Background

Placenta previa is characterized by the abnormal implantation of placental tissue overlying the endocervical os [1]. It is associated with severe maternal and fetal morbidity and mortality [2]. The strongest risk factor for placenta previa is previous cesarean deliveries [3]. Along with the increasing rate of cesarean delivery, the incidence of placenta previa is increasing, and it is estimated be 1 in 200 pregnancies worldwide [4] and 1.24% in Chinese pregnancy women [15]. Secondary outcomes were transfusion blood rate, Apgar, and NICU. Blood transfusion during cesarean delivery was performed by the clinician in accordance with protocol.

Statistical analysis was completed using SPSS 21.0. Statistical assessment of our data was performed using descriptive statistics as well as t-tests, Wilcoxon rank-sum and chi-square test for continuous and categorical variables, respectively. Univariate analysis was performed to determine the role of the type of anesthesia in the outcomes, unadjusted odds ratios or beta coefficients, 95% confidence intervals, and 2-side p values were calculated. Multivariate logistic or line regressions were further performed, and adjusted odds ratios or beta coefficients were calculated, as well. Variables with a p-value < 0.05 in the univariate analysis were entered into the multivariate model. Potential confounders included gestational weeks, gravity, PAS, anterior placenta, previous cesarean delivery, previous placenta previa, antepartum hemorrhage, emergency cesarean delivery, and anesthesia-to-delivery time (min). Given that management for PAS cases is different from that for placenta previa, the results were re-calculated after excluding those cases with placenta previa complication with PAS.

Results

A total of 1234 placenta previa subjects were included in the study; 737 (59.7%) with neuraxial anesthesia and 497 (40.3%) with general anesthesia. Table 1 summarized the baseline distribution of placenta previa subjects. The neuraxial and general groups were similar in maternal age, height, weight, and BMI. Subjects with general anesthesia were delivered earlier, had more gravidities, and had a higher proportion of placenta accreta spectrum, anterior placenta, antepartum hemorrhage, emergency cesarean delivery, and history of cesarean delivery and placenta previa.

Table 1 Maternal characteristics of among included patients
Full size table

Table 2 showed the perioperative data and maternal and neonatal outcomes between the two groups. Estimated blood loss was less (558.96 ± 42.77 ml vs. 1952.51 ± 180.00 ml) and the rate of blood transfusion was lower in the neuraxial group. The preoperative hemoglobin concentration was higher in the general group. However, the postoperative hemoglobin concentration was not different between the two groups. The operating time and anesthesia-to-delivery time were shorter in the neuraxial group. For neonatal outcomes, the Apgar scores were all higher at 1-, 5-, and 10-min in the neuraxial group, and the proportion of neonatal asphyxia and admission to NICU were lower in the neuraxial group.

Table 2 Perioperative data and maternal and neonatal outcomes
Full size table

In the regression models, blood loss was less, and preoperative hemoglobin concentration and Apgar score were higher, and the rate of blood transfusion, neonatal asphyxia, and admission to NICU were lower in the neuraxial group. After adjusting anesthesia-to-delivery time, there was no substantial change in the results. After further adjusting for anesthesia-to-delivery time and other relevant confounding factors (gestational weeks, gravity, PAS, anterior placenta, previous cesarean delivery, previous placenta previa, antepartum hemorrhage, and emergency cesarean delivery), we found that the above results remained significantly (Table 3). After excluding PAS cases, the main results did not materially change, either (Supplement Tables 1, 2 and 3).

Table 3 Regression analysis for factors affecting maternal and neonatal outcomes (neuraxial vs. general)
Full size table

Discussion

In this retrospective analysis of 1234 women with placenta previa, we found that neuraxial anesthesia is associated with several benefits during cesarean delivery in our population, including decreased blood loss, lower need for blood product transfusion, and increased neonatal Apgar score, lower neonatal asphyxia and admission to NICU. We also found anesthesia-to-delivery interval had little influence on the results of the study.

The main strength of the present study is related to the relevant lager sample size in a single center during a relatively short time. Meanwhile, confounding factors were controlled by multivariable analysis to make the results more believable. Further, cases with placenta previa complication with PAS were excluded to recalculate to show the stability of the results. An obvious limitation of the study is its single center retrospective nature and the inherent limitations of retrospective data collection. While we made all efforts to objectively compare anesthesia outcomes between the two groups, it must be acknowledged that the groups likely differed in a priori anesthesia risks. Low-risk patients will be given regional and higher-risk patients a general anesthetic and it is impossible to retrospectively correct for this inevitable bias. In addition, we learned a lot about surgical and anesthetic techniques, as well caring for these patients over the study period and this may have influenced outcomes, including blood loss.

Placenta previa carried a significant risk of antepartum hemorrhage. Our 2017 systematic review and meta-analysis of 29 observational studies found that above half of placenta previa women had antepartum hemorrhage [16]. This cohort finding regarding antepartum hemorrhage is congruent with the previous meta-analysis. Placenta previa women undergoing a general anesthetic have lower preoperative hemoglobin concentration which could be related to their higher incidence of antepartum hemorrhage. Fortunately, there was a little difference in hemoglobin pre- and post-operative and had not found a difference between the postoperative hemoglobin concentration and the two groups. A possible explanation for these findings was that blood transfusion play a big role during labor and delivery. Therefore, adequate blood supply was essential for pregnant women with heavy bleeding and high risk of bleeding, such as placenta previa.

The relationship between anesthesia-to-delivery interval and adverse maternal and neonatal outcomes has been reported in retrospective studies [10, 17]. Delivery within 27 min of anesthesia start was associated with umbilical arterial pH > 7.1, and delivery within 30 min was associated with umbilical arterial pH > 7.0 [17]. In a retrospective cohort study, the authors found that prolonged anesthesia-to-delivery interval was associated with an increased relative risk for neonatal acidosis in planned cesarean deliveries [17].

We found the anesthesia-to-delivery interval was longer in the general anesthesia group. This result was inconsistent with perception. Cystoscopy and separate the adherent abdominal tissue would consume a lot of time in severe patients, such as complication with PAS. That’s why when we excluded patients with PAS, the difference was disappear between the two groups. The general anesthesia women have higher incidence of unfavorable maternal and neonatal outcomes which could be related to their longer anesthesia-to-delivery time. However, after adjusting the anesthesia-to-delivery interval, there has been no real change in the unfavorable maternal and neonatal outcomes between the two groups.

Our data showed that neuraxial anesthesia was associated with better maternal and neonatal outcomes, including less blood loss and transfusion and lower rate of neonatal asphyxia and admission to NICU. Both the patients’ background and the type of anesthesia may have influenced the results. A significantly higher risk of most complications was found in women who had a general anesthesia. The proportion placenta accreta spectrum, anterior placenta and other risk factors are higher in general group. These factors can aggravate maternal and neonatal outcomes [18,19,20,21]. In addition, retrospective and prospective studies also suggested that neuraxial anesthesia is associated with less blood loss and transfusion requirements [13, 22, 23]. Hong JY et al. [13] reported that neuraxial anesthesia received a significantly smaller transfusion than the general anesthesia for patients with placenta previa. Frederiksen MC et al. [22] also found neuraxial anesthesia decreased intraoperative blood loss and the need for blood transfusion in women with placenta previa. Meanwhile, Parekh N et al. [23] found neuraxial anesthesia was associated with a significantly reduced estimated blood loss and reduced need for blood transfusion from a larger consecutive placenta previa cases study.

A major limitation of previous studies is lack of control for confounding factors that are also associated with important outcomes such as blood loss. Given the major baseline differences between the two anesthetic groups, we offered the opportunity to assess these factors in detail through multivariable analysis in this a large single-center study. PAS is a very different from placenta previa regarding management. We further excluded placenta previa complication with PAS to evaluate the results and the results did not materially change. These suggested that neuraxial anesthesia was associated with several benefits during cesarean delivery for placenta previa women.

Placenta previa is the most common cause of massive obstetric hemorrhage and is associated with an increased incidence of massive transfusion, prolonged surgery and length of hospital stay [24]. A multi-disciplinary team (including obstetricians, neonatologists, midwives, anesthetists, critical care staff, ect.) should be approached to management of these patients. Placenta previa will become more frequently encountered by obstetric anesthetics in the future.

Both general and neuraxial anesthesia options have advantages and disadvantages for patients with placenta previa. The ideal anesthetic choice for patients with placenta previa should require individualized planning based on patients’, anesthetic and surgical factors. Patient factors include pregnant women’s preference, predicted difficult airway and contraindications to neuraxial anesthesia, and surgical factors include imaging interpretation predicting extensive or prolonged surgery.

Conclusions

This study presents a paradigm for the anesthetic management of placenta previa that is consistent with current RCOG guidelines and with data presented by other authors. Our study adds to the limited existing literature supporting neuraxial anesthesia is safe and lower risk of hemorrhage for cesarean delivery in women with placenta previa.