Did the first wave of the COVID-19 pandemic impact the cesarean delivery rate? A retrospective cohort study at a primary care center in Switzerland

Introduction

On February 25, 2020, the first case of coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), was detected in Switzerland. On March 11, 2020, the World Health Organization (WHO) declared COVID-19 to be a pandemic [1], and on March 16, 2020 the Swiss government implemented restrictive measures aimed at controlling the spread of the disease [2].

The COVID-19 pandemic has caused a wide range of healthcare challenges, including in pregnancy and birth. Several studies have examined the effect of COVID-19 on pregnancy [3]. For example, in some countries infected women had a higher than expected rate of cesarean delivery (CD) [4], whereas other studies showed that the overall CD rate was either unchanged compared to the pre-pandemic rate [5, 6] or even lower [7, 8]. Whether infection with COVID-19 at the time of delivery is an indication for CD is unknown [9] as there is no clear evidence of vertical transmission [10]. Nonetheless, COVID-status alone seems to be a common indication for CD [11]. In women with severe COVID-19, CD might reduce the risks for both mother and newborn. By contrast, the influence of COVID-19 on the management of pregnancy and birth and on the respective outcomes in uninfected women has been the focus of only limited research.

Since 1985, the WHO has recommended a CD rate no higher than 10–15%, as higher rates do not correspond to significantly better outcomes for either mothers or newborns [12]. The average annual rate at our institution since 2015, when monitoring was implemented, has been 36%. This high rate can partly be explained by the fact that our institution is a referral center with a neonatal intensive care unit (NICU), such that many of its obstetrics patients are those with high-risk deliveries, such as breech and twin births. However, during routine compliance with whom recommendations of regular internal quality controls to assess changes in the CD rate [13], a significant decrease during the first 3 months of the COVID-19 pandemic was observed at our hospital. Thus, the aim of this study was to determine the potential impact of the pandemic on the CD rate, by comparing pregnant patients and their obstetric risk factors before and during the pandemic. We hypothesized (1) that the CD rate decreased during the first wave of the pandemic compared to the same seasonal period in the two years before the pandemic and (2) that the decrease was due to consequences of the pandemic for the hospital setting rather than due to a change in the patient population. We also examined the specific effects of the experience of the obstetrics team and the time of birth in lowering the CD rate.

Subjects and methods

Results

The flowchart (Figure 1) shows the specifics of the data collection, including excluded cases and the data available for the analysis.

Figure 1:

Flowchart of data collection and the analysis, showing excluded cases and data availability.

Between March 1 and May 31, 2020, there were 597 registered deliveries at our hospital, of which 179 were by CD (30%). This result corresponds to an absolute risk reduction of 8.7% (95% confidence interval [95%CI]: 13.3–3.9, p<0.004) compared with the CD rate of the control group, consisting of deliveries from March 1 until May 31 in 2018 and 2019 (CD rate: 38.7%, n=411, total number of births=1,063). The unadjusted OR for CD during vs. before the pandemic was 0.68 (95%CI: 0.55 to 0.84, p=0.0004).

The demographic characteristics, basic health data, and obstetrics data of the patients in the two groups were similar (Table 1). The percentage of patients with a university degree was higher during than before the pandemic period (41.4 vs. 35.6%). An analysis of the delivery characteristics during the pandemic period showed a significant increase in the preterm premature rupture of membranes compared to the control group (3.0 vs. 1.1%) (Table 2). The lower CD rate during the pandemic than in the two prior years applied to the proportion of primary CD (18.0 vs. 13.4%) and secondary CD (20.7 vs. 16.6%). There were more midwife-led births in the pandemic period (5.7 vs. 3.7%) and the births were more often attended by resident physicians or midwives than by attending physicians, chief of service and the department leader during the pandemic period. The number of years of experience of the physician present at birth was slightly lower during the pandemic.

During both periods, nulliparae, induced labor in nullipara, primary CD in nullipara, and previous CD (Robson groups 1, 2A, 2B, and 5) accounted for more than two-thirds of the total number of CDs (Table 3). However, during the pandemic period there were large reductions in the CD rates of all these groups paralleled by an increase in induced deliveries (Robson 2A and 4A).

Neonatal outcome was similar between the two periods, but average umbilical artery pH was lower in the pre-pandemic period as was the rate of infant respiratory distress syndrome during the pandemic period (2.5 vs. 5.8%, Table 4). There was one stillbirth during the pandemic and four stillbirths during the two prior years.

Figure 2 shows the declining CD rate starting in mid to late March 2020 (weeks 3, 4), corresponding to the start of the lockdown in Switzerland, until the end of April and into May 2020.

Figure 2:

CD rates during the pandemic (between March 1 and May 31, 2020) and pre-pandemic (same dates in 2018 and 2019) periods. (Program used for Figure: R, version 4.0.3.).

The unadjusted OR for CD of the pandemic vs. the pre-pandemic cohort was 0.68 (95%CI: 0.55–0.84, p=0.0004). When adjusted for obstetric characteristics and known risk factors for CD, including induction of birth, breech, gestational diabetes mellitus, parity, use of reproductive assistance procedures, psychiatric diagnosis, suspicious or pathological CTG, twin delivery, preeclampsia, patient age, pregnancy duration, the OR and 95% CI were nearly the same (adjusted OR: 0.68, CI 0.53–0.89; p=0.005, Table 5).

In a model additionally fitted for the case lead’s position in the team and the time of delivery, the association was slightly weaker (OR=0.73, 95%CI 0.54–0.99, p=0.04, Table 6).

During the pandemic period of the study, there was no universal testing for COVID-19 except between April 7 and April 24, 2020. Within this period, only one patient tested positive for COVID-19. The patient was asymptomatic and, as planned before knowledge of the positive test result, she underwent scheduled primary CD because of a previous CD. In all other patients, the occurrence of asymptomatic infections and thus the number of undetected cases of COVID-19 in the 3 months from March to May is unknown.

Discussion

Between March and May 2020, the CD rate in our hospital decreased to 30% compared to the pre-pandemic rate of 38% (unadjusted OR 0.68, 95%CI: 0.55–0.84), which confirms our first hypothesis that the CD rate decreased during the first wave of the COVID-19 pandemic. The fact that obstetric characteristics and known risk factors for CD were comparable between the pre-pandemic and pandemic patients, and that adjusting these characteristics led to a very similar odds ratio estimate (adjusted OR: 0.68, CI 0.53–0.89) suggests that the decline in CD was not due to a change in the patient population. Further adjustment for the case lead’s position in the team and the time of delivery resulted in a slightly weaker, but still similar odds ratio estimate (OR=0.73, 95%CI 0.54–0.99). This suggests that these two factors may have contributed to the decrease but cannot explain the decrease alone. These results may be interpreted as indirectly confirming our second hypothesis, that the decrease in CD was due to consequences of the pandemic for the hospital setting rather than due to a change in the patient population. We can only speculate that there were changes in other, unmeasured characteristics in the hospital setting, which were not considered in our study.

Obstetric care by resident physicians and midwives, which was more frequent during the pandemic period, may have been advantageous. This result might be biased since the records might not have consistently included any help that was received by attending physicians, chief of service or the department leader. Nevertheless, the composition of the team might have contributed to the reduced CD rate. However, an experienced physician may have been present at birth because one was called or because of the pregnant woman’s insurance model, which in Switzerland comes with the attendance of a higher ranked physician at birth if the private or semi-private model is chosen.

By contrast, neonatal outcomes were very similar during the two periods. The difference in umbilical artery pH, while significant, was not of clinical importance since, after including the SDs, both values were physiological. Moreover, the overall neonatal outcome was clinically encouraging, with reductions in the rates of NICU admissions and acute respiratory distress syndrome during both periods.

Most studies on the impact of COVID-19 on obstetric care have concentrated on SARS-CoV-2-positive women, but the impact on uninfected pregnant women, and specifically on the CD rate in this group, has been largely ignored. The few authors who examined the overall CD rates at their institutions reported varying results, with higher CD rates in one study [4] and stable CD rates in two others [5, 6]. At the time of submission, only a study from Iceland and one from the USA were similar to ours in their design and aims. The study from Iceland found a reduction of the CD rate between March and May 2020, especially in the rate of elective CD [7]. In the Boston (USA) study, the CD rate between April and July 2020 likewise declined, especially with respect to CDs performed without a trial of labor [8]. Neither study could unequivocally determine the mechanism underlying the reduction, although Einarsdóttir et al. [7] suggested the prioritization of maternity care whereas Sinnott et al. [8] found a lower rate of CD without a trial of labor.

Our study identified a reduction in the CD rate in the three largest Robson groups, i.e., fewer CDs in nullipara, following the induction of labor, and in women with repeat CD. Contrary to the findings of Einarsdóttir et al., there was no decrease in scheduled CD, but there was a more successful trial of labor in nullipara patients, in patients in whom labor was induced, and in those with a trial of labor after a previous CD. The combination of successful inductions of labor and a small rise in instrumental vaginal deliveries could in part explain the lower secondary CD rate. A similar effect was determined in the literature review by Middleton et al. [18], who found that labor induction concluded in fewer CDs and more instrumental vaginal deliveries.

At least at our hospital, during the pandemic physicians may have been even more aware of the negative consequences of CD, given the risk associated with any surgical procedure and the lack of space in the adult intensive care unit (ICU), because of the growing number of COVID-19 cases. This consideration might have led to the (unconscious) decision to choose vaginal delivery over CD for their patients. The higher rate of successful trials of labor supports this hypothesis but definitive evidence is lacking.

The length of hospital stay is much shorter after a vaginal birth than after a CD. Thus, with the restrictions posed by the pandemic, such as with respect to visitors, both patients and physicians may have been more willing to attempt a vaginal birth. This was reflected in the strongly reduced number of overnight stays after birth during the pandemic than during the pre-pandemic period (Table 2). It is important to note that the patients where at no point encouraged to leave the hospital earlier.

With the increasing number of COVID-19 patients in the adult ICU and thus the need for more attending physicians to care for them, senior physicians might have occasionally filled in on the delivery ward. However, whether the composition of the obstetrics teams was higher overall could not be determined. The “personality of the case leader” as well as her/his abilities, knowledge, and handling of emergencies might also have played an important role in lowering the CD rate, but this could not be measured.

A lower CD rate is a desired change in a hospital like Triemli Hospital, where the CD rate is well above the 10–15% recommended by the WHO. The experience during the early phase of the pandemic showed that both a lower rate and a stable neonatal outcome are possible. Accordingly, further efforts should be made to identify the conditions that allow a reduction of the CD rate.

The strengths of this study include its large sample size and the fact that data collection and analysis were done by the same investigator, which ensured that all deliveries were similarly assessed. The limitations include the lack of evaluation of maternal outcome and the difficulty in generalizing our results, as the circumstances at every hospital differ, especially hospitals in different countries. Thus, whether the reduction in the CD rate and the underlying reason(s) will also apply to other hospitals remains to be determined.

Conclusions

The CD rate at our hospital was considerably lower during the first 3 months of the COVID-19 pandemic than during the same months in the two previous years, without a corresponding increase in adverse neonatal outcomes or emergency CD. Instead, the decrease was accompanied by significantly more successful trials of labor in nullipara, in patients with induced labor, and in vaginal birth after a previous CD. The change in the composition of the obstetric team may also have contributed to the reduction in the CD rate. The reasons for the decline in the CD rate may also include a difference in the behavior the obstetrics team and in patients during the pandemic, but this remains somewhat speculative. Further research is therefore needed to address the possible factors that can lower the CD rate. As the CD rate will remain an important issue in the years to come, the consequences of any change with respect to the health and well-being of both mothers and newborns, especially during a pandemic, should be carefully evaluated.