A single intravenous administration of a sub-anesthetic ketamine dose during the perioperative period of cesarean section for preventing postpartum depression: A meta-analysis

Postpartum depression (PPD) is a common depressive episode occurring after childbirth and is associated with substantial morbidity for both mother and infant; prevalence in China is reported at about 22%. Over the past two decades ketamine, an N-methyl-D-aspartate (NMDA) receptor antagonist, has attracted attention as a rapid-acting antidepressant with multiple putative mechanisms relevant to depression, including modulation of glutamatergic transmission, downstream AMPA activation, increases in brain-derived neurotrophic factor (BDNF) and mTOR signalling, synaptogenesis, anti-inflammatory effects and impacts on circadian regulation and sleep. Li and colleagues argued that, despite growing primary studies, no prior meta-analysis had synthesised evidence on whether a single perioperative sub‑anaesthetic intravenous ketamine dose given around cesarean section prevents PPD. The present study therefore aimed to evaluate the efficacy, safety and dose‑response of prophylactic ketamine administered during the perioperative period of elective cesarean section for reducing PPD, examining short‑term (within 1 week) and longer (around 4 weeks) postpartum outcomes as well as adverse events and postoperative pain.

This work is a systematic review and meta-analysis. The investigators searched seven databases in English and Chinese (PubMed, Embase, Web of Science, The Cochrane Library, CNKI, VIP and Wanfang) from inception to 1 December 2020 using keywords related to ketamine, cesarean/delivery, postpartum depression and EPDS. Randomised controlled trials (RCTs) and observational studies meeting eligibility criteria were considered; the extracted text reports that 9 RCTs and 1 retrospective study met inclusion criteria. Eligibility for included participants required American Society of Anesthesiologists (ASA) grade I–II, age 18 or older, elective cesarean section under spinal anaesthesia. Exclusion criteria included pathological obstetric complications, significant cardiovascular disease, psychosis, substance abuse, contraindications to ketamine or spinal anaesthesia, significant haemorrhage and poor compliance. The intervention was a single perioperative intravenous administration of a sub‑anaesthetic ketamine dose, compared with an equal volume of normal saline. Outcomes prespecified in the extraction included PPD score within 1 week postpartum, prevalence of PPD within 1 week, PPD score at about 4 weeks, incidence of dizziness, vomiting and total adverse events, and 48 h postoperative pain score. Two reviewers independently screened studies, extracted data (study and participant characteristics, interventions and outcomes) and assessed quality; disagreements were resolved by a third reviewer. Risk of bias in RCTs was appraised using Cochrane Handbook methods, while the retrospective study was assessed with the Newcastle–Ottawa Scale. Meta-analyses used Review Manager 5.4 and Stata (versions 12.0 and 16.0). Continuous outcomes were pooled as standardised mean differences (SMD) with 95% confidence intervals (CI) because different scales were used; dichotomous outcomes were pooled as risk ratios (RR) with 95% CI. Heterogeneity was assessed with Cochran's Q and I2; the authors predefined use of fixed‑effect models when p > 0.1 and I2 < 50% and random‑effects models otherwise. They also planned subgroup analyses (dose and administration method), leave‑one‑out sensitivity checks, univariate meta‑regression to explore heterogeneity and tests for publication bias using Egger's test and Duval and Tweedie’s trim‑and‑fill procedure.

The search identified 1,176 records initially; after screening 10 studies (9 RCTs and 1 retrospective study) with a total of 2,087 participants were included. Risk-of-bias appraisal indicated that all studies reported random sequence generation; only three reported blinding and allocation concealment clearly, while others were rated as unclear for those domains. There were losses to follow-up in three studies but these were reported as balanced and judged low risk for incomplete outcome data. Primary efficacy outcomes: Ten studies contributed data on PPD score within 1 week postpartum. Pooled analysis using a random‑effects model (heterogeneity I2 = 97%) showed a significant reduction in early PPD scores with ketamine versus control (SMD = -1.10, 95% CI -1.67 to -0.52, p = 0.0002). Eight studies reported prevalence of PPD within 1 week; pooled prevalence was lower in the ketamine group (RR = 0.49, 95% CI 0.33–0.74, p = 0.0007) with moderate heterogeneity (I2 = 55%). Four studies reported PPD score after 4 weeks; pooled effect showed no significant difference between groups (SMD = -0.16, 95% CI -0.32 to -0.01) with p = 0.06 and I2 = 52%, i.e. the short‑term benefit was not demonstrably maintained at about 4 weeks. Safety and secondary outcomes: Eight studies reported vomiting, which occurred more frequently with ketamine (RR = 1.85, 95% CI 1.32–2.59, p = 0.0003; I2 = 0%). Dizziness was reported in six studies and showed a non‑significant trend (RR = 2.9, 95% CI 0.97–8.65, p = 0.06). Seven studies reported total adverse events and found no significant between‑group difference (RR = 0.87, 95% CI 0.58–1.30, p = 0.50). Three studies contributed 48 h postoperative pain scores; pooled analysis indicated lower pain scores with ketamine (SMD = -1.85, 95% CI -3.63 to -0.07, p = 0.04) but with very high heterogeneity (I2 = 99%). Subgroup and moderator analyses: Dose‑stratified subgroup analysis suggested that ketamine at 0.5 mg/kg or higher significantly reduced both PPD score and prevalence within 1 week, whereas 0.25 mg/kg did not. Different administration methods (single perioperative intravenous injection, an analgesic pump delivering ketamine continuously for 48 h postpartum, or a combination) all showed efficacy on the early outcomes, but subgrouping did not eliminate overall heterogeneity. Univariate meta‑regression on PPD score within 1 week identified puerperal age, body mass index (BMI) and administration method as significant moderators, individually explaining 43.33%, 82.50% and 76.21% of between‑study heterogeneity respectively. BMI and age were negatively associated with ketamine response, and continuous infusion via an analgesic pump for 48 h postpartum was associated with greater protection against PPD compared with a single injection. Ketamine dose, the PPD assessment scale and timing within the first week were not significant moderators. Publication bias: Egger’s test suggested potential bias for the two early PPD outcomes, but trim‑and‑fill analysis reportedly found no missing studies and concluded that the main results were unaffected. No significant publication bias was detected for other outcomes according to the reported tests.

Li and colleagues interpreted the pooled findings as evidence that a single perioperative sub‑anaesthetic dose of ketamine administered around elective cesarean section significantly reduces PPD symptom scores and the prevalence of PPD within the first postpartum week, with an apparent minimum effective intravenous dose of about 0.5 mg/kg. The early antidepressant effect did not appear to be sustained to the fourth postpartum week in the limited data available. The study team also observed an analgesic benefit at 48 h and no increase in total adverse events, although vomiting was more frequent with ketamine. The authors emphasised that significant heterogeneity was present across studies and, by meta‑regression, identified age, BMI and mode of administration as important moderators. They proposed biological explanations linking ketamine’s diverse mechanisms to these moderators: BDNF signalling and BDNF Val66Met polymorphism as modulators of ketamine response; sleep and circadian effects (including slow wave sleep and clock‑gene expression) as relevant pathways; and metabolic/immune factors related to BMI and metabolic syndrome that could alter inflammatory status and drug distribution. The investigators noted that continuous administration via an analgesic pump for 48 h may offer greater prophylactic effect than a single bolus, though all administration routes showed some efficacy. On safety, the authors stated that single sub‑anaesthetic ketamine doses commonly produce transient side effects (headache, dizziness, dissociation, nausea, blood pressure changes) which are usually tolerable; they cautioned about contraindications (psychosis, glaucoma, uncontrolled hypertension, cardiac failure, cerebrovascular disease, substance abuse) and recommended perioperative precautions and symptomatic measures (fasting, antiemetics, analgesics, comfortable environment). They also highlighted concerns reported with repeated or high‑dose use (urological, hepatic, cognitive and dependency risks) as reasons to limit exposure. Limitations acknowledged by the authors included substantial heterogeneity that was not fully explained by subgrouping, incomplete reporting of potential moderators across primary studies (for example BDNF genotype and sleep measures were not recorded in all studies), some unclear blinding/allocation concealment across included trials and the inclusion of a retrospective study. They recommended further research to clarify mechanisms, to test adjunctive strategies such as melatonin, to explore dosing and administration regimens that prolong effect, and to examine consistency with ketamine guidance from treatment‑resistant depression practice.

The authors conclude that a single perioperative sub‑anaesthetic intravenous ketamine dose of 0.5 mg/kg or higher during elective cesarean section appears to be a safe and effective prophylactic intervention to reduce PPD symptom scores and prevalence within the first postpartum week, and to lower 48 h postoperative pain. Continuous postoperative delivery via an analgesic pump may confer greater protection than a single intraoperative injection. Age and BMI were identified as negative predictors of response, and vomiting occurred more often with ketamine though total adverse event rates did not differ. The authors call for further research to elucidate mechanisms, optimise administration strategies and evaluate adjuncts such as melatonin to enhance efficacy and safety.