days between 1973 and 1978, and between 1995 and 2007. For all births, we included information on delivery date, level of hospital, maternal age, parity, sex of the newborn, and gestational age at delivery, as well as adverse perinatal outcomes related to post-term, namely intrauterine and neonatal death, Apgar score at 5 minutes, meconium aspiration (ICD-8 code 766.00, ICD-9 code 770B, and ICD-10 codes P24.0 and P24.9), seizures (ICD-8 codes 772.49 and 780.20, ICD-9 code 779A, and ICD-10 code P90), and nerve injury (ICD-8 codes 772.25–772.28, ICD-9 codes 767F-H, and ICD-10 code P14).
For calculating gestational age, antenatal records include information on date of last menstrual period and estimated date of delivery based on LMP as calculated by pregnancy wheel (adjusted to a cycle length of 28 days). In addition, from 1990, estimated date of delivery was available based on ultrasound assessment. For infants born from 1973 through 1978, gestational age at delivery was assessed using information from LMP while for those born from 1995 through 2007, ultrasound assessment of gestational age was used. Ultrasound examination was performed between 16 and 20 weeks in 52 clinics and between 10 and 15 weeks in 3 clinics,9 representing 97% and 3% of the study population, respectively. The ultrasound-based estimated day of delivery is recorded both in antenatal and obstetrical records. To ensure reliable estimates, we included only deliveries for which there was agreement within 14 days between these estimates, leaving 427,595 births in the first time period and 766,671 in the second time period (23,834 births excluded).
Term deliveries were defined as those occurring from 39 gestational weeks + 0 days to 40 weeks + 6 days, late-term as 41 weeks + 0 days to 41 weeks + 6 days, and post-term from 42 weeks + 0 days to 43 weeks + 6 days.
The data were first cross-tabulated according to newborn sex, gestational age, and the 2 time periods (1973–1978 and 1995–2007). Male-to-female newborn ratios were plotted by gestational week separately for the 2 time periods. We calculated rates of pregnancy outcomes by 1000 live births, according to infant sex, gestational age, and time period. For each of the 2 time periods, we used logistic regression to compare post-term and term infants with regard to risks of stillbirth (from 28 completed weeks), neonatal death (0–27 days), low Apgar score (<7) at 5 minutes, meconium aspiration, seizures, and nerve injury. In all comparisons, infant sex was included as an effect modification term, and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. For post-term female infants, we compared the change in risk for each outcome between the 2 periods by calculating the cohort ratio as the ratio between ORs for each time period. In the multivariable analyses, we adjusted for maternal age at birth, parity, and level of hospital (primary, secondary, or tertiary level). The statistical analyses were conducted with the SAS 9.1 software package (SAS Institute, Cary, NC).
The study was approved by one of the Regional Ethical Review Boards in Stockholm, Sweden.
Compared with the first time period (1973–1978), rates of post-term births (≥42 weeks) were lower in the second time period (1995–2007) (Table 1). This decline was more profound among girls (from 15.1% to 7.8% of all pregnancies reaching 39 weeks) than among boys (from 14.7% to 10.5%). During the later period, the mean difference between the estimations of gestational age by using information from LMP and ultrasound was 0.9 days in boys and 2.1 days in girls.
The newborn male-to-female ratios by gestational age at delivery changed remarkably between the 2 time periods (Figure). During the period when gestational age was assessed by LMP (1973–1978), the male-to-female ratio for those born at 39 gestational weeks or later remained constant around 1.0. After pregnancy dating from ultrasound assessments had been established as a routine for more than 5 years (1995–2007), the male-to-female ratio steadily increased from 39 weeks, reaching 1.60 at 43 gestational weeks.
Rates of adverse pregnancy outcomes generally increased by gestational age among both male and female births during both time periods (Table 2). Except for the risk of seizures among female infants in the early period and stillbirths among males in the later period, the risks of adverse pregnancy outcomes were higher for post-term births compared with term births (Table 2). Post-term female newborns tended to have lower rates of adverse pregnancy outcomes compared with their male counterparts during the early period, but higher rates in the late time period.
In the first time period, post-term-born females had, compared with post-term-born males, reduced risks of neonatal death, low Apgar score at 5 minutes, seizures, and nerve injury (Table 3). After the introduction of ultrasound, post-term-born females did not have reduced risk of any adverse perinatal outcome, but on the contrary, had higher risk for stillbirth and meconium aspiration than males. For all outcomes, the odds ratios for post-term females compared with males increased substantially (Table 3).
During the first period (1973–1978), the stillbirth rates in post-term boys and girls were approximately the same. During the second time period, the stillbirth rate among post-term females was 2.2/1000, while among post-term boys, it was 1.4/1000. The observed number of stillbirths among post-term females was 63, while if the rate for post-term males is applied, 41 stillbirths would have been expected. Among stillbirth cases in the second time period, 10% were post-term among males and 12% among females according to ultrasound assessment. According to LMP dating, the percentages were 18 and 25.6, respectively.
When ultrasound examination replaced LMP for assessment of gestational age in Sweden, risks of adverse perinatal outcomes increased among post-term-born female newborns, compared with post-term-born male newborns. Risks related to low Apgar score, meconium aspiration and seizures were higher among female compared with male post-term newborns, and risks related to the other outcomes (stillbirth, neonatal death, and nerve injury) were in the same direction. A secondary finding was the different sex distribution by gestational week in 2 different time periods in Sweden (1973–1978 vs. 1995–2007), indicating an increase in the proportion of males among post-term born infants in the second time period.
An important finding of this study is the difference between observed (n = 63) and expected (n = 41) numbers of stillbirths among post-term girls during the ultrasound-dated time period (based on observed risks for males). Thus, one-third of stillbirths among post-term girls today might be due to incorrect calculation of gestational age and consequent failure of inducing delivery at the proper time. This is further supported by the fact that the proportion of post-term males and females among stillbirths is different when using the LMP for calculation of gestational age (18% for males vs. 26% for females), compared with the ultrasound-derived estimate (10% vs. 12%, respectively). Stillbirth rates in post-term boys and girls were approximately the same in the first period (1973–1978).
After the implementation of ultrasound as the method of estimating date of delivery, the overall rates of post-term pregnancies were dramatically reduced, both in Sweden and in other countries.4,11–13 In accordance with a previous study,27 we found that dating by ultrasound instead of LMP changes the proportion of male and female newborns classified as post-term.27 Consequently, routine ultrasound assessment in the second trimester seems to cause a systematic misclassification of gestational age by sex, and in particular reduces the rate of post-term pregnancies with female fetuses. If post-term pregnancies with female fetuses are misclassified as term pregnancies because of dating error, they may not receive the same supervision as clinically recognized post-term pregnancies. One may speculate that pregnancies with female fetuses classified as post-term might indeed be very post-term and entail particularly high risks of perinatal morbidity and mortality.
When investigating the possibility of other explanations for the different representation of male and female newborns in the 2 time periods, the characteristics of prenatal care in Sweden must be considered. Sweden is a country where the antenatal determination of sex by ultrasound is uncommon, even when feasible. Sex is very rarely recorded in the antenatal records, and, except for the rare cases of sex-linked diseases, fetal sex does not determine the degree or type of perinatal care. It is also highly unlikely that midwives differentially recorded perinatal outcomes after the determination of sex postnatally in the 2 time periods. Differences in reporting accuracy and induction protocols between the 2 time periods are not likely to be alternative explanations because these would affect newborns of the 2 sexes to the same extent.
We used newborn sex as the most common determinant of differential size between fetuses in the second trimester to assess the clinical implications of misclassification of gestations in relation to post-term delivery. One might consider several other situations where this phenomenon might also apply, eg, fetuses with early fetal growth restriction29 or fetuses to mothers who smoke.21,22 In addition, genetically-small fetuses may incorrectly be categorized as term when they in reality are post-term. If such pregnancies are not supervised properly, their risks of adverse perinatal outcomes may be increased. It is possible that the high risk for such adverse outcomes among immigrant women in Sweden might in part be due to this phenomenon.30,31 A systematic misclassification of gestational age by ultrasound should also lead to proportionally fewer boys than girls being considered to be born preterm. Thus, if the “true gestational age” is shorter in boys compared with girls assessed as preterm, this may lead to an overestimation of male-to-female differences in perinatal health in preterm infants.
Presently in Sweden, where maternal care is provided free of charge for all women, 1 routine ultrasound scanning is generally offered and performed between gestational week 16 and 20. Detailed information on exactly which biometric parameters were used to calculate gestational age was not recorded in the Birth Register, but as all the parameters seem to be affected by fetal sex, they can cause a systematic error, when used for ultrasound dating.32–35 It is still unclear whether using crown-to-rump-length measurement in the first trimester for dating of pregnancies would decrease this systematic error, as some studies have reported a sex difference in fetal crown-rump length whereas another has not.32,36 Though ultrasound for assessment of gestational length in general is better than LMP and should be the preferred method for pregnancy dating, it is important to be aware of its shortcomings.4–7
The increased risk for perinatal mortality and morbidity among post-term pregnancies has led to proposals for routine induction before 42 gestational weeks. As evidence of associations between labor induction and birth outcomes is controversial,2 expectant management with fetal monitoring twice a week is an alternative. It has recently been suggested that it would be appropriate to let women make an informed decision about which management they prefer.24 Considering the differences in risks for several adverse perinatal outcomes between males and females classified as post-term, such a decision would, however, be difficult to make without fully informing mothers about the uncertainties of pregnancy-dating. It is also clear that a careful recording of the LMP is most crucial for both research and clinical decision-making.
To avoid the systematic error by ultrasound dating, which increases risks of adverse outcomes for post-term infant girls, we can see 3 alternative ways to refine estimation of the day of delivery: (1) ultrasound measurement of crown-rump length in the first trimester, as early developmental stages may be less susceptible to differences by sex, maternal smoking, or genetic determinants of growth, and thus be more invariant and accurate; (2) determination of fetal sex by ultrasound and use of this information when deciding the time for induction in late-term pregnancies; and (3) use of the combined information from LMP and ultrasound, so that pregnancies categorized as late-term according to ultrasound are induced if estimated day of delivery according to ultrasound was postponed for more than 1 week from the estimated day of delivery by LMP. Of these 3 alternatives, we believe that the use of the combined information from ultrasound and LMP in selected cases would be the most acceptable. Offering the ultrasound examination at the end of the first trimester might lead to a lower rate of detection of congenital anomalies. Determination of fetal sex and using sex-specific ultrasound biometric reference parameters does not solve the problem of misclassification of gestational duration for fetuses with early intrauterine growth restriction, fetuses to mothers who smoke, or genetically small fetuses.33
Among the strengths of this study is the use of a registry with highly reliable data.28 The study was population-based, and included more than 1 million births (n = 1,194,266), and controlled for possible confounding factors such as level of hospital, maternal age, and parity. Deliveries occurring before 39 completed weeks were not included in the current analysis, as this study focuses on post-term outcomes. Possible limitations include the lack of power when studying changes in risks for males and females over the 2 time periods for very rare outcomes, such as stillbirth, neonatal death, and nerve injury.
In conclusion, this study provides evidence that routine use of ultrasound in the second trimester for pregnancy-dating has reduced the proportion of females among post-term singletons pregnancies, but at the same time has increased the risks of adverse perinatal outcomes for female infants born post-term. Acknowledging this shortcoming for ultrasound, when used as a tool for pregnancy dating, it would be prudent to include the combined information from ultrasound and LMP for assessment of gestational length when handling late-term and post-term pregnancies.
We thank Ove Axelsson for valuable advice.
1. Sanchez-Ramos L, Olivier F, Delke I, Kaunitz AM. Labor induction versus expectant management for post-term pregnancies: a systematic review with meta-analysis. Obstet Gynecol
2. Gulmezoglu AM, Crowther CA, Middleton P. Induction of labour for improving birth outcomes for women at or beyond term. Cochrane Database Syst Rev.
3. Norwitz ER, Snegovskikh VV, Caughey AB. Prolonged pregnancy: when should we intervene? Clin Obstet Gynecol
4. Tunon K, Eik-Nes SH, Grottum P. A comparison between ultrasound and a reliable last menstrual period as predictors of the day of delivery in 15,000 examinations. Ultrasound Obstet Gynecol
5. Taipale P, Hiilesmaa V. Predicting delivery date by ultrasound and last menstrual period in early gestation. Obstet Gynecol
6. Neilson JP. Ultrasound for fetal assessment in early pregnancy. Cochrane Database Syst Rev.
7. Waldenstrom U, Axelsson O, Nilsson S. A comparison of the ability of a sonographically measured biparietal diameter and the last menstrual period to predict the spontaneous onset of labor. Obstet Gynecol
8. Waldenstrom U, Axelsson O, Nilsson S, et al. Effects of routine one-stage ultrasound screening in pregnancy: a randomised controlled trial. Lancet
9. (SBU) TSCoTAiHC. Routine ultrasound examination during pregnancy. Stockholm: The Swedish Council on Technology Assessment in Health Care (SBU); 1998:48.
10. Bennett KA, Crane JM, O'Shea P, Lacelle J, Hutchens D, Copel JA. First trimester ultrasound screening is effective in reducing postterm labor induction rates: a randomized controlled trial. Am J Obstet Gynecol
11. Hoffman CS, Messer LC, Mendola P, Savitz DA, Herring AH, Hartmann KE. Comparison of gestational age at birth based on last menstrual period and ultrasound during the first trimester. Paediatr Perinat Epidemiol
12. Gardosi J, Vanner T, Francis A. Gestational age and induction of labour for prolonged pregnancy. Br J Obstet Gynaecol
13. Savitz DA, Terry JW Jr, Dole N, Thorp JM Jr, Siega-Riz AM, Herring AH. Comparison of pregnancy dating by last menstrual period, ultrasound scanning, and their combination. Am J Obstet Gynecol
14. Hogberg U, Larsson N. Early dating by ultrasound and perinatal outcome. A cohort study. Acta Obstet Gynecol Scand
15. (SBU) TSCoTAiHC. Routine ultrasound examination during pregnancy. Stockholm: The Swedish Council on Technology Assessment in Health Care (SBU); 1998:95, 96.
16. Gardosi J, Geirsson RT. Routine ultrasound is the method of choice for dating pregnancy. Br J Obstet Gynaecol
17. Mongelli M, Wilcox M, Gardosi J. Estimating the date of confinement: ultrasonographic biometry versus certain menstrual dates. Am J Obstet Gynecol
18. Tunon K, Eik-Nes SH, Grottum P. The impact of fetal, maternal and external factors on prediction of the day of delivery by the use of ultrasound. Ultrasound Obstet Gynecol
19. Pedersen JF. Ultrasound evidence of sexual difference in fetal size in first trimester. Br Med J
20. Wald N, Cuckle H, Nanchahal K, Turnbull AC. Sex differences in fetal size early in pregnancy. Br Med J (Clin Res Ed)
21. Persson PH, Grennert L, Gennser G, Kullander S. A study of smoking and pregnancy with special references to fetal growth. Acta Obstet Gynecol Scand Suppl
22. Hanke W, Sobala W, Kalinka J. Environmental tobacco smoke exposure among pregnant women: impact on fetal biometry at 20–24 weeks of gestation and newborn child's birth weight. Int Arch Occup Environ Health
23. Thorsell M, Kaijser M, Almstrom H, Andolf E. Expected day of delivery from ultrasound dating versus last menstrual period—obstetric outcome when dates mismatch. BJOG
24. Mandruzzato G, Alfirevic Z, Chervenak F, et al. Guidelines for the management of postterm pregnancy. J Perinat Med
25. Tunon K, Eik-Nes SH, Grottum P. Fetal outcome in pregnancies defined as post-term according to the last menstrual period estimate, but not according to the ultrasound estimate. Ultrasound Obstet Gynecol
26. Smulian JC, Campbell WA, Rodis JF, Feeney LD, Fabbri EL, Vintzileos AM. Gender-specific second-trimester biometry. Am J Obstet Gynecol
27. Henriksen TB, Wilcox AJ, Hedegaard M, Secher NJ. Bias in studies of preterm and postterm delivery due to ultrasound assessment of gestational age. Epidemiology
28. Cnattingius S, Ericson A, Gunnarskog J, Kallen B. A quality study of a medical birth registry. Scand J Soc Med
29. Goldenberg RL, Davis RO, Cliver SP, et al. Maternal risk factors and their influence on fetal anthropometric measurements. Am J Obstet Gynecol
30. Essen B, Bodker B, Sjoberg NO, et al. Are some perinatal deaths in immigrant groups linked to suboptimal perinatal care services? BJOG
31. Saastad E, Vangen S, Froen JF. Suboptimal care in stillbirths—a retrospective audit study. Acta Obstet Gynecol Scand
32. Moore WM, Ward BS, Jones VP, Bamford FN. Sex difference in fetal head growth. Br J Obstet Gynaecol
33. Schwarzler P, Bland JM, Holden D, Campbell S, Ville Y. Sex-specific antenatal reference growth charts for uncomplicated singleton pregnancies at 15–40 weeks of gestation. Ultrasound Obstet Gynecol
34. Johnsen SL, Rasmussen S, Sollien R, Kiserud T. Fetal age assessment based on femur length at 10–25 weeks of gestation, and reference ranges for femur length to head circumference ratios. Acta Obstet Gynecol Scand
35. Caughey AB, Snegovskikh VV, Norwitz ER. Postterm pregnancy: how can we improve outcomes? Obstet Gynecol Surv
36. Becker S, Ural S, Fehm T, Bienstock J. Fetal gender and sonographic assessment of crown-rump length: implications for multifetal pregnancy reduction. Ultrasound Obstet Gynecol