Impact of labor analgesia on mode of delivery and neonatal outcomes in Japan: A retrospective cohort study

Labor analgesia (LA) is associated with the potential hazard of high-risk delivery, such as cesarean section (CS) and instrumental vaginal delivery (IVD), and adverse neonatal outcomes such as neonatal asphyxia and respiratory distress. The objective was to examine the impact of LA on mode of delivery and neonatal outcomes and to counsel pregnant women about a potentially higher risk and allow them to decide LA or non-LA. This retrospective cohort study containing 5,184 pregnant women analyzed the association between LA and both mode of delivery and neonatal outcomes. LA increased the risk of IVD (Adjusted Odds Ratio [AOR] 3.25, 95% confidence interval [95%CI] 2.51–4.20) but decreased that of CS (AOR 0.52, 95%CI 0.44–0.60). Two factors (advanced maternal age [AOR 1.70, 95%CI 1.33–2.17] and primipara [AOR 4.72, 95%CI 3.30–6.75]) increased the risk of IVD. We should carefully consider the indication of LA for cases with these two factors since LA can increase the risk of IVD and adverse neonatal outcomes.


Unfunded studies
Enter: The author(s) received no specific funding for this work. Since this study was an observational retrospective study, written informed consent was waived and we proceeded using an opt-out approach. If the data are held or will be held in a public repository, include URLs, accession numbers or DOIs. If this information will only be available after acceptance, indicate this by ticking the box below. For example: All XXX files are available from the XXX database (accession number(s) XXX, XXX.).

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
• If the data are all contained within the manuscript and/or Supporting Information files, enter the following: All relevant data are within the manuscript and its Supporting Information files.
• If neither of these applies but you are able to provide details of access elsewhere, with or without limitations, please do so. For example: Data cannot be shared publicly because of [XXX]. Data are available from the XXX Institutional Data Access / Ethics Committee (contact via XXX) for researchers who meet the criteria for access to confidential data.
The data underlying the results presented in the study are available from (include the name of the third party Luke's International University for researchers who meet the criteria for access to confidential data. The address for contact office of the Research Support Unit of St. Luke's International University is as follows e-mail: kenkyukikaku@luke.ac.jp and contact information or URL Childbirth, as one of the most important life events for most women, requires them to go 27 through a series of burdens during labor and delivery. The magnitude of labor pain is often 28 compared with a finger amputation [1]. Although the pain itself is not life-threatening for 29 pregnant women and infants, it can transiently impair the cognitive function of pregnant women, 30 especially memory function, during early puerperium [2]. Furthermore, the pain is related to 31 the risk of postpartum depression, and the degree of pain is also associated with the severity of 32 depression [3]. Researchers reported that analgesia can relieve these symptoms [4]. Currently, 33 labor analgesia (LA) with various methods is commonly used in modern obstetric practice and 34 includes spinal anesthesia, epidural analgesia, and combined spinal-epidural analgesia [5]. 35 However, the side effects, especially those related to the mode of delivery and neonatal 36 outcomes, are of great concern. Some studies reported that LA is associated with the potential 37 hazards of cesarean section (CS), instrumental vaginal delivery (IVD), and adverse neonatal 38 where the average number of deliveries is 1,400 per annum. We selected samples of pregnant 65 women based on the inclusion and exclusion criteria shown in Fig 1. We included all pregnant 66 women who were admitted for delivery in the study periods. The exclusion criteria were as 67 follows Stillbirths and miscarriages were excluded. Since a mode of delivery was defined as 68 the primary outcome, cases with elective CS were excluded. For the same reason, cases with 69 multiple pregnancy and malpresentation that would have resulted in elective CS were also 70 excluded. Cases of delivery at a gestational age of less than 36 weeks were excluded because 71 these could affect a mode of delivery and neonatal outcomes. Finally, cases with elective LA, 72 cases in which LA was unavailable for LA, and cases with the lack of data due to outborn were 73 excluded. All of the cases with elective LA had postoperative congenital heart disease, and the 74 attending physician decided to conduct labor analgesia without the patient's wishes. The one 75 patient with thrombocytopenia associated with Sjögren's syndrome who could not receive LA 76 was excluded. 77

Data collection
79 All data were obtained by chart review from the electronic medical records. Maternal data 80 included age at delivery, which was dichotomized as advanced maternal age, defined as giving 81 birth at age 35 or older [19], or not, height, pre-delivery body weight, pre-delivery body mass 82 index, parity, duration of the second stage of labor, and whether or not LA was applied. Neonatal 83 data included gestational age, gender, Apgar score (1 minute, 5 minute), admission to the 84 neonatal intensive care unit (NICU), neonatal death, which is defined as death within 28 days 85 of birth, and birth weight. The estimated fetal weight was replaced with the actual birth weight 86 because the estimated fetal weight is not always accurate due to inter-examiner reliability. 87 Pregnant women who received LA for vaginal delivery received either epidural analgesia, 88 combined spinal-epidural analgesia, or dural puncture epidural. 89

90
The primary outcome of our study was the mode of delivery, and the secondary outcomes were 91 neonatal admission to the NICU and Apgar score. The indications for admission to the NICU 92 were gestational age less than 36 weeks, birth weight less than 2,300 g, the occurrence of a 93 respiratory disorder requiring supplemental oxygen or more, and when deemed necessary by 94

96
When the association between mode of delivery and application of LA was examined, we 97 performed the analysis using two different outcomes related to the mode of delivery: 98 spontaneous vaginal delivery (SVD) and ECS ("SVD versus ECS") and SVD and IVD ("SVD 99 versus IVD") with the reasoning that the mode of delivery is often automatically determined 100 by the degree of the fetal head station. When immediate delivery is considered, ECS was 101 selected in cases of high fetal head station, while IVD was selected in cases of low fetal head 102 station. The high fetal head station means at +2 or more. The final decision is made by the 103 attending obstetrician. Multivariate logistic regression analysis was performed to calculate the 104 AOR and 95%CI, to examine the impact of LA on the mode of delivery and neonatal outcomes. 105 In addition, we focused only on pregnant women who received LA and attempted to identify 106 the characteristics of patients with factors that make LA risky. This study was approved by the 107 institutional review board of our institute (Approval code: 21-R113) and was conducted 108 according to the principles of the Declaration of Helsinki. Since this study was an observational 109 retrospective study, the requirement for written informed consent was waived, which allowed 110 us to proceed with our study in an opt-out manner. All data were analyzed using SPSS Statistics 111

114
The total number of pregnant women who were admitted for delivery was 6,404. After 115 considering inclusion and exclusion criteria, 5,184 women were finally included in the study 116 population (Fig 1). The mean maternal age at delivery was 34.6 years with a standard deviation 117 (SD) of 4.45 (Table 1). 118 The proportion of primipara was higher in the LA group (73%) than in the non-LA 127 group (56%, p < 0.001) ( Table 1). Primiparous women were more likely to receive LA (57%) 128 than multiparous women (39%). 129 To examine the association between the mode of delivery and the application or not 130 of LA, a univariate analysis was performed for IVD and ECS against SVD (Table 2). Compared 131 with SVD, both IVD and ECS were associated with a higher proportion of admission to the 132 NICU (IVD: 8.0%, p = 0.008 ECS: 12.8%, p < 0.001) and lower Apgar score (IVD 1 min/5 133 min: mean7.63/8.85, SD 1.09/0.603, p < 0.001/< 0.001 ECS 1 min/5 min: mean 7.56/8.82, SD 134 1.26/0.781, p < 0.001/< 0.001). The rate of LA was significantly higher in IVD (80%, p < 0.001) 135 than SVD (50%) and conversely lower in ECS (40%, p < 0.001). 136 Primary outcome: Impact of labor analgesia on mode of delivery 144 First, we extracted the risks that may affect the mode of delivery and then examined the impact 145 of LA on the mode of delivery. We hypothesized that the risks were LA, advanced maternal age, 146 lower maternal height, larger estimated fetal weight, and primipara. The primary outcome for 147 mode of delivery is shown in Table 3. We analyzed two groups: one compared SVD and ECS, 148 excluding IVD ("SVD versus ECS"), and the other compared SVD and IVD, excluding ECS 149   In "SVD versus ECS," those of advanced maternal age and primipara exhibited a 157 higher AOR (AOR 2.39, 6.00 95%CI 2.05-2.78, 4.95-7.28). Lower maternal height and larger 158 estimated fetal weight increased the risks of ECS, while the use of LA decreased the risk of 159 ECS (AOR 0.52, 95%CI 0.44-0.60). 160 Similar to the "SVD versus ECS" results, advanced maternal age and primipara also 161 exhibited a higher AOR (AOR 1.66, 5.13 95%CI 1.33-2.05, 3.77-6.99) in the "SVD versus The secondary outcomes, including neonatal outcomes and Apgar score, are shown in Table 3. 168 LA lowered the Apgar score at 1 minute (AOR 2.06, 95%CI 1.50-2.83), but it did not influence 169 the Apgar score at 5 minutes or the NICU admission rate. Being a primipara was the strongest 170 factor that could cause all adverse neonatal outcomes after adjusting for covariates including 171 LA prescription. 172 When the population was restricted to only those who received LA, we found that 173 pregnant women with the following two risk factors demonstrated a higher risk of IVD: 174 advanced maternal age (AOR 1.70, 95%CI 1.33-2.17) and primipara (AOR 4.72, 95%CI 3.30-175 6.75) ( Table 3). Neither lower maternal height nor larger estimated fetal weight were 176 significantly associated with an increased risk of IVD . 177

178
This study included more than 5,000 pregnant women, approximately half of whom were of 179 advanced maternal age, and who were admitted to our hospital for delivery from 2017 to 2021. 180 In a multivariate analysis, LA increased the risk of IVD but decreased that of ECS. Moreover, 181 advanced maternal age and primipara increased the risks of both ECS and IVD. Among those 182 who received LA, two factors of advanced maternal age and primipara increased the risk of 183 IVD; the AOR of primipara was highest at 4,72 (95%CI 3.30-6.75) and that of advanced 184 maternal age was 1.70 (95%CI 1.33-2.17). 185 To the best of our knowledge, our study is one of few reports from Japan that presents 186 an association between LA and mode of delivery as well as neonatal outcomes. The novelty is 187 that we sought to identify the characteristics of pregnant women who may need to be cautious 188 in selecting LA. Our study is valuable in that we analyzed primarily Asian patients, who are 189 recognized to be at an increased risk of CS [6]; we also included a large number of patients of 190 advanced maternal age. 191 As our results showed that LA decreased the risk of CS, many recent reports showed 192 that LA is not associated with increased risk of CS or that LA even reduces CS rates [13,15]. 193 Despite a previous study published several decades ago, which concluded that epidural 194 analgesia was associated with a significantly increased risk of CS among primipara [6], a recent 195 Cochrane Review in 2018 that included 40 trials and more than 11,000 pregnant women found 196 that epidural analgesia did not increase the risk of CS [14]. The reason that LA is not associated 197 with and does not increase the ECS rate may be because factors associated with more painful 198 labor are themselves associated with an increased risk of CS (e.g., fetal malrotation, fetal-pelvic 199 disproportion, dysfunctional labor) [20]. Thus, we consider that pain relief might help to avoid 200 ECS. However, no studies compared the degree of pain with the CS rate, and it is unclear why 201 LA lowers the CS rate. Further investigation is needed to determine the pathophysiological 202

reasons. 203
Our results showed that LA also increases the risk of IVD. While some studies 204 likely to undergo IVD is due to stalled labor, followed by risk to fetal well-being [8]. 210 We found that the use of LA is associated with adverse neonatal outcomes. The effects 211 of LA on neonatal outcomes are controversial and are the same as those on mode of delivery. 212 Although a meta-analysis reported in the Cochrane Review in 2018 denied an association 213 between epidural analgesia and adverse neonatal outcomes, such as admission rate to the NICU 214 and an Apgar score of less than 7 at 5 minutes [14], many other studies confirmed the 215 association [7,11,12]. In general, Apgar scores are lower in newborns delivered by CS [22, 216 23], but in our study, the Apgar score was lower in the LA group even though the CS rate 217 decreased, which suggests that LA itself must lower the Apgar score. Kumar et al. reported that 218 the Apgar score was significantly lower with use of epidural analgesia because of respiratory 219 distress, and they considered that a large volume of distribution at a steady state in neonates 220 resulted in a much longer terminal elimination half-life of fentanyl over several hours [11]. 221 Most newborns with neonatal asphyxia are able to establish fetal-to-neonatal transition after 222 appropriate resuscitation immediately after birth. This means that even if the Apgar score at 1 223 minute is low, the Apgar score at 5 minutes will be higher after appropriate resuscitation. This 224 may be the reason why the results of our study did not reveal statistically significant differences 225 when the Apgar score at 5 minutes was less than 7. Researchers suggested that LA may elicit 226 more severe adverse effects in a newborn in the immediate postnatal period than what the Apgar 227 score at 1 minute purportedly represents. Neonatologists are expected to exhibit resuscitation 228 skills for cases of neonatal asphyxia, especially in cases of LA. To promote resuscitation skills, 229 we should provide a training and educational system to maintain higher skills and knowledge. 230 Even if neonatal resuscitation is learned through simulation, this skill is difficult to maintain 231 without practical experience, and thus, aggressive resuscitation practices are important. The 232 centralization of facilities where LA is available might ensure safe deliveries. We should 233 carefully consider LA, especially in facilities without personnel, such as pediatricians, 234 neonatologists, and midwives, who attend deliveries. 235 According to our results, LA may reduce the risk of ECS but increase the risk of IVD 236 and lower the Apgar score at 1 minute. In particular, the increased risk of IVD was significantly 237 higher when the pregnant women who were primiparous and of advanced maternal age were 238 administered LA. While the decrease in ECS is favorable in terms of surgical invasiveness and 239 the safety of the next delivery, the increase in IVD is a serious issue. IVD is associated with 240 many potential risks to both the mother (e.g., perineal and vaginal tears) and the newborn (e.g., 241 brachial plexus injury with shoulder dystocia, subgaleal and intracranial hemorrhage) [21]. 242 Prior et al. reported that the Apgar score was lower in infants delivered by IVD than in those 243 delivered by CS [22]. Based on our results, we suggest that pregnant women at an increased 244 risk of IVD, such as those of advanced maternal age (i.e., 35 years of age or older), lower 245 maternal height, larger estimated fetal weight, and primipara, should carefully consider LA. 246 Hence, when obstetricians and/or anesthesiologists confirm a pregnant woman's requests to 247 receive LA, explaining not only the advantages but also the disadvantages and risks and 248 obtaining appropriate informed consent is important. Moreover, pregnant women with these 249 risk factors may be guided to receive alternative treatments, such as acupuncture, hypnotism, The results of our study are based on a limited population, and as mentioned above, 256 we should be cautious about perceptions of the indications for and the risks of LA. Determining 257 the indications for LA on a case-by-case basis and recognizing that LA is not beneficial for 258 every pregnant woman is necessary because the population of pregnant women and facilities 259 vary depending on the country, region, and other circumstances. The environment surrounding 260 childbirth is changing; for example, advanced maternal age is increasing in all developed 261 countries due to women's career advancement and lifestyle changes, and thus, we need to 262 consistently update our research findings on the impacts of LA. In addition, the medical 263 principle behind lower Apgar scores is not completely understood, and since the topic of LA 264 remains controversial, further research is needed. 265 Our study also exhibits some limitations. The impact of the method of LA on the 266 mode of delivery was not considered. In Tables 1 and 2, although statistically significant 267 differences were observed in terms of maternal age, height, pre-delivery body weight, pre-268 delivery body mass index, gestational age, sex of newborn, and birth weight, these differences 269 are not noteworthy in clinical practice because these factors were continuous variables and the 270 number of cases that met the inclusion criteria was high. Similarly, in the multivariate analysis, 271 we should focus on the interpretation of the comparison of the AOR of continuous variables, 272 such as maternal height and estimated birth weight, and that of categorical variables, such as 273 LA, advanced maternal age, and primipara. In addition, establishing the cut-off for maternal 274 height and estimated birth weight is very difficult, and thus, the results of our study are unlikely 275 to affect actual practice without specific values. 276 In conclusion, although LA is an important means to relieve pain during labor and 277 delivery, it can increase the risk of IVD and an Apgar score of less than 7 at 1 minute. Therefore, 278 we should carefully consider the indication of LA in cases of advanced maternal age and 279 primipara. Furthermore, it is desirable to maintain and improve resuscitation skills through 280 training and certification systems and to intensify facilities where LA is available.  However, the side effects, especially those related to the mode of delivery and neonatal 36 outcomes, are of great concern. Some studies reported that LA is associated with the potential 37 . Therefore, it is urgent that data from Japan be 53 analyzed due to the high proportion of advanced maternal age. 54 We believe that this study, which includes an Asian population with a high rate of 55 advanced maternal age, can contribute to the still controversial knowledge of the risks 56 associated with LA. In our study, we used a retrospective chart review and multivariate analysis 57 to examine the impact of labor analgesia on the mode of delivery and neonatal outcomes at our 58 hospital and aimed to counsel pregnant women about a potentially higher risk and allow them 59 61 Study population 62 We conducted a retrospective cohort study of all pregnant women who were admitted for 63 delivery from April 2017 to August 2021 at St Luke's International Hospital, Tokyo, Japan, 64 where the average number of deliveries is 1,400 per annum. We selected samples of pregnant 65 women based on the inclusion and exclusion criteria shown in Fig 1. We included all pregnant 66 women who were admitted for delivery in the study periods. The exclusion criteria were as 67 follows Stillbirths and miscarriages were excluded. Since a mode of delivery was defined as 68 the primary outcome, cases with elective CS were excluded. For the same reason, cases with 69 multiple pregnancy and malpresentation that would have resulted in elective CS were also 70 excluded. Cases of delivery at a gestational age of less than 36 weeks were excluded because 71 these could affect a mode of delivery and neonatal outcomes. Finally, cases with elective LA, 72 cases in which LA was unavailable for LA, and cases with the lack of data due to outborn were 73 excluded. All of the cases with elective LA had postoperative congenital heart disease, and the 74 attending physician decided to conduct labor analgesia without the patient's wishes. The one 75 patient with thrombocytopenia associated with Sjögren's syndrome who could not receive LA 76 was excluded. 77

Data collection
79 All data were obtained by chart review from the electronic medical records. Maternal data 80 included age at delivery, which was dichotomized as advanced maternal age, defined as giving 81 birth at age 35 or older [19], or not, height, pre-delivery body weight, pre-delivery body mass 82 index, parity, duration of the second stage of labor, and whether or not LA was applied. Neonatal 83 data included gestational age, gender, Apgar score (1 minute, 5 minute), admission to the 84 neonatal intensive care unit (NICU), neonatal death, which is defined as death within 28 days 85 of birth, and birth weight. The estimated fetal weight was replaced with the actual birth weight 86 because the estimated fetal weight is not always accurate due to inter-examiner reliability. 87 Pregnant women who received LA for vaginal delivery received either epidural analgesia, 88 combined spinal-epidural analgesia, or dural puncture epidural. 89

Outcomes and definitions 90
The primary outcome of our study was the mode of delivery, and the secondary outcomes were 91 neonatal admission to the NICU and Apgar score. The indications for admission to the NICU 92 were gestational age less than 36 weeks, birth weight less than 2,300 g, the occurrence of a 93 respiratory disorder requiring supplemental oxygen or more, and when deemed necessary by 94

Statistical analysis 96
When the association between mode of delivery and application of LA was examined, we 97 performed the analysis using two different outcomes related to the mode of delivery: 98 spontaneous vaginal delivery (SVD) and ECS ("SVD versus ECS") and SVD and IVD ("SVD 99 versus IVD") with the reasoning that the mode of delivery is often automatically determined 100 by the degree of the fetal head station. When immediate delivery is considered, ECS was 101 selected in cases of high fetal head station, while IVD was selected in cases of low fetal head 102 station. The high fetal head station means at +2 or more. The final decision is made by the 103 attending obstetrician. Multivariate logistic regression analysis was performed to calculate the 104 AOR and 95%CI, to examine the impact of LA on the mode of delivery and neonatal outcomes. 105 In addition, we focused only on pregnant women who received LA and attempted to identify 106 the characteristics of patients with factors that make LA risky. This study was approved by the 107 institutional review board of our institute (Approval code: 21-R113) and was conducted 108 according to the principles of the Declaration of Helsinki. Since this study was an observational 109 retrospective study, the requirement for written informed consent was waived, which allowed 110 us to proceed with our study in an opt-out manner. All data were analyzed using SPSS Statistics 111

Patient characteristics 114
The total number of pregnant women who were admitted for delivery was 6,404. After 115 considering inclusion and exclusion criteria, 5,184 women were finally included in the study 116 population (Fig 1). The mean maternal age at delivery was 34.6 years with a standard deviation 117 (SD) of 4.45 (Table 1). 118 The proportion of primipara was higher in the LA group (73%) than in the non-LA 127 group (56%, p < 0.001) ( Table 1). Primiparous women were more likely to receive LA (57%) 128 than multiparous women (39%). 129 To examine the association between the mode of delivery and the application or not 130 of LA, a univariate analysis was performed for IVD and ECS against SVD (Table 2). Compared 131 with SVD, both IVD and ECS were associated with a higher proportion of admission to the 132 NICU (IVD: 8.0%, p = 0.008 ECS: 12.8%, p < 0.001) and lower Apgar score (IVD 1 min/5 133 min: mean7.63/8.85, SD 1.09/0.603, p < 0.001/< 0.001 ECS 1 min/5 min: mean 7.56/8.82, SD 134 1.26/0.781, p < 0.001/< 0.001). The rate of LA was significantly higher in IVD (80%, p < 0.001) 135 than SVD (50%) and conversely lower in ECS (40%, p < 0.001). 136 Table 2 First, we extracted the risks that may affect the mode of delivery and then examined the impact 145 of LA on the mode of delivery. We hypothesized that the risks were LA, advanced maternal age, 146 lower maternal height, larger estimated fetal weight, and primipara. The primary outcome for 147 mode of delivery is shown in Table 3. We analyzed two groups: one compared SVD and ECS, 148 excluding IVD ("SVD versus ECS"), and the other compared SVD and IVD, excluding ECS 149    The secondary outcomes, including neonatal outcomes and Apgar score, are shown in Table 3. 168 LA lowered the Apgar score at 1 minute (AOR 2.06, 95%CI 1.50-2.83), but it did not influence 169 the Apgar score at 5 minutes or the NICU admission rate. Being a primipara was the strongest 170 factor that could cause all adverse neonatal outcomes after adjusting for covariates including 171 LA prescription. 172 When the population was restricted to only those who received LA, we found that 173 pregnant women with the following two risk factors demonstrated a higher risk of IVD: 174 advanced maternal age (AOR 1.70, 95%CI 1.33-2.17) and primipara (AOR 4.72, 95%CI 3.30-175 6.75) ( Table 3). Neither lower maternal height nor larger estimated fetal weight were 176 significantly associated with an increased risk of IVD . 177

178
This study included more than 5,000 pregnant women, approximately half of whom were of 179 advanced maternal age, and who were admitted to our hospital for delivery from 2017 to 2021. 180 In a multivariate analysis, LA increased the risk of IVD but decreased that of ECS. Moreover, 181 advanced maternal age and primipara increased the risks of both ECS and IVD. Among those 182 who received LA, two factors of advanced maternal age and primipara increased the risk of 183 IVD; the AOR of primipara was highest at 4,72 (95%CI 3.30-6.75) and that of advanced 184 maternal age was 1.70 (95%CI 1.33-2.17). 185 To the best of our knowledge, our study is one of few reports from Japan that presents 186 an association between LA and mode of delivery as well as neonatal outcomes. The novelty is 187 that we sought to identify the characteristics of pregnant women who may need to be cautious 188 in selecting LA. Our study is valuable in that we analyzed primarily Asian patients, who are 189 recognized to be at an increased risk of CS [6]; we also included a large number of patients of 190 advanced maternal age. 191 As our results showed that LA decreased the risk of CS, many recent reports showed 192 that LA is not associated with increased risk of CS or that LA even reduces CS rates [13,15]. 193 Despite a previous study published several decades ago, which concluded that epidural 194 analgesia was associated with a significantly increased risk of CS among primipara [6], a recent 195 Cochrane Review in 2018 that included 40 trials and more than 11,000 pregnant women found 196 that epidural analgesia did not increase the risk of CS [14]. The reason that LA is not associated 197 with and does not increase the ECS rate may be because factors associated with more painful 198 labor are themselves associated with an increased risk of CS (e.g., fetal malrotation, fetal-pelvic 199 disproportion, dysfunctional labor) [20]. Thus, we consider that pain relief might help to avoid 200 ECS. However, no studies compared the degree of pain with the CS rate, and it is unclear why 201 LA lowers the CS rate. Further investigation is needed to determine the pathophysiological 202

reasons. 203
Our results showed that LA also increases the risk of IVD. While some studies 204 23], but in our study, the Apgar score was lower in the LA group even though the CS rate 217 decreased, which suggests that LA itself must lower the Apgar score. Kumar et al. reported that 218 the Apgar score was significantly lower with use of epidural analgesia because of respiratory 219 distress, and they considered that a large volume of distribution at a steady state in neonates 220 resulted in a much longer terminal elimination half-life of fentanyl over several hours [11]. 221 Most newborns with neonatal asphyxia are able to establish fetal-to-neonatal transition after 222 appropriate resuscitation immediately after birth. This means that even if the Apgar score at 1 223 minute is low, the Apgar score at 5 minutes will be higher after appropriate resuscitation. This 224 may be the reason why the results of our study did not reveal statistically significant differences 225 when the Apgar score at 5 minutes was less than 7. Researchers suggested that LA may elicit 226 more severe adverse effects in a newborn in the immediate postnatal period than what the Apgar 227 score at 1 minute purportedly represents. Neonatologists are expected to exhibit resuscitation 228 skills for cases of neonatal asphyxia, especially in cases of LA. To promote resuscitation skills, 229 we should provide a training and educational system to maintain higher skills and knowledge. 230 Even if neonatal resuscitation is learned through simulation, this skill is difficult to maintain 231 without practical experience, and thus, aggressive resuscitation practices are important. The 232 centralization of facilities where LA is available might ensure safe deliveries. We should 233 carefully consider LA, especially in facilities without personnel, such as pediatricians, 234 neonatologists, and midwives, who attend deliveries. 235 According to our results, LA may reduce the risk of ECS but increase the risk of IVD 236 and lower the Apgar score at 1 minute. In particular, the increased risk of IVD was significantly 237 higher when the pregnant women who were primiparous and of advanced maternal age were 238 administered LA. While the decrease in ECS is favorable in terms of surgical invasiveness and 239 the safety of the next delivery, the increase in IVD is a serious issue. IVD is associated with 240 many potential risks to both the mother (e.g., perineal and vaginal tears) and the newborn (e.g., 241 brachial plexus injury with shoulder dystocia, subgaleal and intracranial hemorrhage) [21]. In conclusion, although LA is an important means to relieve pain during labor and 277 delivery, it can increase the risk of IVD and an Apgar score of less than 7 at 1 minute. Therefore, 278 we should carefully consider the indication of LA in cases of advanced maternal age and 279 primipara. Furthermore, it is desirable to maintain and improve resuscitation skills through 280 training and certification systems and to intensify facilities where LA is available. 281 282 Dear Emily Chenette, Editor-in-Chief of PLOS ONE On behalf of all co-authors of this manuscript, I would like to thank the editor and the reviewers for a thorough review of our manuscript and constructive advice, which have improved the manuscript considerably. We carefully went over the important comments/suggestions from the reviewers and revised the manuscript. Please find our point-by-point responses to the reviewers' comments below. All changes we made in this revision are highlighted in yellow. We hope that you will find the revisions and comments satisfactory.
Reviewer #1: #1. Your institute often accept the foreign pregnant women from many countries. Is it necessary to discuss about the racial difference in this study?
Response to the Reviewer: We appreciate the reviewer's question. We think that it was difficult to strictly extract race including one half or one quarter patients and exclude them. It is true that our hospital sees many foreign patients, but the majority of pregnant women who deliver in our hospital is Japanese. Furthermore, the majority of the few foreign patients is Chinese or Korean, thus we believe that not excluding the foreign patients does not have a significant impact on the results of this study dealing with Asian.
#2. I would like to know that the precise indications for deciding IVD, which may help to clear the precise delivery outcome on LA.
Response to the Reviewer: In our hospital, when immediate delivery is considered, cesarean section is selected in cases of high fetal head station, while instrumental vaginal delivery is selected in cases of low fetal head station. The high fetal head station means at +2 or more. The final decision is made by the attending obstetrician. I have added the above text to the main text. (Page 8, Line 103-104) Response to Reviewers #3. In Table 3, only the AOR and 95 % CI have shown in the IVD vs SVD with LA patients and suggested significant differences. However, the differences in estimated body weight among three groups (SVD, IVD and ECS) seem much closer and suggest no significant differences in Table 1. To avoid the misrecognition you should suggest more precise data in Table 3 or in the draft.
Response to the Reviewer: In Table 1, the univariate analysis was performed, whereas in Table 3, the multivariate analysis was performed. Not only the method of analysis but also the number of cases covered was different. The body weight was a continuous variable, whereas the sample size was large, which is the reason for the significant difference, and we consider that the significant difference has a little clinical significance.

Reviewer #2:
Although your manuscript is interesting and well-written, I think it does not have merit enough for publication in PLOS One, because it is only a retrospective cohort (please, include the study design in the title). There are lots of randomized controlled trials demonstrating maternal and neonatal outcomes with labor analgesia and a Cochrane systematic review with more than 11,000 patients included. I suggest you try another journal and follow STROBE recommendations for observational studies, with a copy of STROBE checklist for cohort studies.
Response to the Reviewer: We appreciate the reviewer's peer review. We followed the reviewer's advice to add the study design to the title and submit the STROBE checklist. Although, several retrospective cohort studies have been published before, there are very few reports from Asia. We consider that the present study has great clinical significance with increase of the advanced maternal age and popularization of the labor analgesia. 2. The idea that the premise of this study was to identify people who should avoid labor analgesia seems problematic from the start. Perhaps the focus should be on counseling patients about a potentially higher risk of needing an operative vaginal delivery and allowing the patient to decide rather than advising them that they can't get analgesia.

Response to the Reviewer:
As the reviewer's advice, the aim is to counsel patients and allow them to make informed decision. We followed the advice and revised. (Page 2, Line 17-18) Introduction: 3. May be worth mentioning that labor pain is also associated with postpartum depression and anxiety.
Response to the Reviewer: We followed the advice and revised with the additional reference. (Page 3, Line 31-33) 4. I'm again troubled that the focus of this analysis is to identify people who should avoid labor analgesia. There are outcomes worse than requiting an operative delivery, and pain and depression may be among such outcomes.
Response to the Reviewer: As we responded in the question 2, we revised the statement to focus on counseling and allowing patients to decide rather than advising them that they can't receive labor analgesia. (Page 4, Line 59-Page 5, Line 60) Methods: 5. "The estimated fetal weight was substituted for the actual birth weight because the estimated fetal weight is not always accurate due to inter-examiner reliability." Consider re-wording this as "The estimated fetal weight was replaced with the actual birth weight…" Response to the Reviewer: We followed the advice and revised. (Page 7, Line 86) 6. Consider rewording this "When termination is considered," as "When immediate delivery is considered" since termination has many meanings.
Response to the Reviewer: We followed the advice and revised. (Page 8, Line 101) 7. What are the indications for operative delivery at this institution? This may inform concerns related to this rate being higher. Many countries would interpret an increase in operative delivery over cesarean to be a success.
Response to the Reviewer: In our hospital, when immediate delivery is considered, cesarean section is selected in cases of high fetal head station, while instrumental vaginal delivery is selected in cases of low fetal head station. The high fetal head station means at +2 or more. The final decision is made by the attending obstetrician. I have added the above text to the main text. (Page 8, Line 103-104) As the reviewer concerns, a decrease of cesarean sections is desirable in obstetric aspects.
In this study, we would like to argue that labor analgesia brings not only good things but also bad things in neonatal aspect. Therefore, following the reviewer's advice, we revised the statement to focus on counseling and allowing patients to decide rather than advising them that they can't receive labor analgesia. (Page 4, Line 59-Page 5, Line 60) Results: 8. "When the population was restricted to only those who received LA, we found that pregnant women with the following four risk factors demonstrated a higher risk of IVD: advanced maternal age, lower maternal height, larger estimated fetal weight, and primipara (Table 3)." To be fair, you should do the same analysis on those who did NOT receive analgesia because you might find the same risk factors for operative delivery, so the analgesia may not have impacted this as much as you propose. 9. Indeed, for advanced maternal age and some other factors, analyzing only those with analgesia has a lower aOR than when you included everyone. Please also do the same analysis on those without analgesia.
Response to the Reviewer: As suggested, we performed the same analysis as follows. In addition, some modification was required in the analysis data.
Among patients without labor analgesia, only primiparous had higher AOR, whereas, among those with labor analgesia, not only primiparous but also advanced maternal age are associated with increased risk of IVD. The results of the analysis are provided in this letter and some have been added to Table 3 in the revised manuscript.
10. The discussion mentions that operative delivery is associated with perineal tearing but this was not actually evaluated in this study. Please consider adding this analysis as it would help advise the results and discussion.
Response to the Reviewer: Certainly, perineal tearing is one of the important complications of instrumental vaginal delivery. However, the outcome of the present study are the mode of delivery and the neonatal outcomes, and the complications that occur after delivery was not our interest.
In fact that the perineal tearing is highly recognized as a complication of instrumental vaginal delivery is referred in the manuscript [20].

Discussion:
11. The aforementioned additional analysis on those who did NOT receive analgesia must be performed. These results should be reviewed after this sentence: "Among those who Response to the Reviewer: Please refer the response to Comment 8 and 9 above.
12. I remain unconvinced of the conclusions but await additional analyses.
Response to the Reviewer: We have made corrections as described above. We appreciate your re-review.
Reviewer #4: In the Abstract Adjusted Odds Ratio -put the confidence interval in the results. Introduction Last paragraph -just put the purpose of the study. It is not the place to describe the methodology.
Response to the Reviewer: We followed the advice and revised. (Page 2, Line 21-23)

In Materials and Methods
Describe inclusion and exclusion criteria. Figure 1 contains the exclusion criteria and does not have the inclusion ones. Both should be described in the methodology.
Response to the Reviewer: We followed the advice and revised. (Page 6, Line 66-77) Lines 62 to 65 -the included text is disconnected from the rest.
Response to the Reviewer: We confirmed that the submission guideline stated, "Figure captions must be inserted in the text of the manuscript, immediately following the paragraph in which the figure is first cited (read order).". We would appreciate if the reviewer could point out specifically if we seem to be mistaken.

Bibliographic references
Standardize the presentation of bibliographic references.
Reference 19 -has the year of publication in 2 locations.
As can be seen below: 19. Wong  Response to the Reviewer: We followed the advice and revised. (Page 25, Line 331-333) Reviewer #5: In general, the manuscript is fit for publication. In discussion section, data should be analyzed carefully, and conclusion should be made cautiously. Adverse neonatal and fetal outcome, cesarean section and instrumental vaginal delivery may be strongly associated with and even may be caused by advanced maternal age, lower maternal height, larger fetal weight, and primipara, and other risk factors, but not by LA. Perhaps, LA should not be restricted or prohibited in these pregnant women and those in labor with the four kinds of risk factors mentioned above. This should be re-analyzed carefully and added into the discussion section.
Response to the Reviewer: We appreciate the reviewer's advice. We did the same analysis on those without labor analgesia. The results of the analysis are provided in this letter and have been added to Table 3 in the revised manuscript.
As the reviewer's advice, the aim is to counsel patients and allow them to make informed decision. We followed the advice and revised. Response to the Reviewer: We followed the advice and revised. We appreciate the reviewer's peer review. We adhere to the guidelines when submitting.