gms | German Medical Science

For more than 200 years now, we have been using different methods to try and determine the estimated date of delivery for a pregnancy, initially by way of calculation and later also obstetric ultrasonography [21]. Originally intended to provide a rough idea of the pregnancy stages, an estimated date of delivery (EDD) during the 40^th week of pregnancy (40+0) is increasingly being seen as a strict cut-off and it is perceived as a risk for a pregnancy to continue beyond this.

Over the last ten years, the number of post-term inductions has increased dramatically. In 2008, of the 658,201 hospital births in Germany, 36,395 were post-term inductions [4], which equates to 5.53 percent. In 2017, however, of the 743,241 hospital births in Germany, 93,632 of them were post-term inductions [15]. This corresponds with a 12.6 percent share.

In 2010, for the first time, the German Society for Gynaecology and Obstetrics published a guideline on how to proceed in the event of an overdue or post-term pregnancy. This version of the guideline [9], which is currently still valid, recommends that, induction of labour be suggested for a pregnant woman going beyond her EDD by ten days, even if there is no clinical evidence of a pathology. Here, decisions regarding obstetric care are heavily based on the EDD.

However, there has recently been mounting evidence that the methods used for determining the due date contain significant sources of error. This applies to the Naegele's rule [20] method of calculation but particularly also to ultrasound measurement techniques [18], [22], [34]. Moreover, current data suggests that the average gestational age of 280 days from the start of the last menstrual period (LMP), which is the standard used to calculate the EDD, is too short [16], [19]. In France, for instance, the EDD is based on a gestational age of 41+0 weeks [6]. Although the due date is generally estimated using the gestational age based on the menstrual cycle or by mapping an ultrasound measurement on a chart with a broad confidence interval, it is regarded as a diagnosis and deemed authoritative [24]. The fact that only a very small percentage of pregnancies actually end on the EDD [17] is ignored here. It is assumed that deviations from the EDD are down to calculation inaccuracies rather than an individual foetal maturation process. The notion that pregnancies might possibly proceed according to their own individual rhythm and the maturity of a newborn does not necessarily have to correlate with the estimated gestational age is not discussed.

In preparation for this study, during the period between November 2016 and April 2017, an extensive online review of the literature was conducted using the keywords maturation, maturity, newborn, gestational age, length of gestation, due date, prolonged pregnancy and post-term pregnancy, and their equivalent terms in German, with the necessary combinations of these terms. The databases used were Medline, Old Medline and the Cochrane Library (Wiley). No search restrictions were applied in terms of time period or language.

The aim of the study was to scrutinise whether the EDD had sufficient explanatory power to predict the physiological end of a pregnancy and the birth of a mature newborn.

The prospective clinical study was conducted in 2017 at a university obstetric hospital with approximately 2,400 births per year. The principal author of the study determined the morphological and neurological maturity of 100 newborns within the first 72 hours after delivery. An assessment of the clinical signs of maturity of a child seemed to be a suitable method for verifying the validity of the EDD for the individual newborn. Based on the available literature, we selected the New Ballard Score as a positively evaluated instrument for the most accurate possible maturational assessment of a newborn [1], [3], [5], [12], [28], [31]. The scoring used by this instrument is based on six physical and six neuromuscular criteria. The Ethics Commission of the Institutional Review Board at Ludwig-Maximilians-Universität (LMU) in Munich (Ethikkommission der Medizinischen Fakultät der Ludwig-Maximilians-Universität München) gave the study a positive evaluation.

By setting the minimum estimated gestational age at 259 days after the start of the LMP (37+0), premature births were excluded. Due to the influences on the intra-uterine maturational processes described in the literature, which cannot be fully explained using gestational age alone, the principal author stipulated the following additional exclusion criteria: maternal gestational diabetes or diabetes mellitus [13], [26], maternal hypertonia or pre-eclampsia [30], [35] and multiple birth [7], [25]. Newborns from non-Caucasian ethnic groups were excluded due to the described deviation in the New Ballard Score [2], as were children in a generally poor condition who could not be subjected to an additional examination.

The principal author had access to the neonatal unit, the files of the newborns and also to the parents and newborns themselves who were mostly with their mothers for rooming-in. Generally, the parents of the child were informed verbally and in writing the day before the examination and their consent was obtained. Here it was emphasised that the children would not be subject to undue strain due to the examination. Even on the actual day of the exam, the parents had the option of refusing participation at any time despite the fact that they had, in principle, already given their consent in advance. First, based on the admissions list from the neonatal unit I created a list of all newborns who could be considered for an examination. Additionally, the exclusion criteria were discussed with the parents again to rule out any gaps in the documentation. The EDD was taken from the Mutterpass (maternity record) in each case. If the EDD was corrected over the course of the pregnancy, the new EDD was used.

Ballard et al. found no differences in the results of postnatal maturation assessments conducted up to 96 hours after birth (with the exception of premature babies born earlier than 27 weeks) and thus did not specify a preferred time of examination [5]. Nevertheless, for the purposes of this study, 72 hours was stipulated as the maximum age of the newborn being examined.

• Maternal data: Age, parity, height, pre-pregnancy weight
• Birth: How labour started, breech presentation (according to the New Ballard Score, in this case examination must be >24 hours postpartum)
• Child: EDD according to maternal record, sex, age at time of examination (in hours), gestational age based on the New Ballard Score

The sample comprised 100 children; given the limited scope of a master’s thesis, a power computation was not conducted beforehand. The data collected on the total of 100 children in the sample enabled us to assess both the estimated gestational age and maturity based on a neonatal examination, depending on the child’s sex. Moreover, the study verified the correlation between the maturity according to the neonatal examination and the estimated gestational age and explored whether there were any differences between the maturation of the children born via spontaneous labour and the group of children whose delivery was not spontaneous.

The principal author analysed the data using SPSS 24. The analysis provided key figures, distributions and standard deviations. Cross tables were created and the relative risk was determined. The specified level of significance was p<0.05.

The arithmetic mean of the estimated gestational age in the sample was 280.37 days, the median was exactly 280 days (40+0). The lowest value in the sample was 263 days after the LMP (37+4), the highest was 300 days after the LMP (42+6). For a total of 49 percent of the sample, their pregnancies went beyond the EDD. For n=4 of the women in the sample, this was officially classified as a post-term pregnancy (over 294 days after the LMP (42+0)); n=13 went beyond their EDD by more than ten days and n=18 by more than eight days.

The average gestational age according to the New Ballard Score of the total sample (N=100) was 278.4 days. Although the male newborns (n=52) had an average gestational age of 277.15 days thus making them 2.6 days less mature than female newborns, this difference was not significant at p=0.127.

At first glance, the estimated gestational age of the children in the sample according to the calculation was only marginally lower than the gestational age revealed by the maturation assessment. The mean difference between the gestational age according to the New Ballard Score and the calculated gestational age was -1.970 days. However, in the sample there was a significant difference between these two parameters of t₍₉₉₎=2.015 and p=0.047. What is striking here is that the children born after their EDD showed a larger deviation from the calculated gestational age than the children born before their EDD. A separate analysis of the data for births up to and including, or after the EDD showed that for births up to and including the 40th (40+0) week of pregnancy (n=51) the calculated gestational age conformed with the gestational age according to neonatal examination, however this finding was not significant as t₍₅₀₎=-1.064 and p=0.292. Nevertheless, it was evident that the earlier babies were born before the calculated due date, the more mature they were found to be according to the New Ballard Score. For births in the sample which, according to the EDD, took place after the 40th (40+0) week of pregnancy (n=49), the estimated gestational age deviated significantly at t₍₄₈₎=3.990 from the gestational age revealed by the neonatal examination (p<0.001). The conclusion we can draw from this data is that the longer after the EDD the children are born, the less mature they are as compared to the estimated gestational age (see Figure 1 [Fig. 1]). When we tested for the correlation between the gestational age reached and the difference between maturity according to neonatal examination and the calculated gestational age, a significant negative correlation of r=-0.591 and p<0.001 was observed (see also Figure 2 [Fig. 2]).

Breaking down the data on the discrepancy between the gestational age based on neonatal examination and the estimated gestational age by how labour started (spontaneous or induced) reveals considerable differences. In the case of a spontaneous start of labour (n=77), the difference between the gestational age based on neonatal examination and the estimated gestational age was insignificant at -0.16 days. For children delivered via primary Caesarean section (n=10), this difference was, on average -2.2 days and for induced deliveries (n=13), on average, -12.54 days. If we look separately at the group of births that were post-term inductions (n=11), the deviation increases to -14.55 days. The t-test indicated a significant difference between spontaneous and induced deliveries (p=0.001 at t₍₉₈₎=3.594) when it came to the variance between maturity according to the neonatal examination and the estimated gestational age. For the induced labours, the probability of an immaturity that deviates from the expected gestational age by >7 days was higher than for labours that began spontaneously (RR=3.35; [95% CI 1.89-4.15]).

Within the framework of the study, it was not possible to examine all the children born during the period of investigation that did not meet any exclusion criteria. The sample was purely random (subject to the time constraints of the principal author, the availability of the parents and linguistic barriers) and cannot be considered representative. Nevertheless, it is expected that the trend indicated by the present study would also be observed in a larger scale study.

In order to ensure that the principal author was not aware of the estimated gestational age when checking for the exclusion criteria of prematurity, children whose files contained exclusion criteria were firstly listed separately by name. This was later compared with the overall list of children and the children concerned were excluded from the sample. This was a necessary step to avoid a direct connection being made between the name and the gestational age of the child. Although Smith et al. determined that knowing the estimated gestational age did not influence the result of the examination according to the New Ballard Score [28], this precautionary measure was still taken in order to minimise the risk of a confirmation bias.

The findings fundamentally challenge the correlation between the estimated gestational age and maturity according to neonatal examination which was previously thought to be firmly established. Although the scores for establishing the gestational age of a newborn were not developed with the purpose of verifying diagnoses made during maternity or prenatal care, the use of the New Ballard Score nevertheless provides us with an opportunity to scrutinise the unambiguous finding of the present study using a larger sample. The finding being: Gestational age does not reflect the maturity of a child, particularly in births occurring after the EDD.

Due to the low number of cases, the apparently substantial correlation between induced labour and deviation from the gestational age according to neonatal examination must be qualified. The group of inductions and planned c-sections (n=23) is small and, in 11 cases, there is also an overlap with women from the group whose labour was overdue by more than one week and one day (41+1). Consequently, it cannot be ascertained whether it is the induced labour or the overdue pregnancy that is correlated with the marked deviation from the gestational age according to the neonatal examination. However, no effect of the deviation is observed in children born by spontaneous delivery (n=77), despite the fact that there were also nine children in this group who were overdue by more than one week and one day (41+1). One possible conclusion could be that with the processes leading to a spontaneous delivery, the foetus undergoes an additional maturational spurt. This question would have to be addressed by future research.

Although compared to other similar instruments the New Ballard Score can be described as a good assessment tool, questions still remain about its explanatory power. In earlier studies it was already determined that, in assessments of newborns using the New Ballard Score, the maturity of children born before their EDD tended to be overestimated and that of children born after their EDD, underestimated [3], [31]. It must therefore be discussed whether the method selected to determine maturity (New Ballard Score) basically always shows deviations or whether the deviations in maturity from the estimated gestational age were correctly identified. Studies evaluating methods of determining maturity have so far not assessed this aspect [8], [10], [12], [27], [29], [31], [32]. However, it is not only neonatal examinations according to the Ballard Score that indicate this alleged overestimation of the gestational age of infants born before their EDD and underestimation for those born after their EDD but rather this also occurs when other methods are used to establish maturity [5], [10], [11], [23]. Even with different collectives, different study settings and using different examination methods [8], [27], [29], [32] a consistent trend can be observed, which corresponds with the findings presented here. This would appear to suggest that the maturation processes that take place during pregnancy are not always congruent with the estimated gestational age.

The present study shows how inadequate the available methods of calculating the gestational age actually are in terms of accurately reflecting the clinical process occurring in utero which might possibly be specific to each individual but in any case certainly differs from the estimate. There is an urgent need for research into both the issue of actual duration of pregnancy and the robustness of the EDD as a diagnostic value as well as into the assumption that has been made to date that a child’s maturation process in utero runs linear to the estimated gestational age and that this process can be diagnosed using calculations. Moreover, further research is required to address the reasons for growth or maturation retardation in utero [14].

It should be reconsidered whether the steps taken in clinical practice when a pregnancy goes beyond the EDD should be guided by a method of calculation, the validity of which is questionable. Undoubtedly, there are children affected by dysmaturity. As shown by a recent study [33], for this small number of children, it is important that their condition is identified and that labour is induced. For the remaining majority of children, however, it could be the case that the necessary physiological length of pregnancy is longer or, due to their growth rhythms, simply differs from our expectations. The maturation of a child cannot be calculated before it is born and, to date, cannot be satisfactorily diagnosed.

Determining gestational age to the day is certainly not required. Thus, determining an exact date on which a pregnancy should end and labour begin should only be relevant in a legal (employment) context such as the start of maternity leave, and should not remain a medical necessity.

The authors declare that they have no competing interests.