Prophylactic esketamine to reduce postpartum depression in primiparae

Abstract

BACKGROUND

Postpartum depression (PPD) is a common complication after childbirth, especially in primiparae.

OBJECTIVE(S)

This trial sought to evaluate whether prophylactic administration of esketamine during the perinatal period could prevent PPD in primiparae without predisposition to prenatal depression.

DESIGN

A prospective, double-blind, multicentre, randomised controlled trial.

SETTING

Three academic hospitals.

PATIENTS

Primiparae scheduled for elective caesarean section with an Edinburgh Postnatal Depression Scale (EPDS) score less than 10.

INTERVENTIONS

Postnatal women were randomly assigned to receive either i.v. esketamine in a single dose of 0.25 mg kg⁻¹ followed by 80 mg of esketamine as an adjunct to 24-h patient-controlled intravenous analgesia (PCIA). Women in the control group received an equal volume of saline.

MAIN OUTCOME MEASURES

The primary outcome was the total incidence of PPD within 3 months postpartum. Secondary outcomes included postoperative EPDS scores, numeric rating scale scores, sufentanil consumption, the number of effective presses for postoperative intravenous analgesia and adverse events.

RESULTS

A total of 322 patients were included in the modified intention-to-treat analysis. The total incidence of PPD in the esketamine group (11.59%) was significantly less than the control group (20.89%) [adjusted relative ratio (RR), 0.57; 95% CI, 0.35 to 0.94; P = 0.028] as was the incidence at 7 days postpartum (4.89 vs. 15.19%; adjusted RR, 0.32; 95% CI, 0.15 to 0.72; P = 0.005). However, there were no significant differences in PPD incidence and EPDS scores at 1, 2 and 3 months postpartum, respectively. Several mild central nervous events, such as dizziness (10.98%), hallucination (10.37%) and dissociation (5.49%), were observed during esketamine treatment.

CONCLUSION

Peri-operative adjunctive administration of esketamine is relatively safe and can prevent PPD in primi-parae without predisposition to prenatal depression in the short term.

TRIAL REGISTRATION

Clinicaltrials.gov. Identifier: NCT04860661

KEY POINTS

• Prophylactic esketamine reduced postpartum depression in primiparae at 7 days.

• The long-term benefits of esketamine are not significant.

• Central nervous events related to esketamine were generally mild and transient.

Introduction

Postpartum depression (PPD) is a common disorder that begins within the first 4 or 6 weeks postpartum.¹ Being a primipara is a well known risk factor for perinatal depression,^2–5 and 70% of new mothers may experience transient mood swings within 10 postpartum days⁶ that may affect maternal quality of life, and have a persistent negative impact on the mother–infant relationship and the infant's health, and development.^7,8 Therefore, it is essential to explore ways to reduce the incidence of PPD, especially in primiparae.

In PPD, like other common subtypes of depressive disorders, glutamatergic dysfunction and neuronal damage are present in the dorsolateral prefrontal cortex of those affected.⁹ The N-methyl-D-aspartic (NMDA) receptor serves as the binding or regulating target for both classical antidepressants and novel fast-acting antidepressants.¹⁰ In recent times, the NMDA receptor antagonist ketamine has attracted wide attention for its rapid improvement of depressive symptoms and reduction of suicidal ideas.¹¹ It can be safely used during the perinatal period, such as for analgesia in caesarean section, and to reduce the risk of PPD.^12,13 However, the central nervous side effects of ketamine limit its use in the clinic.

Esketamine is the S-enantiomer of ketamine and has a greater affinity with NMDA receptors. Like ketamine, esketamine can treat refractory depression and quickly eliminate suicidal intentions.^14,15 Due to its higher potency and fewer side effects, the effect of esketamine on PPD has been reported in recent studies.^16–19 However, the results of these studies are contradictory due to the heterogeneity of study cohorts, dosing regimens and the influence of confounding factors. Yang et al.¹⁸ reported that in women with prenatal depression [Edinburgh postnatal depression scale (EPDS) > 9], a regimen of intra-operative bolus (0.25 mg kg⁻¹) + 48 h PCIA (2 mg kg⁻¹) significantly reduced the incidence of PPD at 42 days postpartum. Conversely, Liu et al.¹⁹ found no preventive effect within 6 months postpartum using a loading dose of 0.25 mg kg⁻¹ combined with a lower dose PCIA (1.25 mg kg⁻¹) in a mixed-risk cohort, including both unaffected and depressed women. Therefore, much more work is needed to assess the therapeutic efficacy of esketamine in the prevention of PPD.

In this multicentre trial, we sought to evaluate the preventive effect of esketamine on PPD in primiparous women undergoing elective caesarean section.

Methods

Study design

This investigator-initiated, multicentre, prospective, double-blind, placebo-controlled randomised trial was conducted from 1 May 2021 to 31 March 2024 in three hospitals (Tongji Hospital, Maternal and Child Health Hospital of Hubei Province, and Jingmen Central Hospital) in China. The protocol was registered in ClinicalTrials.gov (NCT04860661) on 20 April 2021. Written informed consent was obtained from each participant before randomisation. This trial adheres to the applicable Consolidated Standards of Reporting Trials (CONSORT) guidelines.

Ethics

Ethical approval for this study (No. 2021S051) was provided by the Ethics Committee of Tongji Medical College of Huazhong University of Science and Technology, Wuhan, China on 22 March 2021.

Study participants

Women were assessed for eligibility on the day before their surgery. The trial enrolled primiparae aged 18 to 40 years with a single pregnancy who were scheduled for elective caesarean section under combined spinal-epidural anaesthesia (CSEA). The exclusion criteria included: the ASA physical status of class III or higher, increased intracranial pressure, pregnancy-induced hypertension, severe heart diseases, untreated or undertreated hyperthyroidism, severe liver and kidney function damage, mental or intellectual disability, drug or alcohol abuse, prenatal EPDS score at least 10, contraindications to spinal anaesthesia. EPDS is a symptom self-rating scale used for maternal depression, applicable both before and after delivery.²⁰ It encompasses 10 items, with each item being scored on a 4-point scale (ranging from 0 to 3), and the total score of EPDS ranges from 0 to 30 points. The Chinese version of EPDS has been validated and a score of at least 10 adopted as the screening criterion for maternal depression.^21,22 Baseline EPDS scores of all patients were evaluated in the ward on the day before surgery.

Randomisation, blinding and standardisation

Patients were randomly assigned to the esketamine group or control group in a 1 : 1 ratio. Randomised numbers were generated by an independent statistician using SAS software version 9.2 (SAS Inc., Cary, North Carolina, USA) with a block size of four, stratified by trial site. Allocations were concealed in sequentially numbered opaque envelopes and opened by investigators who were responsible for preparing the study drug and did not participate in data collection or follow-up. All women, anaesthesiologists, surgeons and researchers responsible for intra-operative data collection and postoperative follow-up were blinded to group allocation. Group assignments were not revealed until the completion of the 3-month follow-up.

Before the start of the trial, staff from different centres received standardised training, regarding the determination of inclusion and exclusion criteria, the execution of randomisation and blinding, the interpretation of outcomes, follow-up procedures and also the management and recording of data. The primary research centre conducted rigorous data review and quality control, such as checks for plausibility ranges.

Interventions

Upon arrival in the operating room, routine monitoring including electrocardiogram, pulse oximetry and noninvasive blood pressure was attached. Upper limb venous access was established, and 500 ml of 0.9% saline was infused at a rate of 10 ml kg⁻¹ h⁻¹. Oxygen was delivered to the patient at 6 l min⁻¹ via the anaesthesia breathing circuit. Women were placed in the left lateral position, and CSEA was performed by experienced anaesthesiologists at the L2–L3 or L3–L4 intervertebral space. Intrathecal injection of 1.5 to 2 ml of 0.5% ropivacaine aimed to achieve the target anaesthesia level at T4–T6. The sensory level was assessed using a 22-guage short-bevel needle and was defined as the highest dermatome with a markedly diminished or absent perception of sharpness. Vital signs were closely monitored, and if hypotension occurs, i.v. administration of ephedrine (6 to 12 mg) and balanced salt solution were used to maintain haemodynamic stability. At 15 min after clamping the neonatal umbilical cord, women in the esketamine group received 0.25 mg kg⁻¹ i.v. esketamine (Jiangsu Hengrui Pharmaceuticals Co., Ltd., China) diluted to 10 ml with 0.9% saline at a constant speed over 60 s. The esketamine dose was calculated based on ideal body weight (IBW) determined by the Lemmens formula: IBW (kg) = 22 × height² (m).²³ The control group received an equivalent volume of saline. All received 2 mg of tropisetron i.v. at the end of surgery, and PCIA was used as routine postoperative analgesia for 24 h. The PCIA pump contained 100 μg of sufentanil and 80 mg of esketamine for the esketamine group or only 100 μg of sufentanil for the control group, both diluted to 100 ml with 0.9% saline. The pump was set with a background infusion rate of 2 ml h⁻¹, a single bolus dose of 2 ml and a lockout interval of 8 min. If patients experienced severe central nervous side effects during the observation period, they were treated with low doses of midazolam or propofol and subsequently withdrawn from the trial. Patients were asked to rate their postoperative pain intensity at rest using an 11-point Numeric Rating Scale for 24 hours after surgery. A dose of 50 mg flurbiprofen axetil i.v. was administered as a rescue analgesic if the pain score was at least 4. Neonates were clinically monitored by experienced obstetric nurses for 72 h after delivery in accordance with the routine monitoring procedures of the ward. Any observed abnormalities potentially related to drug exposure were documented.

Outcomes

The primary outcome was the total incidence of PPD within 3 months postpartum, which was evaluated by the EPDS scores at 7 days, 1, 2, and 3 months after delivery through face-to-face or telephone interviews, and defined as a score of at least 10.

The secondary outcomes include EPDS scores assessed before surgery, and at 7 days, 1, 2 and 3 months after surgery. Within 24 h after surgery, NRS score, ranging from 0 (indicating painless) to 10 (indicating worst pain), was used to estimate postoperative pain intensity. The number of effective presses of the PCIA and sufentanil consumption were also recorded. Adverse events associated with esketamine were monitored within 1 h after drug administration and on the first day after surgery, including nausea or vomiting, dizziness, hallucination, nystagmus, dissociation and diplopia, and treated if necessary. The duration of surgery, estimated blood loss and Apgar scores of newborns at 1 and 5 min after birth were also recorded.

Sample size

The primary outcome of this trial was the total incidence of PPD after caesarean section. According to a previous study, peri-operative administration of ketamine reduced the prevalence of PPD from 19.6 to 12.8%.¹³ Power Analysis and Sample Size (PASS) 15.0.5 software was used to calculate the sample size. With a type I error rate of 0.05 (α = 0.05), a power of 90% (β = 0.1) and a dropout rate of 20%, 1248 patients were needed.

During the trial, new evidence emerged from published single-centre RCTs demonstrating esketamine's superior antidepressant efficacy over ketamine.^16–19 Prompted by these findings and ethical imperatives to minimise unnecessary participant exposure, we prospectively recalculated the sample size using Li et al.'s²⁴ meta-analysis data (control group PPD: 22.37% vs. esketamine group: 8.38%). This evidence-driven adjustment reduced the target from 1248 to 336 patients while maintaining 90% power (α = 0.05). Separately, during routine Data Safety Monitoring Board (DSMB) safety surveillance (unblinded for adverse event monitoring only), preliminary trends suggesting early benefit at 7 days were noted. Critically, this nonprespecified observation did not influence sample size recalculation or trial conduct. All efficacy endpoint analyses remained fully blinded until trial completion.

Statistical analysis

All the analyses were conducted primarily in the cohort included in the modified intention-to-treat analysis, that is all patients who received the randomised allocated intervention (esketamine/saline) and completed at least one follow-up PPD screening. Patients who required conversion to general anaesthesia or failed to complete one follow-up were excluded. The Kolmogorov–Smirnov test was applied to determine the normality of the data distribution. Continuous variables with normal distribution were described as mean ± SD and intergroup differences were evaluated using ANOVA or independent Student's t test. The nonnormally distributed data were described as median [IQR] and compared by the Mann–Whitney U test. In addition, the pseudo-median differences (MED) and 95% CI were calculated with Hodges–Lehmann estimate. Categorical variables were presented as numbers (%) and analysed by the χ² test, continuity corrected χ² test or Fisher's exact test. Generalised linear model was used to calculate the adjusted relative ratios (RR) and 95% CI.

Additionally, to investigate whether the effects of esketamine vary among different groups, subgroup analysis was performed based on individual and clinical characteristics of patients, such as age, BMI and comorbidities. The interactions between the treatment effect and each predefined subgroups were assessed separately using logistic regression. A two-tailed P value of less than 0.05 was considered statistically significant. Statistical Package for Social Sciences (SPSS) software version 27.0 and R statistical software version 4.4.1 were used for statistical analysis. Specific statistical strategies are detailed in Supplement 1.

Results

Study population

A total of 1263 primiparae were initially screened, and 336 were eligible and randomly assigned to either esketamine (n = 168) or control group (n = 168) (Fig. 1). In the esketamine group, 1 primipara was converted to general anaesthesia due to failed intrathecal anaesthesia, 3 were lost to follow-up and 164 were finally included in the modified intention-to-treat analysis. In the control group, 4 primiparae were converted to general anaesthesia, 6 were lost to follow-up and 158 were finally included in the analysis. The baseline individual and clinical characteristics as well as intra-operative data of primiparae and neonates were well balanced between the two groups (Table 1). Baseline data of patients in each centre are presented in Supplement 2. The median age of all primiparae was 29 years [28 to 32]. The baseline median EPDS scores in the esketamine group and control group were 6 and 5, respectively.

Fig. 1.

Flow diagram of participants.

EPDS, Edinburgh Postnatal Depression Scale.

Table 1.

Baseline characteristics and intra-operative data of primiparae and neonates

Variables	Esketamine group (n = 164)	Control group (n = 158)	P value
Age (years)	29 [27 to 32]	29.5 [28 to 32]	0.200
Height (cm)	160 [157 to 163]	160 [156 to 163]	0.848
Weight (kg)	70.0 [64.6 to 77.9]	70.4 [63.4 to 77.1]	0.607
Prenatal BMI (kg m−2)	27.2 [25.3 to 29.6]	27.3 [24.9 to 29.7]	0.826
Duration of education (years)			0.759
<9	0	0
9 to 12	26 (15.85)	23 (14.56)
>12	138 (84.15)	135 (85.44)
Gestational age (days)	273 [267 to 277]	272.5 [268 to 277]	0.912
Comorbidities
GDM	22 (13.41)	22 (13.92)	>0.999
Gynaecological diseasesa	18 (10.98)	22 (13.92)	0.500
History of gynaecological surgeryb	13 (7.93)	20 (12.66)	0.199
Hypothyroidism	9 (5.49)	11 (6.96)	0.648
Autoimmune disease	6 (3.66)	6 (3.80)	>0.999
Sinus tachycardia	17 (10.37)	17 (10.76)	>0.999
Hct (%)	36.6 [34.3 to 38.7]	35.8 [34 to 38]	0.061
Baseline EPDS score	6 [4 to 7]	5 [3 to 8]	0.971
Duration of surgery (min)	55 [47 to 65]	58 [47 to 70]	0.296
Apgar score
1 min	9 [9 to 9]	9 [9 to 9]	0.510
5 min	9 [9 to 10]	9 [9 to 9]	0.157
Blood loss (ml)	200 [200 to 300]	200 [200 to 347]	0.871

Data are presented as median [IQR] or n (%). BMI, body mass index; EPDS, Edinburgh Postnatal Depression Scale; GDM, gestational diabetes mellitus; Hct, haematocrit.

^aHysteromyoma, teratoma, ovarian chocolate cyst, polycystic ovary syndrome, ectopic pregnancy and septate uterus.

^bLaparoscopic myomectomy, laparoscopic oophorocystectomy, teratoma resection, salpingectomy and transcervical resection of septum.

^cAntiphospholipid antibody syndrome, rheumatoid arthritis, psoriasis, Hashimoto's thyroiditis and Graves’ disease.

Outcomes

The total incidence of PPD within 3 months postpartum in the esketamine group (11.59%) was significantly lower than that in the control group (20.89%) (adjusted RR 0.57; 95% CI, 0.35 to 0.94; P = 0.028), also the incidence of PPD at 7 days postpartum (4.89 vs 15.19%; adjusted RR 0.32, 95% CI, 0.15 to 0.72, P = 0.005). However, the differences between the two groups disappeared at 1, 2 and 3 months postpartum. The EPDS scores in the esketamine group were also significantly lower than those in the control group at 7 days postpartum (MED − 1, 95% CI, −1 to 0, P = 0.039), while there were no significant intergroup differences at 1, 2 and 3 months postpartum (Table 2).

Table 2.

The prevalence of postpartum depression and Edinburgh Postnatal Depression Scale scores

	Esketamine group (n/total)	Control group (n/total)	Estimated effects (95% CI)a	P value
The prevalence of PPD
7 days	8/164 (4.89)	24/158 (15.19)	Adjusted RR: 0.32 (0.15 to 0.72)	0.005
1 month	13/158 (8.23)	16/149 (10.74)	Adjusted RR: 0.82 (0.43 to 1.59)	0.561
2 months	8/149 (5.37)	13/145 (8.97)	Adjusted RR: 0.71 (0.33 to 1.52)	0.381
3 months	7/146 (4.79)	10/143 (6.99)	Adjusted RR: 0.79 (0.31 to 2.00)	0.615
EPDS scores
7 days	4 [2 to 6]	5 [2 to 8]	Median difference: −1 (−1 to 0)	0.039
1 month	4 [2 to 7]	[2 to 7]	Median difference: 0 (−1 to 0)	0.576
2 months	4 [2 to 7]	4 [2 to 7]	Median difference: 0 (−1 to 1)	0.694
3 months	3 [2 to 5]	4 [2 to 7]	Median difference: 0 (−1 to 0)	0.108

Data are presented as median [IQR] or n (%). EPDS, Edinburgh Postnatal Depression Scale; PPD, postpartum depression.

^aAdjusted for: age, prenatal BMI, duration of education, gestational age, comorbidities, sinus tachycardia, Hct and baseline EPDS score.

Subgroup analysis showed no significant interaction between esketamine's prophylactic effect on PPD within 3 months postpartum and the predefined subgroups (Fig. 2). Additionally, subgroup analyses and the forest plot for PPD at 7 days postpartum are shown in Supplement 2, with no observed interactions between subgroups.

Fig. 2.

Subgroup analyses and forest plot for postpartum depression within 3 months postpartum.

BMI, body mass index; EPDS, Edinburgh Postnatal Depression Scale; GDM, gestational diabetes mellitus; NA, not available.

NRS scores were similar between the esketamine group and control group at two postoperative hours, 1 [0 to 2] vs. 1 [0 to 2], at 6 h, 2 [1 to 4] vs. 2 [1 to 4], at 12 h, 2 [1 to 4] vs. 2 [1 to 4], at 18 h, 2 [1 to 4] vs. 2 [1 to 4] and at 24 h, 2 [1 to 4] vs. 2 [1 to 5] (all P > 0.05) (Fig. 3). The median consumption of sufentanil in PCIA in the esketamine group and control group was 60 and 62 μg, respectively, with no significant intergroup differences (P = 0.130). Although the number of effective presses in PCIA in the esketamine group, 6 [3 to 10] was lower than that in the control group, 7 [3.5 to 11], there was no statistical difference (P = 0.140).

Fig. 3.

Postoperative pain-related indicators.

(a) Numeric Rating Scale (NRS) Pain Scores at 2, 6, 12, 18 and 24 h postpartum; (b) the number of effective presses of the PCIA; (c) cumulative sufentanil consumption 24 h postoperatively. Data are presented as violin plots, which display the full distribution and probability density of the data. Within each violin, the thick black horizontal line represents the median, and the thin dashed black lines indicate the first and third quartiles. The width of the violin at any point reflects the proportion of data points at that value. The extreme ends of the plots represent the maximum and minimum observed values.

During caesarean section, nausea or vomiting occurred in five primiparae (3.05%) in the esketamine group within 1 h after single-dose administration of esketamine, and none in the control group (Table 3). However, this difference was not statistically significant (P = 0.078). Notably, the incidences of dizziness (10.98% vs. 0, P < 0.001), hallucinations (10.37% vs. 0, P < 0.001) and dissociation (5.49% vs. 0, P = 0.008) in the esketamine group were significantly higher. Nevertheless, these adverse reactions were of brief duration and vanished before the end of surgery. The patients did not require extra interventions. Among 17 primiparae with hallucinations, 12 reported that this experience was euphoric, 3 were unpleasant and 2 were neutral. Nystagmus (1.81%) was also observed in the esketamine group, while there were no significant differences between the two groups (P = 0.259).

Table 3.

Adverse events during or after caesarean section

	Esketamine group (n = 164)	Control group (n = 158)	P value
Within 1 h after drug administration
Nausea or vomiting	5 (3.05)	0	0.078
Dizziness	18 (10.98)	0	<0.001
Hallucination	17 (10.37)	0	<0.001
Euphoric	12 (7.32)	0	<0.001
Unpleasant	3 (1.83)	0	0.259
Neutral	2 (1.22)	0	0.499
Nystagmus	3 (1.81)	0	0.259
Dissociation	9 (5.49)	0	0.008
24 h postpartum
Nausea or vomiting	26 (15.85)	20 (12.67)	0.430
Dizzinessa	21 (12.80)	9 (5.70)	0.034
I	19 (11.59)	9 (5.70)	0.075
II	1 (0.61)	0	>0.999
III	1 (0.61)	0	>0.999
Diplopia	1 (0.61)	0	>0.999

Data are presented as n (%).

^aGrade I: vertigo episodes do not affect daily activities during or after the attack. Grade II: daily activities are interrupted during the attack but fully resume quickly after. Grade III: most daily activities are self-manageable after the attack. Grade IV: most daily activities cannot be self-managed after the attack. Grade V: all daily activities require assistance after the attack.

During postoperative analgesia, 26 primiparae (15.85%) in the esketamine group and 20 (12.67%) in the control group experienced nausea or vomiting, with no significant difference between the groups (P = 0.430). Tropisetron was intravenously administered to three patients in the esketamine group and two patients in the control group. The remaining patients experienced mild symptoms and recovered spontaneously within hours. The incidence of dizziness (12.80 vs. 5.70%) was higher in the esketamine group (P = 0.034). Most cases (90.48%) were Grade I, and daily activities of patients were not affected. Although one case was Grade II and one was Grade III, which was moderate, patients did not require additional treatment, and none of them developed severe dizziness. One primipara in the esketamine group reported mild diplopia, which emerged on the day of surgery and vanished on the following day. No patients experienced hallucination or dissociation after surgery. No adverse reactions such as sedation or feeding difficulties were observed in infants after breastfeeding.

Discussion

This trial found that in primiparae without predisposition to prenatal depression, the administration of esketamine reduced the risk of PPD and EPDS scores at 7 days postpartum, but the observed effects were not long-term. Although central nervous side effects of esketamine occurred during medication, most symptoms were mild and self-limiting.

Esketamine, a derivative of ketamine, is believed to have more potent and more enduring antidepressant effects.²⁵ Although the terminal half-life of esketamine is approximately 11 h in humans, preclinical and clinical studies indicate that esketamine and ketamine exhibit a long-lasting antidepressant effect lasting about several weeks.^26–30 A recent animal experiment showed that after a single dose of ketamine, the drug was trapped and isolated from metabolic degradation, sustainably inhibiting burst firing and blocking NMDA receptors, so producing a continuous antidepressant effect.²⁸ In published clinical studies, the antidepressant effect and duration of ketamine and esketamine vary widely.³¹

The prenatal mental state and drug dose critically influence esketamine's preventive efficacy against PPD. In this trial, we evaluated primiparae without prenatal depression (EPDS < 10) and observed that esketamine significantly reduced PPD incidence at 7 days postpartum, yet this effect attenuated at 1 month postpartum. This transient efficacy aligns with studies in similar low-risk cohorts.^18,32,33 Chen et al.³² reported that a single intravenous injection (0.25 mg kg⁻¹) and PCIA (1 mg h⁻¹) with esketamine for 48 h reduced PPD risk at 7 days postpartum but not at 2 weeks. Han et al.³³ gave esketamine (0.5 mg kg⁻¹) by PCIA and noted benefits limited to 14 days despite higher esketamine exposure. Crucially, divergent outcomes emerge in high-risk cohorts.^18,34 A large multicentre clinical trial demonstrated that an intravenous infusion of 0.2 mg kg⁻¹ of esketamine reduced PPD risk for women with prenatal depressive symptoms within 42 days postpartum.³⁴ Another trial found that low-dose esketamine (1 mg kg⁻¹) by PCIA reduced PPD risk at 7 days postpartum, while high-dose (2 mg kg⁻¹) extended the reduction in PPD risk to 42 days postpartum.¹⁸ Current research indicates that the pharmacodynamic response to esketamine appears to differ between euthymic and subclinically depressed parturients. Our exclusion of prenatal depression intentionally isolated esketamine's biological effects in this nondepressed cohort, a group representing the majority seen in clinical practice, but often underrepresented in PPD trials. Future studies should stratify participants by baseline EPDS scores to establish precision protocols.

A statistically significant median difference of 1 point in EPDS scores was also observed at day 7, though its clinical interpretation warrants caution. The minimal clinically important difference (MCID) value serves to identify interventions conferring substantive individual-level improvement. In patients at risk of depression, a change of 3 or 4 points in EPDS after intervention is regarded as deterioration or improvement of MCID.³⁵ Our observed 1-point difference falls below this threshold, indicating limited symptomatic benefit at the individual level. Nevertheless, in our euthymic cohort, the difference represents a 16.7% relative reduction from baseline symptoms. Moreover, esketamine significantly reduced PPD incidence at 7 days. Thus, while individual EPDS changes were modest, the compound evidence demonstrates tangible short-term prevention at the group level.

Ketamine was originally used in obstetric surgery for its sedative and analgesic properties. Adhikari et al.³⁶ suggested that intravenous ketamine reduced pain scores at 2 and 6 h postoperatively. Han et al.³³ indicated that esketamine as an adjuvant in PCIA effectively relieved pain within 48 h after caesarean delivery. However, in our trial, the pain scores within 24 h after surgery did not differ between the groups. Similarly, Liu et al.¹⁹ also found no difference between groups in pain scores on the first and second days postpartum. Two randomised controlled trials have shown that the analgesic effect of esketamine did not significantly reduce resting pain but did play a role in reducing motion pain.^16,32 Furthermore, a large clinical trial found that 0.25 mg kg⁻¹ of esketamine given to parturients before incision did not reduce intra-operative or postoperative pain but may benefit those with higher pain levels.³⁷ Therefore, the adjunctive analgesic effect of esketamine may not be ideal and appears to be related to the dose and intensity of pain.

In our trial, none of the patients experienced serious adverse reactions. Central nervous adverse reactions such as dizziness, hallucinations and dissociation increased within 1 h after esketamine administration. These reactions were transient and most subsided before the end of surgery. Consistent with other studies,^38,39 most patients with hallucinations also reported euphoria. However, although reported hallucinations were mainly characterised as positive experiences, these phenomena demand vigilant clinical monitoring. Continuous pulse oximetry and electrocardiography should be in place during and after esketamine administration; individual susceptibility varies, with rare cases progressing to distress or agitation. Benzodiazepines, such as midazolam, should be prepared in advance to manage emergency symptoms. During the first 24 h postoperatively, the primary central nervous side effect observed was dizziness, which was generally mild (Grade I) and did not affect daily activities. Overall, while esketamine can cause central nervous side effects, they are typically mild and short-lived.

Our study has several limitations. First, several confounding factors for PPD, such as socio-economic status, family support, stressful life events, marital satisfaction and newborn health status, were not collected or analysed, which may affect our findings.⁵ Second, central nervous side effects following intra-operative intravenous administration of esketamine may lead to detection bias. To mitigate this bias, we employed independent and blinded EPDS score evaluation to enhance the reliability of the data. Third, the selection of PCIA over neuraxial or regional analgesic techniques primarily was driven by the need to maintain therapeutic esketamine concentrations through continuous delivery and ensure uniform background analgesia. We acknowledge this approach contradicts Enhanced Recovery After Caesarean (ERAC) principles, as PCIA may delay recovery and ambulation. Future studies should evaluate combining intrathecal opioids⁴⁰ with intravenous esketamine to align with ERAC protocols without compromising PPD prophylaxis. Fourth, the reduced sample size may diminish precision in long-term outcome estimates and constrain subgroup analyses. However, the recalculated sample size maintained greater than 90% power for detecting clinically significant differences. The observation of preliminary trends during DSMB safety monitoring was unblinded and not prespecified. While this allowed real-time risk-benefit assessment, it introduces potential bias. Thus, these exploratory observations from safety monitoring should be interpreted as hypothesis-generating only.

Conclusion

Peri-operative intravenous esketamine combined with PCIA can reduce the short-term risk of PPD in primiparae without tendency to prenatal depression. However, the long-term benefits are not significant. Although esketamine has relatively higher central nervous adverse events, most are mild, brief and self-limiting. Overall, esketamine presents a favorable risk–benefit profile for PPD prophylaxis in this cohort, though clinicians should remain alert to manageable central nervous side effects.