The purpose of this study was to find out the frequency of antibiotic use during term deliveries in a developed country as well as the indications and types of antibiotics used in a comprehensive study cohort. The other aim was to find out whether the antibiotic use is associated with infant health during the first three months of life. The data about antibiotics during delivery were collected from the maternity records and combined with the data from the online survey of the participants of the prospective HELMi cohort [16]. All 1019 infants were born between March 2016 and March 2018 at gestational weeks 37—42 without known congenital defects.
Overall, antibiotics were used during 36.6% of all deliveries. The number is slightly larger than in a corresponding Danish study including 706 women, where the frequency was 33% [18] and slightly smaller comparing to frequency in the US, > 40% [19]. In our study, 23.7% of the mothers who gave birth vaginally received antibiotics. Comparable to that, in a Canadian cohort study including 198 infants, 27% of vaginally born infants were exposed to intrapartum antibiotics mostly due to GBS prophylaxis or PPROM [20]. In our cohort, the indication for antibiotics was non-prophylactic in less than 5% of the deliveries.
In our cohort, all mothers with intended vaginal delivery were screened routinely for GBS with PCR-based test at the time of admission to delivery ward, and 17% of all mothers received antibiotics for proven or unclear GBS carriage status. The number is equivalent to a worldwide amount of GBS carriers (18%) [4]. GBS screening policy is heterogeneous [5]. In a study covering 95 countries, approximately two thirds of countries reported a GBS screening policy, most often based either on rectovaginal culture or clinical risk factors and only few reported point-of care PCR-screening [5]. In maternity hospitals in Helsinki and Uusimaa district PCR-screening performed by midwives by point of care test is used, since it proves rapid information of the current GBS status of the mother, is easy to perform and effective in preventing GBS early-onset disease in infants (0.15 cases during first 72 h/1000 live births between years 2016 and 2017). While discussing the potential side effects of antibiotics, it should be acknowledged that intrapartum antibiotic prophylaxis is the only effective treatment so far to prevent GBS early-onset disease and therefore effective screening at the time of delivery is necessary [21].
All mothers who gave birth by CS received antibiotics as prophylaxis due to birth mode as recommended in international guidelines, since prophylaxis reduces the incidence of maternal wound infections and endometritis [22, 23]. The effect is especially profound in women undergoing emergency CS and lower in non-laboring women with intact membranes undergoing elective CS [23]. In our study, CSs covered 16.9% of the deliveries corresponding to the national rate of 16.7% in Finland [24]. This is a slightly smaller number than a corresponding recent number of Northern Europe 22.4%, worldwide percentage being 18.6% but with large country-specific variations [25]. The proportion of CS, however, has been on the rise during the last decades [25]. Reasons for the phenomenon are complex, varying from maternal characteristics to ethical considerations [26]. As a result, an increasing number of infants are exposed to antibiotics during delivery [19].
Administration of prophylactic antibiotics after the cord is clamped can prevent infant´s exposure to antibiotics during CS. However, current WHO recommendation is to administer prophylactic antibiotic before surgical incision based on meta-analyses concluding that this policy reduces the risk of surgery [27]. On the contrary, recent large Swiss study did not confirm this finding [28]. At the time of our study, local hospital guidelines instructed to start prophylactic antibiotics 30–60 min before incision.
In our study, higher pre-pregnancy BMI of the mother was statistically significantly associated with a higher frequency of CS, and therefore also the use of intrapartum prophylactic antibiotics. Other studies have shown that delivery progresses more slowly when maternal BMI increases [29, 30], leading more frequently to emergency CSs.
Intriguingly, the antibiotic use during delivery was associated with gastrointestinal function of the infant. Defecation frequency of the infant during the 17-week follow-up period was increased significantly when the mother had received intrapartum antibiotics. The difference may not be clinically relevant, but supports the hypothesis that antibiotics used during delivery affect the gastrointestinal tract and infant microbiota [31, 32]. During the first three weeks, those infants who received antibiotics during vaginal delivery had higher defecation frequency comparing to infants born by CS, which may indicate that antibiotics received just before birth during CS may not modify the infant microbiome as strongly as antibiotics received earlier. The finding may also be explained by different the types of antibiotics used during delivery since benzylpenicillin was more frequently used in vaginal deliveries and cefuroxime in CSs. In addition, the median dosage of antibiotics during delivery was 2 among vaginal deliveries and 1 among CSs supporting the result. Importantly, intrapartum antibiotics did not associate with greater amount of crying or did not carry along an increased risk for impaired health of the infant during the first 17 weeks.
Earlier studies of the association between the growth of the infant and antibiotics during pregnancy and early age have reported that there is some variation between different antibiotics. For example, especially macrolides and beta-lactams such as penicillin associated with a greater weight gain during infancy and childhood [33,34,35]. This may be explained by variable transplacental transmission of different antibiotics. It also seems that the timing of the antibiotic course is of major importance. For example, one study reported that antibiotics taken during the first week of life were associated with smaller growth whereas the later use of antibiotics during the first year was associated with a greater growth and weight gain [36].
We observed a positive association between intrapartum antibiotics and the weight of the infant at the age of three months. The infants who had been exposed to antibiotics during delivery weighted on average 140 g more than the others and the finding was not explained by the birth mode. Although the weight difference disappeared by four months the results are in line with previous studies suggesting that the antibiotics increase weight gain in children most likely through modification of the intestinal microbiota [35].
Strengths and limitations
To our knowledge, this study cohort of term pregnancies included in the analyses of antibiotic use is one of the largest reported. This cohort represents well the general population as the number of CSs (16.9%) corresponded accurately to the number of CSs in Finland (16.7% in 2017 [24]). Also, the data were comprehensive and reliable, collected from one hospital district with uniform patient charts and clinical practice. Notably, the data on infant health was prospectively collected on weekly online questionnaires covering the 17-week observation period [16]. As a limitation, we did not have the data on infant heights at three months of age. Importantly, there were no major health concerns related to infant health when being exposed to antibiotics during delivery.
As another limitation, there was slight uncertainty of the data recorded in patient charts of a few mothers. In such cases, the entire maternity record was thoroughly reviewed. However, missing data were rare and occasional unclear markings did not affect the main results.
