Volume 100, Issue 4 p. 684-693
ORIGINAL RESEARCH ARTICLE
Open Access

Augmentation of labor with oxytocin and its association with delivery outcomes: A large-scale cohort study in 12 public hospitals in Nepal

Helena Litorp

Corresponding Author

Helena Litorp

Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden

Department of Global Public Health, Karolinska Institutet, Stockholm, Sweden

Correspondence

Helena Litorp, Department of Women's and Children's Health, Uppsala University, Sjukhusvagen 7, 753 09 Uppsala, Sweden.

Email: [email protected]

Search for more papers by this author
Avinash K. Sunny

Avinash K. Sunny

Golden Community, Lalitpur, Nepal

Search for more papers by this author
Ashish Kc

Ashish Kc

Department of Women's and Children's Health, Uppsala University, Uppsala, Sweden

Society of Public Health Physicians Nepal, Kathmandu, Nepal

Search for more papers by this author
First published: 19 May 2020
Citations: 14

Funding information

The study was funded by Grand Challenges Canada, Canada; Swedish Research Council, Sweden; Laerdal Foundation for Acute Medicine, Norway, and Einhorn family foundation, Sweden.

Abstract

Introduction

The use of oxytocin to augment labor is increasing in many low-resource settings; however, little is known about the effects of such use in contexts where resources for intrapartum monitoring are scarce. In this study, we sought to assess the association between augmentation of labor with oxytocin and delivery outcomes.

Material and methods

We conducted a cohort study in 12 public hospitals in Nepal, including all deliveries with and without augmentation of labor with oxytocin, but excluding elective cesarean sections, women with missing information on augmentation of labor, and women without fetal heart rate on admission. Bivariate and multivariate logistic regression calculating the crude and adjusted risk ratio (aRR) with corresponding 95% CI were performed, comparing (a) intrapartum stillbirth and first-day mortality (primary outcome); and (b) intrapartum monitoring, mode of delivery, postpartum hemorrhage, bag-and-mask ventilation of the newborn, Apgar score, and neonatal death before discharge (secondary outcomes) among women with and without oxytocin-augmented labor.

Results

The total cohort consisted of 78 931 women, of whom 28 915 (37%) had labor augmented with oxytocin and 50 016 (63%) did not have labor augmented with oxytocin. Women with augmentation of labor had no increased risk of intrapartum stillbirth and first-day mortality (aRR 1.24, 95% CI 0.65-2.4), but decreased risks of suboptimal partograph use (aRR 0.71, 95% CI 0.68-0.74), suboptimal fetal heart rate monitoring (aRR 0.50, 95% CI 0.48-0.53), and emergency cesarean section (aRR 0.62, 95% CI 0.59-0.66), and increased risks of bag-and-mask ventilation (aRR 2.1, 95% CI 1.8-2.5), Apgar score <7 at 5 minutes (aRR 1.65, 95% CI 1.49-1.86), and neonatal death (aRR 1.93, 95% CI 1.46-2.56).

Conclusions

Although augmentation of labor with oxytocin might be associated with beneficial effects, such as improved monitoring and a decreased risk of cesarean section, its use may lead to an increased risk of adverse perinatal outcomes. We urge for a cautious use of oxytocin to augment labor in low-resource contexts, and call for evidence-based guidelines on augmentation of labor in low-resource settings.

Abbreviations

  • aRR
  • adjusted risk ratio
  • CI
  • confidence intervals
  • CS
  • cesarean section
  • NePeriQIP
  • Nepal Perinatal Quality Improvement Project
  • WHO
  • World Health Organization
  • Key Message

    In this large-scale cohort study in 12 public hospitals in Nepal, augmentation of labor with oxytocin was associated with improved intrapartum monitoring and decreased risk of emergency cesarean section, but an increased risk of adverse perinatal outcomes.

    1 INTRODUCTION

    Augmentation of labor is the process of stimulating the frequency, duration, and intensity of uterine contractions after the onset of labor either by intravenous oxytocin infusion or artificial rupture of membranes, and it is used to treat prolonged labor and potentially avert cesarean section (CS). The World Health Organization (WHO) recommends that oxytocin to augment labor should only be used when there is a valid indication and potential benefit, when the mother is closely monitored in terms of fetal heart rate and uterine contractions, and the mother is not left unattended.1 However, ambiguous use of oxytocin to augment labor has been reported from both high- and low-resource settings, which has raised concern regarding its potential harmful effects.2-7

    Although augmentation of labor may be effective to shorten the first and second stages of labor,8-10 little to no effect has been demonstrated on CS rates.8-13 During uterine contractions, the maternal spiral arteries are compressed and placental perfusion is strangulated. As oxytocin increases the intensity of uterine contractions and decreases the resting time between contractions,14 it has been suggested that augmentation of labor with oxytocin increases the risk of fetal asphyxia.15 Although previous studies have reported increased risk of uterine hyperstimulation8, 11, 12 and fetal heart rate anomalies8, 12 during augmentation of labor with oxytocin, several large randomized controlled trials in high-resource contexts have failed to establish an association between augmentation of labor with oxytocin and adverse perinatal outcomes.8-12 In spite of reports of an increasing use of oxytocin to augment labor in low-resource contexts,2, 3, 5, 16 no previous studies have, to our knowledge, explored the effects of oxytocin to augment labor in terms of potential benefits and harms in such settings, where intrapartum surveillance is often substandard.17-19 In this large-scale study in 12 public hospitals in Nepal, we sought to assess the association between augmentation of labor with oxytocin and delivery outcomes, including; (a) intrapartum stillbirth and first-day mortality (primary outcome); and (b) intrapartum monitoring, mode of delivery, postpartum hemorrhage, bag-and-mask ventilation of the newborn, Apgar score, and neonatal death before discharge (secondary outcomes).

    2 MATERIAL AND METHODS

    We conducted a cohort study using data collected on all deliveries between 14 April 2017 and 17 October 2018 in 12 public hospitals in Nepal. The study was based on secondary data from a large stepped-wedged randomized controlled trial conducted to evaluate the effect of a quality improvement intervention on perinatal care in Nepal (NePeriQIP), specifically targeting the quality of neonatal resuscitation of non-breathing newborns.20, 21

    2.1 Setting

    Nepal is a low-income country that, despite a number of social challenges and struggles to build up its midwifery workforce,22 has made impressive improvements in maternal and child outcomes during the last decades.23 The under-5 mortality rate has declined from 91/1000 live births in 2001 to 38/1000 live births in 2014,24 and the maternal mortality ratio has declined from 553/100 000 live births in 2000 to 186/100 000 live births in 2017.25 Although only 69% of women undergo the recommended number of four or more antenatal care visits, 84% of women attend antenatal care at least once during their pregnancy and 58% of births are assisted by a skilled birth professional.26 Of the total births in Nepal in 2016, CS accounted for 9%.26

    Hospitals included in the current study had between 1000 and 11 000 deliveries per year, were government-funded referral centers for maternal and newborn care, and had similar resources to provide comprehensive emergency obstetric care. With a total number of deliveries of about 100 000 in the 12 hospitals, the NePeriQIP data collection accounted for almost 29% of the total health facility deliveries in Nepal during the study period. Despite most of them being in the flatlands, the hospitals were different in terms of service coverage and diverse in relation to ethnicity, language, and religion. Mid-Western Regional and Seti Zonal Hospitals were located in the most disadvantaged regions in Nepal in terms of literacy, access to services, and life expectancy. Bheri Zonal Hospital in Nepalgunj had a large number of minority Muslim communities, whereas in Bharatpur, some of the most fringe communities came for maternal and sick newborn services. The hospitals’ labor units were led by skilled birth attendants, fetal heart rate monitoring was performed using Pinard fetoscope or hand-held Doppler, and the guidelines on fetal heart rate monitoring at the hospitals were in adherence with WHO guidelines (ie, to be performed for 60 seconds every 15-30 minutes in the first stage of labor and for 60 seconds every five minutes in the second stage of labor).27 If fetal heart rates are outside the normal range (110-160 beats per minute), the guidelines stipulate monitoring should be extended to cover three consecutive uterine contractions. Oxytocin was readily available for induction or augmentation of labor at all the hospitals, as it is listed as an emergency drug under the Free Essential Drug List in Nepal,28, 29 and was administered either with a motor-driven pump or an infusion drip.30 For augmentation of labor, national clinical guidelines recommend that the initial oxytocin infusion rate is set at 1-4 milliunits/minute (2-8 drops/minute) and that the infusion rate is gradually increased at intervals of at least 20 minutes until a contraction pattern of three to four contractions per 10 minutes is established.30 With either method of infusion, the guidelines stipulate that frequency, strength, and duration of uterine contractions, as well as the fetal heart rate, be carefully monitored. However, the guidelines provide no recommendation to increase fetal heart rate monitoring if the woman has labor augmented with oxytocin.30 No pain relief during labor was offered in any of the hospitals.

    2.2 Participants

    We included all women consenting to participate in the NePeriQIP study20, 21 who were admitted with recorded fetal heart rate and for whom oxytocin to augment labor (yes vs no) was registered in the data collection form. Women with missing data on oxytocin use, women who had absent or no recording of fetal heart rate on admission, and women who underwent elective CS were excluded from the analyses.

    2.3 Data collection and management

    As part of the NePeriQIP data collection, a data surveillance system was established in all the hospitals to collect data on deliveries.20, 21 For obstetric variables, data were extracted from the Maternity Registers and medical records by trained data collectors using a data retrieval form. For sociodemographic variables, data were collected by the same data collectors through semi-structured interviews with mothers before discharge. Completed forms were then assessed by a data coordinator at the hospital for completeness and accuracy before being entered digitally into the data base by the data entry and management team. For entering and cleaning data, we used the Census and Survey Processing System.

    2.4 Variables and outcomes

    The exposure variable was oxytocin to augment labor (yes vs no) as registered in the data collection form. The primary outcome variable was intrapartum-related mortality defined as intrapartum stillbirth (no breathing 10 minutes after delivery) or neonatal death within the first 24 hours of life (yes vs no). Secondary outcomes included suboptimal partograph use (partograph completely or partially filled vs not filled), suboptimal fetal heart rate monitoring (fetal heart rate monitoring performed according to protocol,27 or fetal heart rate monitored more than once during labor vs fetal heart rate monitored only once during labor or not at all), bag-and-mask ventilation of the infant after birth (yes vs no), Apgar score <7 at 5 minutes (yes vs no), emergency CS (yes vs no), instrumental delivery (yes vs no), postpartum hemorrhage (yes vs no), and neonatal death before discharge (yes vs no). We chose to include intrapartum monitoring, that is partograph use and surveillance of fetal heart rate, as outcome variables because we wanted to assess whether women who had labor augmented with oxytocin had been properly monitored given the potential increased risk of adverse obstetric outcomes associated with augmentation of labor.8, 11, 12, 15

    2.5 Statistical analyses

    The cleaned data were exported to Statistical Package for Social Sciences (SPSS) version 25 (IBM Corp.) for analysis. We used descriptive statistics to compare the socio-economic and obstetric characteristics of women with and without augmentation of labor with oxytocin, and applied the Pearson chi-squared test for categorical variables and the independent t test for continuous variables to determine statistically significant differences between the groups. We considered a P-value <.05 as statistically significant. We further compared primary and secondary outcomes between exposed and unexposed using the Pearson chi-squared test and bivariate logistic regression, and used multivariate logistic regression to adjust the estimate for variables previously reported to be confounders for the outcome under investigation, for example maternal age, ethnicity, educational level, gestational age, complications during pregnancy, mode of delivery, birthweight etc.31, 32 We presented the results as crude and adjusted risk ratios (aRR) including 95% CI. Results on outcome measures were also stratified as per hospital level (high-, medium-, and low-volume) as defined by the NePeriQIP randomization process,21 in order to explore the generalizability of findings across hospital settings. Missing data were excluded from the analyses.

    2.6 Ethical approval

    Written informed consent was obtained from the mothers before inclusion in the NePeriQIP study and confidentiality was guaranteed. The study was approved by the Ethical Review Board of Nepal Health Research Council (reference number 26-2017) on 16 March 2017. For the current study, the researchers worked with a database coded as per the patient's identity.

    3 RESULTS

    There were 104 223 women admitted to the hospitals included in the NePeriQIP study during the study period. After exclusion of women with elective CS (5650, 5.4%), missing data on oxytocin to augment labor (15 814, 15%), and absent or no recording of fetal heart rate on arrival (3828, 3.7%), the total cohort consisted of 78 931 women of which 28 915 (37%) had labor augmented with oxytocin and 50 016 (63%) did not have labor augmented with oxytocin (Figure 1). Women who had labor augmented with oxytocin were more often younger, more often from a disadvantaged ethnic group, had a higher level of education, were more often giving birth to their first child, had higher gestational age, were more often admitted to the high-volume hospitals, were more often induced with prostaglandins or oxytocin, were less likely to have experienced complications during pregnancy or delivery, and more often gave birth to a normal-weight infant (Table 1). The overall proportion of women who had labor augmented with oxytocin was higher in high-volume than in medium- and low-volume hospitals (Table 2).

    Details are in the caption following the image
    Flow chart of inclusion of participants. NePeriQIP, Nepal Perinatal Quality Improvement Project; CS, cesarean section
    TABLE 1. Characteristics of women with and without augmentation of labor with oxytocin in 12 public hospitals in Nepal. Includes only women admitted with fetal heart rate
    Characteristic Oxytocin to augment labor P valuea

    Yes

    N = 28 915

    No

    N = 50 016

    N % N %
    Maternal age (SD) 23.6 (4.1) 24.0 (4.3) <.001
    Ethnicity
    Advantaged group 14 410 50% 25 540 51% .001
    Disadvantaged group 14 505 50% 24 476 49%
    Missing 0 0% 0 0%
    Education level
    Illiterate 944 3.2% 1888 3.8% <.001
    Literate 2076 7.1% 3855 7.7%
    Basic education 3650 13% 6526 13%
    Secondary 14 896 51% 22 656 44%
    Higher 1637 5.6% 3281 6.2%
    Missing 5711 20% 11 810 24%
    Parity
    Para 0 15 573 54% 21 903 44% <.001
    Para 1 8859 31% 17 912 36%
    Para ≥2 4480 16% 10 195 20%
    Missing 3 0% 6 0%
    Gestational age
    22-27 weeks 268 0.9% 1831 3.7% <.001
    28-32 weeks 397 1.4% 1619 3.3%
    33-36 weeks 3317 11% 6484 13%
    37-41 weeks 21 909 76% 35 342 71%
    42-45 weeks 947 3.3% 1415 2.8%
    Missing 2133 7.3% 3325 6.6%
    Inclusion of study participants across hospitals
    Western Regional Hospital 3180 11% 5831 12% <.001
    Midwestern Regional Hospital 1537 5.3% 3791 7.6%
    Bardiya District Hospital 36 0.1% 1063 2.1%
    Bharatpur Zonal Hospital 6856 24% 9676 19%
    Seti Zonal Hospital 3348 12% 6983 14%
    Nuwakot District Hospital 145 0.5% 1156 2.3%
    Koshi Zonal Hospital 3214 11% 5376 11%
    Rapti Subregional Hospital 1651 5.7% 3743 7.5%
    Prithiva Chandra Hospital 118 0.4% 651 1.3%
    Lumbini Zonal Hospital 4912 17% 5781 12%
    Bheri Zonal Hospital 3309 11% 4181 8.4%
    Pyuthan District Hospital 609 2.1% 1784 3.6%
    Fetal heart rate at admission
    Normal 28 863 99.8% 49 877 99.7% .007
    Abnormal 52 0.2% 139 0.3%
    Stage of labor at admission
    Not in labor 4433 15% 9528 19% <.001
    Latent phase of labor 16 055 56% 23 020 46%
    First stage of labor 7659 27% 14 593 29%
    Second stage of labor 766 2.6% 2755 5.5%
    Missing 2 0% 120 0.20%
    Induction of labor
    No 16 044 56% 31 758 64% <.001
    With prostaglandins 6146 21% 7604 15%
    With amniotomy 662 2.3% 1204 2.4%
    With oxytocin 3940 14% 3175 6.3%
    Missing 2123 7.3% 6275 13%
    Complications during pregnancy or labor
    Multiple pregnancy 256 0.9% 537 1.1% .011
    Premature rupture of membranes 127 0.3% 84 0.3% <.001
    Severe anemia 33 0.1% 123 0.2% <.001
    Chorioamnionitis 5 0% 5 0% <.001
    Hypertensive disorders 71 0.2% 148 0.3% <.001
    Antepartum/intrapartum hemorrhage 6 0% 56 0.1% <.001
    Birthweight
    <2500 g 3743 13% 9491 19% <.001
    >2500 g 23 095 80% 37 200 74%
    Missing 2076 7.2% 3325 6.6%
    Sex of the baby
    Male 15 607 54% 26 969 54% .88
    Female 13 308 46% 23 047 46%
    Missing 0 0% 0 0%
    • a Pearson chi-squared test for categorical variables and the independent t test for continuous variables.
    TABLE 2. Proportion of women receiving oxytocin to augment labor in high-, medium-, and low-volume hospitals
    Hospital

    Number and percentage of deliveries with labor augmented with oxytocin

    N = 28 915

    N %
    High-volume hospitals 18 162 41%
    Bharatpur Zonal Hospital 6856 42%
    Lumbini Zonal Hospital 4912 46%
    Koshi Zonal Hospital 3214 37%
    Western Regional Hospital 3180 35%
    Medium-volume hospitals 9845 35%
    Seti Zonal Hospital 3348 32%
    Bheri Zonal Hospital 3309 44%
    Rapti Sub-regional Hospital 1651 31%
    Mid-Western Regional Hospital 1537 29%
    Low-volume hospitals 908 16%
    Pyuthan District Hospital 609 25%
    Prithivi Chandra Hospital 118 15%
    Nuwakot District Hospital 145 11%
    Bardiya District Hospital 36 3.3%

    Details of delivery outcomes and intrapartum monitoring among women with and without augmentation of labor with oxytocin are presented in Table 3. After multivariate regression analyses accounting for socio-economic and obstetric characteristics, including events during pregnancy and delivery (Table 4), we found that women with augmentation of labor had no increased risk of intrapartum stillbirth and first-day mortality (aRR 1.24, 95% CI 0.65-2.4), but a lower risk of suboptimal partograph use (aRR 0.71, 95% CI 0.68-0.74), a lower risk of suboptimal fetal heart rate monitoring (aRR 0.50, 95% CI 0.48-0.53), and a lower risk of CS (aRR 0.62, 95% CI 0.59-0.66). On the other hand, they had a higher risk of instrumental delivery (aRR 2.60, 95% CI 2.40-2.80), a higher risk of bag-and-mask ventilation of the newborn (aRR 2.10, 95% CI 1.80-2.50), a higher risk of Apgar score <7 at 5 minutes (aRR 1.65, 95% CI 1.49-1.86), and a higher risk of neonatal death before discharge (aRR 1.93, 95% CI 1.46-2.56). Overall, results were similar regardless of hospital level (high-, medium-, and low-volume), except for results in low-volume hospitals related to suboptimal partograph use, emergency CS, postpartum hemorrhage, bag-and-mask ventilation, Apgar score <7 at 5 minutes, and neonatal death before discharge, of which some failed to reach statistical significance in the multivariate analyses (see Supplementary material, Tables S1-S3).

    TABLE 3. Delivery outcomes and intrapartum monitoring among women with and without augmentation of labor with oxytocin in 12 public hospitals in Nepal. Includes only women admitted with fetal heart rate
    Delivery outcomes and intrapartum monitoring Oxytocin to augment labor P-valuea

    Yes

    N = 28 915

    No

    N = 50 016

    N % N %
    Primary outcome
    Intrapartum stillbirth and first day mortality among newborns
    Yes 64 0.2% 130 0.3% .28
    No 28 851 99.8% 49 886 99.7%
    Secondary outcomes
    Progress of labor recorded in partograph
    Yes, completely filled 10 092 35% 12 968 26% <.001
    Yes, partially filled 8423 29% 14 817 30%
    No 10 400 36% 22 231 44%
    Fetal heart rate monitoring performed during labor
    Yes, as per protocol 15 069 52% 20 881 42% <.001
    Yes, sporadically ( >once) 8739 30% 12 798 26%
    Yes, only once 4350 15% 11 868 24%
    No 757 2.6% 4469 8.9%
    Mode of delivery
    Spontaneous vaginal 23 338 81% 38 708 77% <.001
    Instrumental vaginal 1891 6.5% 1049 2.1%
    Emergency cesarean section 3561 12% 9688 19%
    Assisted breech delivery 111 0.4% 297 0.6%
    Postpartum hemorrhage
    Yes 67 0.2% 155 0.3% .046
    No 28 848 100% 49 861 100%
    Bag-and-mask ventilation
    Yes 439 1.6% 346 0.7% <.001
    No 27 637 96% 48 261 99%
    Apgar score
    <7 at 5 minutes 1136 3.9% 1553 3.1% <.001
    >7 at 5 minutes 27 773 96% 48 447 97%
    Neonatal death
    Dead at discharge 234 0.8% 422 0.8% .60
    Alive at discharge 28 590 99% 49 419 99%
    • a Pearson chi-squared test.
    TABLE 4. Bivariate and multivariate logistic regression analyses of delivery outcomes and intrapartum monitoring among women with and without augmentation of labor with oxytocin in 12 public hospitals in Nepal. Includes only women admitted with fetal heart rate
    Delivery outcomes and intrapartum monitoring Oxytocin to augment labor
    No Yes Yes
    Crude RR 95% CI Adjusted RR 95% CI
    Primary outcome
    Intrapartum stillbirth and first day mortality 1.0 0.85 0.63-1.15 1.24a 0.65-2.40
    Secondary outcomes
    Suboptimal partograph use 1.0 0.70 0.69-0.72 0.71b 0.68-0.74
    Suboptimal fetal heart rate monitoring 1.0 0.44 0.43-0.46 0.50c 0.48-0.53
    Emergency cesarean section 1.0 0.57 0.55-0.60 0.62d 0.59-0.66
    Instrumental delivery 1.0 2.70 2.50-2.90 2.60d 2.40-2.80
    Postpartum hemorrhage 1.0 0.75 0.56-0.99 0.80e 0.55-1.20
    Bag-and-mask ventilation 1.0 2.20 1.90-2.60 2.10a 1.80-2.50
    Apgar score < 7 at 5 minutes 1.0 1.30 1.20-1.40 1.65a 1.49-1.86
    Neonatal death before discharge 1.0 0.96 0.82-1.10 1.93a 1.46-2.56
    • a Adjusted for maternal age, ethnicity, educational level, parity, gestational age, fetal heart rate at admission (normal vs absent), induction of labor, complications during pregnancy or labor, birthweight, suboptimal partograph use, suboptimal fetal heart rate monitoring, and mode of delivery.
    • b Adjusted for maternal age, ethnicity, educational level, parity, gestational age, fetal heart rate at admission (normal vs absent), induction of labor, complications during pregnancy or labor, birthweight, and suboptimal fetal heart rate monitoring.
    • c Adjusted for maternal age, ethnicity, educational level, parity, gestational age, fetal heart rate at admission (normal vs absent), induction of labor, complications during pregnancy or labor, birthweight, and suboptimal partograph use.
    • d Adjusted for maternal age, ethnicity, educational level, parity, gestational age, fetal heart rate at admission (normal vs absent), induction of labor, complications during pregnancy or labor, birthweight, suboptimal partograph use, and suboptimal fetal heart rate monitoring.
    • e Adjusted for maternal age, ethnicity, educational level, parity, gestational age, fetal heart rate at admission (normal vs absent), induction of labor, complications during pregnancy or labor, birthweight, and mode of delivery.

    4 DISCUSSION

    In this large-scale cohort study exploring the association between oxytocin to augment labor and delivery outcomes in 12 public hospitals in Nepal, we found no difference in intrapartum stillbirth and first-day mortality between women with and without augmentation of labor with oxytocin. On the other hand, after multivariate analyses accounting for socio-economic and obstetric characteristics and events, we found that women who had labor augmented with oxytocin had a lower risk of suboptimal intrapartum monitoring and CS, but a higher risk of instrumental delivery, bag-and-mask ventilation, Apgar score <7 at 5 minutes, and neonatal death before discharge.

    The main strength of our study was the large sample size collected in 12 public hospitals distributed over the country and in total accounting for almost a third of Nepal's facility births during the study period, which make our results generalizable and applicable to many other health-care settings in low-income countries. The data collection teams were organized to secure data quality data, and repeated validity checks were undertaken. Nevertheless, missing data were a problem in some variables. For example, more women than reported most likely suffered from obstetric complications such as hypertensive disorders. It is, however, difficult to speculate on whether either of the groups under comparison had a larger underreporting of complications than the other, and how this might have affected the results. The current study was based on secondary data from the NePeriQIP study,20 hence, there is a risk that the study was under-powered to detect an association between augmentation of labor and our primary outcome. The results on secondary outcomes, such as intrapartum monitoring, mode of delivery, and adverse perinatal outcomes do, however, add valuable evidence to the current literature, hence the study still serves its purpose. Although all hospitals were referral centers with similar resources to provide emergency obstetric care, the use of oxytocin to augment labor differed across hospital levels, with higher use of augmentation of labor in high-volume hospitals. To account for this, we stratified results by hospital level, which produced similar results as for the total cohort. Lastly, another limitation of our study was that it provided no information on the clinical events that proceeded the decision to augment labor with oxytocin, whether oxytocin was used according to guidelines,30 whether there was a problem with uterine hyperstimulation, or what dosage of oxytocin was used; nor did we have data on important obstetric complications such as anal sphincter injury33, 34 and uterine rupture,35 which might have been more common in the group of women who had labor augmented with oxytocin.

    In line with previous publications on the effect of augmentation of labor with oxytocin, which have reported small or statistically non-significant associations between augmentation of labor and adverse perinatal outcomes,8-12 our large-scale study found no increased risk of intrapartum stillbirth and first-day mortality. On the other hand, we found an increased risk of neonatal resuscitation, low Apgar score, and neonatal death before discharge among women who had labor augmented with oxytocin, which suggests that oxytocin to augment labor might have harmful effects when used in low-resource contexts. Although our conflicting results regarding perinatal outcomes might be due to a type II bias, as described above, Bakker et al, who analyzed low pH in the umbilical cord artery at birth among women who had labor augmented with oxytocin, suggest the association between augmentation of labor and fetal acidosis is most likely the result of increased uterine activity and not of the oxytocin itself.35 This is also in line with a UK case-control study, which found that oxytocin did not increase base-line fetal heart rate or short-term variability when given according to a routine regimen.14 As our study provides no information on whether indications of augmentation of labor with oxytocin were valid, to what extent clinical guidelines were adhered to, and if study participants were subjected to uterine hyperstimulation, we recommend that future studies from low-resource settings should focus on health-care professionals’ monitoring of uterine contractions during augmentation of labor with oxytocin, the role of hyperstimulation in relation to adverse perinatal outcomes, and evaluation of clinical guidelines for labor augmentation. Such evidence will be imperative to ensure a cautious use of oxytocin to augment labor.3, 5, 36 The lack of statistically significant associations between augmentation of labor and adverse perinatal outcomes found in low-volume hospitals was most likely due to a smaller sample size in these hospitals.

    As emphasized by the WHO,1 augmentation of labor should only be performed if the mother is properly monitored. A positive effect of augmentation of labor with oxytocin in our setting was that it seemed to have triggered an increased commitment among health-care professionals, resulting in better partograph use and fetal heart rate monitoring than among those women who did not have labor augmented with oxytocin. The quality of intrapartum monitoring was, however, far from meeting the WHO standards, as reported from other low-resource settings,17-19 and the majority of women only had fetal heart rate monitoring performed sporadically or once. In contrast to previous studies that have reported a limited effect of augmentation of labor with oxytocin on CS rates,8-13 women in our study who received augmentation of labor with oxytocin had a significantly lower risk of CS than women who received no augmentation of labor, even after adjusting for socio-economic and obstetric characteristics and events. These results are in line with an intervention study in Zanzibar, where the CS rate increased as the total oxytocin use to argument labor dropped from 22% to 12%, although the authors do not interpret the oxytocin drop to have had a casual effect on the CS rate.3 Given the global concern of CS overuse,31, 37-39 our results highlight that augmentation of labor, as part of active management of labor, has a potential to avert unnecessary operations.

    5 CONCLUSION

    Although augmentation of labor with oxytocin in low-resource settings might be associated with beneficial effects such as improved partograph use, improved fetal heart rate monitoring, and a lower risk of CS, its use may also lead to an increased risk of adverse perinatal outcomes such as bag-and-mask ventilation, low Apgar score, and neonatal death. We urge for a cautious use of oxytocin to augment labor in low-resource contexts, and call for future studies to evaluate and suggest guidelines for augmentation of labor suitable for low-resource settings, to ensure optimal outcomes for mothers and newborns.

    ACKNOWLEDGMENTS

    We would like to thank Rejina Gurung, Omkar Basnet, and the data collection team for their thorough work in managing the data collection for the current study.

      CONFLICT OF INTEREST

      None.