Efficacy and safety of dexamethasone in postoperative recovery following hysterectomy: a systematic review and meta-analysis

Abstract

Objectives

Hysterectomy, a common surgical procedure, is frequently associated with moderate-to-severe postoperative pain and a high incidence of postoperative nausea and vomiting (PONV). Dexamethasone, a corticosteroid, may help alleviate these symptoms; however, existing evidence is largely drawn from mixed surgical populations and does not specifically address its efficacy and safety in hysterectomy patients. This meta-analysis provides a focused and updated synthesis of randomised controlled trials (RCTs) in this population, incorporating time-stratified pain outcomes and subgroup analyses by dose, surgical approach, timing and route of administration to evaluate the role of dexamethasone in postoperative recovery.

Design

Systematic review and meta-analysis using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach.

Data sources

PubMed, Scopus, Google Scholar and The Cochrane Central Register of Controlled Trials (CENTRAL) were searched through 1 November 2024.

Eligibility criteria for selecting studies

We included RCTs comparing dexamethasone with placebo for postoperative outcomes in hysterectomy patients.

Data extraction and synthesis

Two independent reviewers used standardised methods to search, screen and code included studies. Risk of bias was assessed using the Cochrane Collaboration and Evidence Project tools. Meta-analysis was conducted using random effects models. Findings were summarised in GRADE evidence profiles and synthesised qualitatively.

Results

15 RCTs (1362 patients) were included. Dexamethasone significantly reduced PONV (risk ratio (RR): 0.53, 95% CI 0.47 to 0.61, p<0.00001, I²: 0% high certainty) and pain scores at 24 hours (mean difference (MD): −0.20, 95% CI −0.35 to −0.05, p=0.009, I²=0%, moderate certainty), 8–12 hours (MD: −0.60, 95% CI −0.88 to −0.31, p<0.0001, I²: 27%, moderate certainty and 4 hours (MD: −0.43, 95% CI −1.07 to 0.21, p=0.19, 93%, moderate certainty). It also decreased the use of rescue antiemetics (RR: 0.57, 95% CI 0.43 to 0.75, I²: 39%, high certainty) and postoperative opioid consumption (standardised MD: −0.48, 95% CI −0.90 to −0.05, p=0.03, I²: 74%, low certainty). The effects of rescue analgesics and hospital stay duration were nonsignificant. Subgroup analyses showed consistent antiemetic efficacy of dexamethasone across doses, timings, routes and procedures. For pain, greater analgesic effects were seen with higher doses and perineural administration, particularly at 8–12 hours. The risk of bias was low in most studies, but evidence of publication bias was observed for the pain score outcome.

Conclusions

Dexamethasone is an effective adjunct in hysterectomy, significantly reducing PONV and postoperative pain at 8–12 and 24 hours, particularly with 4–10 mg doses. Benefits are consistent across routes, timings and surgical approaches, with greater early analgesia after perineural use. It reduces opioid consumption but has a limited effect on rescue analgesia, supporting its role as a complementary analgesic. While generally considered safe, current safety data are limited, highlighting the need for further research. These results support its use in multimodal recovery protocols and identify priorities for future studies in high-risk and diverse surgical populations.

PROSPERO registration number

CRD42024608067.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• This systematic review and meta-analysis included 15 randomised controlled trials focused exclusively on the role of dexamethasone in patients undergoing hysterectomy.

• Comprehensive subgroup analyses were conducted by dose, timing, route of administration and surgical approach.

• The Grading of Recommendations, Assessment, Development and Evaluation framework was used to assess the certainty of evidence across outcomes.

• The risk of bias was low in most studies, and heterogeneity was generally low for primary outcomes.

• Some methodological limitations include potential publication bias and limited data for certain subgroups (eg, oral administration, high-dose regimens) due to the lack of available data.

Introduction

The incidence of hysterectomy remains significantly high, with over 6 million procedures performed globally each year.¹ While estimates suggest that hysterectomy rates are substantial in many regions, they vary significantly depending on healthcare systems, cultural factors and demographic patterns.² Abdominal hysterectomy is associated with moderate-to-severe postoperative pain,³ which is commonly believed to be a risk factor for postoperative nausea and vomiting (PONV).⁴ Dexamethasone, a corticosteroid widely used perioperatively, is known to reduce PONV following a hysterectomy.⁵ First reported as an antiemetic in 1981,⁶ studies have demonstrated that dexamethasone significantly reduces PONV when used prophylactically compared with placebo (normal saline) or ondansetron.⁷ Beyond PONV prevention, dexamethasone has also shown efficacy in reducing postoperative pain in various surgeries, including arthroplasty, hysterectomy and laparoscopic procedures.8,10

Dexamethasone likely inhibits prostaglandin synthesis and reduces serotonin turnover in the central nervous system, contributing to its anti-nausea and analgesic effects.⁷ Additionally, it relieves pain by inhibiting peripheral phospholipase activity, lowering pain-related products from the cyclooxygenase and lipoxygenase pathways.¹¹ With 60%–80% of PONV cases linked to inadequate prophylactic antiemetics, PONV can lead to complications like delayed recovery and prolonged hospital stays.¹² Dexamethasone and ondansetron are both low-cost, but dexamethasone may be more cost-effective perioperatively due to its added benefits in reducing pain and opioid use.¹³ On the other hand, a single 40 mg intravenous dose of dexamethasone, when administered preoperatively, reduces both pain and CRP levels within 48 hours, correlating with improved recovery outcomes.¹⁴ However, dexamethasone is not without risks, including hyperglycaemia, increased infection risk and delayed wound healing, which necessitate careful consideration of optimal dosing to balance its benefits and potential side effects.¹⁵ Given its potential role within opioid-sparing strategies, dexamethasone may contribute to reduced opioid use and related adverse effects in perioperative pain management.¹⁶

Despite its widespread use, the efficacy and safety of dexamethasone in enhancing postoperative recovery after hysterectomy remain inadequately defined. While multiple trials17,31 have assessed its role in preventing PONV and improving analgesia, previous reviews have either lacked a hysterectomy-specific focus or failed to stratify outcomes by time point, dose, route of administration and surgical approach. This meta-analysis addresses these gaps by systematically evaluating high-quality randomised controlled trials (RCTs) to determine dexamethasone’s effects on PONV and pain scores at 4, 8–12 and 24 hours, along with secondary outcomes such as rescue analgesia, antiemetic use, opioid consumption and hospital stay. By offering a detailed synthesis across clinically relevant subgroups, our findings enhance the applicability of evidence to perioperative protocols and support more individualised, evidence-based care in hysterectomy patients.

Methodology

This systematic review and meta-analysis was reported following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines³² and conducted following the Cochrane Handbook for Systematic Reviews of Interventions.³³ A prospectively registered protocol is available on PROSPERO (International Prospective Register of Systematic Reviews)³⁴ with ID: CRD42024608067.

Literature search

We conducted a comprehensive search from inception to 1 November 2024, to identify RCTs comparing dexamethasone with placebo for postoperative outcomes in hysterectomy patients. Searches were performed in bibliographic databases, including MEDLINE (via PubMed) and The Cochrane Central Register of Controlled Trials (CENTRAL). The search strategy incorporated medical subject headings (MeSH) and keywords, including (“Dexamethasone”) AND (“abdominal hysterectomy” OR laparoscopic OR vaginal OR total OR “uterine surgery”) OR (“gynecologic surgery”) AND (“postoperative outcomes”). Additionally, Scopus was used for its broad coverage of multidisciplinary literature, and Google Scholar was employed to supplement traditional database searches. Studies retrieved from these sources were carefully evaluated for methodological quality and publication credibility. An attempt to identify additional studies not found by the primary search methods was made by reviewing the reference lists from identified studies. The search strategy for each database is shown in online supplemental table s1.

Inclusion and exclusion criteria

We included RCTs that met the following criteria: patients undergoing hysterectomy (Population), dexamethasone administration (Intervention), comparison to placebo or normal saline (Comparison) and relevant postoperative outcomes such as PONV, pain scores, incidence of rescue analgesia, incidence of rescue antiemesis, postoperative opioid consumption and length of hospital stay (Outcomes). Studies involving multiple groups with different types of interventions compared with placebo were included only if a direct comparison between dexamethasone and placebo could be established.

For this review: ‘Rescue analgesia’ and ‘rescue antiemetics” were defined as any additional analgesic or antiemetic medication administered postoperatively to manage inadequate pain relief or persistent nausea/vomiting, respectively.

Studies were excluded if they were not RCTs, including patients who were undergoing other surgical procedures concomitantly, reported insufficient data or were duplicate studies. Exclusion criteria also encompassed pregnant women, individuals with insulin-dependent diabetes, obesity (typically defined as body mass index (BMI) ≥30 kg/m²) or high risk for PONV (defined as those with multiple established risk factors such as a personal or family history of PONV or motion sickness, and use of postoperative opioids).

Selection of studies

Two reviewers independently screened the titles and abstracts to select studies meeting the inclusion criteria. We removed records that were ongoing or unpublished studies or were published as abstracts or conference proceedings. Only the most recent information was included where datasets were overlapping or duplicated. We retrieved full texts for potential studies, resolving any discrepancies by consensus.

Data extraction and management

Two independent reviewers performed data extraction and extracted study population characteristics, interventions and outcomes from eligible RCTs. Discrepancies in data extraction were resolved through discussion and consensus. In cases where studies reported multiple intervention subgroups based on varying doses of dexamethasone,^{17 28 30} they were combined into a single intervention group following Cochrane Handbook guidelines.³³ This approach ensured that data from each study contributed only one comparison (intervention vs placebo), preventing duplication of the shared placebo group. Data were initially extracted from tables. For data not available in tables, attempts to contact authors were made; if the authors did not respond or did not have current contact information, the data were abstracted from available figures.^{17 29} Dichotomous data on the presence or absence of adverse effects were extracted and converted to incidence, while continuous data were recorded using mean and SD. Data presented only as median and range were converted to means and SD.³⁵ Similarly, in some studies, the Visual Analogue Scale (VAS) for pain was reported on a 100 mm scale^{17 18 29} and converted to a 10 mm scale for consistency by dividing the values by 10. In multiarm trials,³¹ only the dexamethasone and placebo arms were included to ensure direct comparison. Data (sample sizes, event counts, means and SD) were then pooled using standard formulas, preserving data independence and avoiding unit-of-analysis errors.

Risk of bias assessment

The risk of bias for all included RCTs was assessed using the Cochrane Collaboration’s Revised Risk of Bias tool (RoB 2).³⁶ In line with RoB 2 guidance, we prespecified the incidence of PONV and postoperative pain scores at multiple time intervals (4 hours, 8–12 hours and 24 hours) as primary outcomes for bias assessment. The RoB 2 tool evaluates five domains: (1) bias arising from the randomisation process, (2) bias due to deviations from intended interventions, (3) bias due to missing outcome data, (4) bias in the measurement of outcomes and (5) bias in the selection of the reported results. Risk of bias was assessed at the outcome level for each primary outcome. However, for clarity of reporting, we summarised an overall study-level risk of bias based on the highest level of risk assigned across the primary outcomes. The overall risk of bias was categorised as: Low risk if all domains were rated low for all primary outcomes; Some concerns if at least one domain raised concerns, but none were rated high risk; High risk if at least one domain was rated high risk, or if multiple domains raised concerns that collectively reduced confidence in the study’s results. Two reviewers independently assessed the risk of bias, with any discrepancies resolved through discussion and consensus.

Statistical analysis

Data were analysed using Review Manager (RevMan V.5.4.1). For dichotomous outcomes, risk ratios (RRs) with 95% CIs were calculated using the Mantel-Haenszel method. For continuous outcomes measured on the same scale (eg, pain scores, length of hospital stay), mean differences (MDs) with 95% CIs were used. Due to variations in opioid types, units and reporting, standardised MDs (SMDs) with 95% CIs were applied for postoperative opioid consumption. A random-effects model was used for data synthesis, considering differences in surgical techniques and anaesthesia protocols. Statistical significance was set at p≤0.05. Heterogeneity was assessed using Higgins’ I² statistic, with <50%, 50%–75% and >75% indicating low, moderate and high heterogeneity, respectively. To address high heterogeneity in some outcomes, leave-one-out sensitivity analyses were performed. Subgroup analyses based on dexamethasone timing, administration route, dosage and surgical approach of hysterectomy were conducted to validate findings on PONV and postoperative pain. Publication bias was evaluated via visual inspection of funnel plots. Egger’s test was run for the assessment of publication bias for the outcomes with at least 10 studies on R V.4.4.0 using the packages ‘meta’ and ‘metasens’.

Certainty of evidence assessment

We applied Grading of Recommendations, Assessment, Development and Evaluation (GRADE) principles^{37 38} to assess the confidence in our pooled effect estimates, considering the risk of bias, consistency, directness, precision and reporting bias. Two reviewers independently assessed the certainty of evidence for each pooled analysis, with discrepancies resolved by consulting a third reviewer. RCTs were initially assigned high confidence ratings but downgraded to moderate, low or very low if issues were identified in the GRADE criteria. Confidence levels were defined as: high (true effect likely close to estimate), moderate (true effect probably close but may differ), low (true effect may vary significantly) and very low (true effect likely differs substantially).

Patient and public involvement

None.

Results

Trial identification

The searches identified a total of 2769 articles. After screening titles and abstracts, we identified 17 as potentially eligible for inclusion and retrieved their full-text versions. On reviewing these, one study’s full text could not be located,³⁹ and another study was deemed ineligible because it lacked a control group and primarily focused on evaluating dose ranges.⁴⁰ Ultimately, 15 RCTs17,31 met the inclusion criteria for this review. Figure 1 shows the Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart.

Figure 1. PRISMA flow diagram. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Characteristics of included studies

Table 1 summarises the characteristics of the 15, including RCTs, detailing key aspects. The studies in this meta-analysis span from inception to 2024 and include 1362 patients. Of these, 767 were in the dexamethasone group and 595 in the placebo group, who underwent either general or regional anaesthesia for hysterectomy. The hysterectomy procedures varied by technique, comprising both open and laparoscopic approaches. Details on the demographics and baseline characteristics of patients included in the studies, covering age, BMI, outcomes reported, duration of surgery and anaesthesia, American Society of Anesthesiologists class, are provided in online supplemental table s2. Dexamethasone was administered either as a single dose or in combination, with doses ranging from 2.5 to 15 mg. The timing of administration varied, occurring either before or after anaesthesia induction.

Table 1. Characteristics of trials included in the meta-analysis.

Study ID	Sample size	Country	Type of hysterectomy	Route of administration	Interventions	Anaesthetic technique	Rescue analgesics	Rescue antiemetics
Thangaswamy 200917,	55	India	Laparoscopic	Intravenous	D 4 mg vs 8 mg vs placebo, all 2 hours before induction	Fentanyl 2.0 lg/kg, propofol 2–3 mg/kg, intravenous maintained with nitrous oxide and 1%–1.5% isoflurane in 33% oxygen	Fentanyl 20 μ/kg IV PCA	Ondansetron 0.1 mg/kg IV
Ammar and Mahmoud, 201218	60	Egypt	Open	Perineural	D 8 mg vs placebo, both after induction	Propofol 1.5–2 mg/kg and fentanyl 3 µg/kg, IV maintained with isoflurane 1MAC, cis-atracurium 2 µg/kg/min and fentanyl 1 µg/kg/hour	Morphine 1 mg IV PCA	Ondansetron 4 mg IV
Chu et al, 200819	149	Taiwan	Laparoscopic	Intravenous	D 5 mg vs placebo, both 15 min after induction	Fentanyl 2 g/kg, lidocaine 1 mg/kg, propofol 2 mg/kg, IV maintained with 8%–12% desflurane in oxygen.	Ketorolac with or without meperidine	Ondansetron 4 mg IV
Sekhavat et al, 201920	100	Iran	Open	Intravenous	D 8 mg vs placebo, both at the end of surgery	Fentanyl 2 lg/kg, propofol 2 mg/kg IV maintained with ventilation of N2O-O2 (50–50) (propofol 100–200 mg/kg/min IV; maintenance anaesthesia)	Morphine 5 mg IV	Metoclopramide 10 mg IV
Wang et al, 199921	74	Taiwan	Open	Intravenous	D 8 mg vs placebo, both at the end of surgery	Lidocaine 2% 0.3 mL/kg, lidocaine 2% (with epinephrine); epidural catheter. (midazolam 2.5 mg, IV for sedation)	Diclofenac 75 mg IM	Metoclopramide 10 mg IV
Tzeng et al, 200222	76	Taiwan	Open	Intravenous	D 5 mg vs placebo, both at the end of surgery	Lidocaine 0.3 mL/kg 2%, lidocaine 2% 3 mL (with epinephrine), epidural catheter (midazolam 2.5–5 mg IV for sedation)	Diclofenac 75 mg IM	Ondansetron 4 mg IV
Wang et al, 200223	76	Taiwan	Open	Intravenous	D 5 mg vs placebo, both before surgery	Lidocaine 2% 0.3 mL/kg lidocaine 2% 3 mL (with epinephrine); epidural catheter.	Diclofenac 75 mg IM	Droperidol 1.25 mg IV
Deshpande et al, 201724	60	India	Open	Perineural	D 4 mg vs placebo, both at the end of surgery	Bupivacaine 0.5% 17.5 mg; intrathecally	Tramadol 1 mg/kg IV	Not reported
Khatiwada et al, 201225	80	Nepal	Open	Intravenous	D 4 mg vs placebo, both 1 hour before surgery	Bupivacaine 0.5% 3.4 mL; spinal needle	Tramadol 50 mg IV	Ondansetron 4 mg IV
Leopold et al, 201126	82	Germany	Open	Oral	D 8 mg vs placebo, both 2 hours before surgery	Fentanyl 0.2–0.5 mg, propofol 2–3 mg /kg, IV maintained using 3–5% of desflurane in oxygen	Piritramide 0.05–0.01 mg/kg	Not reported
Kashanian et al, 202127	102	Iran	Open	Intravenous	D 8 mg vs placebo, both 1 hour before surgery	Dexamethasone IV	Not reported	Not reported
Jokela et al, 200928	120	Finland	Laparoscopic	Intravenous	D 5 mg vs D 10 mg vs D 15 mg vs placebo, all before induction	Not reported	Oxycodone 1 mg/mL IV	Droperidol 0.5 mg IV; second line Ondansetron 4 mg IV.
Mathiesen et al, 200929	76	Denmark	Open	Placebo: Oral Dexa: Intravenous	D 8 mg vs placebo, both before induction	Propofol (variable rate) and remifentanil 0.5 mg/kg/min).	Morphine 2.5 mg	Ondansetron 4 mg IV
Ho et al, 200130	172	Taiwan	Open	Intravenous	D 10 mg vs D 5 mg vs D 2.5 mg vs placebo, all after surgery	Lidocaine 2% 0.3 mL/kg (with 1:100 000 epinephrine), lidocaine 3 mL 2% (with epinephrine). (Midazolam 2.5±5 mg IV; maintenance anaesthesia)	Diclofenac 75 mg IM	Ondansetron 4 mg IV
Wang et al, 200031	120	Taiwan	Open	Intravenous	D 10 mg vs placebo, both before anaesthesia induction D 10 mg vs placebo, both after anaesthesia induction	Anaesthesia was induced with 2–2.5 mg/kg IV propofol, 0.2 mg of IV glycopyrrolate, and 2 mg/kg IV fentanyl, 0.15 mg/kg IV vecuronium. Anaesthesia was maintained with 1.0% to 2.5% oxygen isoflurane, 0.6 mg of IV glycopyrrolate and 3 mg of IV neostigmine for tracheal extubation.	Morphine 2 mg IV and then PCA pump delivered 1 mg IV on demand	Droperidol IV 1.25 mg

D, dexamethasone ; IM, intramuscular; IV, intravenous; PCA, principal component analysis.

Risk of bias in included studies

Risk of bias was assessed at the outcome level using the RoB 2 tool, and for clarity of presentation, we summarised the risk of bias at the study level, based on the highest level of risk observed across primary outcomes. Most studies were rated as low risk of bias. One study was judged to have high risk due to missing outcome data (domain 3),²⁷ and three studies had some concerns related to the randomisation process (domain 1).^{21 23 25} The findings were visually presented using traffic lights and summary plots, offering a clear and concise overview of potential biases across the studies (figure 2).

Figure 2. Summary of risk of bias (green: low; yellow: unclear; red: high). Summary of risk of bias items presented as the individualised risk of bias categories for each included study (A) and categories presented as percentages across all included studies (B).

Results of meta-analysis

Figures35 visually represent the pooled results of the incidence of PONV and postoperative pain scores as forest plots.

Figure 3. Forest plots comparing primary outcomes for dexamethasone versus placebo in patients following hysterectomy. Forest plot of incidence of postoperative nausea and vomiting (A), Forest plot of postoperative pain score (B).

Figure 5. Subgroup analysis forest plots comparing primary outcomes for dexamethasone versus placebo in patients following hysterectomy for postoperative nausea and vomiting. Subgroup analysis forest plot of incidence of postoperative nausea and vomiting based on dose of dexamethasone (A). Subgroup analysis forest plot of incidence of postoperative nausea and vomiting based on timings of dexamethasone administration (B). Subgroup analysis forest plot of incidence of postoperative nausea and vomiting based on route of dexamethasone administration (C). Subgroup analysis forest plot of incidence of postoperative nausea and vomiting based on type of hysterectomy procedure (D).

Primary outcomes

Incidence of PONV

For the outcome of PONV, data from 15 studies were analysed, with a total of 1362 patients (767 in the dexamethasone group vs 595 in the placebo group).17,31 The meta-analysis demonstrated a significant decrease in the incidence of PONV in patients receiving dexamethasone compared with placebo (RR 0.53, 95% CI 0.47 to 0.61, p<0.00001; figure 3A). There was no evidence of significant heterogeneity between studies (p=0.89, I²=0%) and certainty of evidence was high.

Postoperative pain score

Postoperative pain scores at 24 hours, 8–12 hours and 4 hours were analysed and presented in figure 3B.

24-hour postoperative pain score

For the outcome of postoperative pain data from eight studies were analysed, with 678 patients (391 in the dexamethasone group vs 287 in the placebo group).1718 22,24 27 29 30 The meta-analysis demonstrated a significant decrease in postoperative pain scores in patients receiving dexamethasone compared with placebo (MD: −0.20, 95% CI −0.35 to −0.05, p=0.009). There was no evidence of significant heterogeneity between studies (p=0.98, I²=0%) and certainty of evidence was moderate.

8−12-hour postoperative pain score

For the outcome of postoperative pain, data from seven studies were analysed, with 572 patients (354 in the dexamethasone group vs 218 in the placebo group).1718 22,24 27 30 The meta-analysis demonstrated a significant decrease in postoperative pain scores in patients receiving dexamethasone compared with placebo (MD: −0.60, 95% CI −0.88 to −0.31, p<0.0001). There was evidence of significant heterogeneity between studies (p=0.02, I² = 27%) and certainty of evidence was moderate.

4-hour postoperative pain score

For the outcome of postoperative pain data from 8 studies were analysed, with 682 patients (393 in the dexamethasone group vs 289 in the placebo group).1718 22,24 27 29 30 The meta-analysis demonstrated a significant decrease in postoperative pain scores in patients receiving dexamethasone compared with placebo (MD: −0.43, 95% CI −1.07 to 0.21, p=0.19). There was evidence of significant heterogeneity between studies (p<0.00001, I²=91%) and certainty of evidence was moderate.

Secondary outcomes

Incidence of rescue analgesics

For the outcome of rescue analgesia, data from 7 studies were analysed, with a total of 747 patients (448 in the dexamethasone group vs 299 in the placebo group).1921,23 25 28 30 The meta-analysis showed no significant difference in the incidence of rescue analgesia between patients receiving dexamethasone and those receiving a placebo (RR: 0.89, 95% CI 0.77 to 1.03, p=0.12; I²=11%), (figure 4A). The certainty of evidence was high.

Figure 4. Forest plots comparing secondary outcomes for dexamethasone versus placebo in patients following hysterectomy. Forest plot of incidence of rescue analgesics (A), forest plot of incidence of rescue antiemetics (B), forest plot of postoperative opioid consumption (C), forest plot of length of hospital stays (D).

Incidence of rescue antiemetics

For the outcome of rescue antiemetics, data from 9 studies were analysed, with a total of 883 patients (485 in the dexamethasone group vs 398 in the placebo group).19,2325 29 The meta-analysis demonstrated a significant decrease in the incidence of rescue antiemetics in patients receiving dexamethasone compared with placebo (RR: 0.57, 95% CI 0.43 to 0.75, p<0.0001; I²=39%) (figure 4B). The certainty of evidence was high.

Postoperative opioid consumption

For the outcome of postoperative opioid consumption, data from 5 studies were analysed, with a total of 391 patients (234 in the dexamethasone group vs 157 in the placebo group).^{17 18 28 29 31} The meta-analysis demonstrated a significant decrease in postoperative opioid consumption in patients receiving dexamethasone compared with placebo (SMD: −0.48, 95% CI −0.90 to −0.05, p=0.03, I²=74%) (figure 4C). The certainty of evidence was low.

Length of hospital stay

For the outcome of the length of hospital stay, data from 4 studies were analysed, with a total of 405 patients (255 in the dexamethasone group vs 150 in the placebo group).^{17 22 27 30} The meta-analysis demonstrated no significant difference in the length of hospital stay in patients receiving dexamethasone compared with placebo (MD: −0.13, 95% CI −0.35 to 0.08, p=0.11, I²=78%) (figure 4D). The certainty of evidence was moderate.

Sensitivity analyses

We conducted a sensitivity analysis for pain scores at all time points by excluding studies that used the perineural route of dexamethasone administration instead of the intravenous route.^{18 24} This exclusion reduced heterogeneity to 0% for the 8–12-hour and 4-hour time points. Notably, the 4-hour pain score results remained non-significant (figure 1A). Additionally, a separate sensitivity analysis was performed for all pain score points by excluding studies with a high or unclear risk of bias.^{23 27} At the 4-hour time point, this exclusion did not meaningfully reduce heterogeneity, and the statistical significance of the results for all time points showed only a slight change (figure 1B).

For postoperative opioid consumption, excluding the study by Ammar and Mahmoud¹⁸ significantly reduced heterogeneity (I²=0%) without affecting the statistical significance (online supplemental figure S1C). Similarly, removing the study by Ho et al³⁰ for the length of hospital stay reduced heterogeneity (I²=0%) without altering the nonsignificant results (online supplemental figure S1D).

Subgroup analyses

For the incidence of PONV based on the following

Dose of dexamethasone

The subgroup analysis of dexamethasone on the basis of doses for incidence of PONV revealed a highly significant overall effect for both the 4–5 mg (RR 0.52, 95% CI 0.43 to 0.64, p<0.00001, I²: 0%) and 8–10 mg doses (RR 0.52, 95% CI 0.44 to 0.62, p<0.00001, I²: 0%). In contrast, the 15 and 2.5 mg doses showed non-significant results (p=0.29 and p=0.19, respectively). While formal subgroup comparison did not indicate statistically significant differences between dose groups, the evidence suggests that doses in the 4–10 mg range consistently provide the most reliable antiemetic benefit (figure 5A).

Timing of dexamethasone administration

The RCTs in this review varied in the timing of dexamethasone administration, with some patients receiving it before anaesthesia induction and others afterward. In both scenarios, dexamethasone significantly reduced the incidence of PONV whether administered before induction (RR: 0.55, 95% CI: 0.47 to 0.65, p<0.00001, I²: 0%) or after (RR: 0.51, 95% CI: 0.41 to 0.64, p<0.00001, I²: 20%), The test for subgroup differences was non-significant, indicating that dexamethasone is similarly effective in reducing PONV regardless of the timings of administration (figure 5B).

Route of dexamethasone administration

The subgroup analysis for PONV based on intravenous, perineural and oral route of administration of dexamethasone showed highly significant overall results for all subgroups (RR: 0.53, 95% CI: 0.47 to 0.61, p<0.00001, I²=0%), The absence of significant subgroup differences suggests that dexamethasone provides consistent antiemetic effectiveness regardless of the route of administration (figure 5C).

Type of procedure

The subgroup analysis for PONV, based on surgical approach (open vs laparoscopic hysterectomy), demonstrated highly significant results within both subgroups (RR: 0.30, 95% CI: 0.24 to 0.38, p<0.00001, I²=0%). The absence of significant subgroup differences suggests that dexamethasone provides consistent antiemetic effectiveness regardless of the surgical technique type (figure 5D).

For postoperative pain score based on the following

Dose of dexamethasone

The subgroup analysis of dexamethasone based on doses for postoperative pain score revealed a highly significant overall effect for the 8–10 mg dose (RR −0.33, 95% CI −0.54 to −0.11, p=0.004, I² = 0%). The lower doses, 2.5 mg and 4–5 mg, have no significant effect on pain control (figure 6A). While formal subgroup comparison did not indicate statistically significant differences between dose groups, the evidence suggests that doses in the 8–10 mg range consistently provide the most reliable analgesic benefit.

Figure 6. Subgroup analysis forest plots comparing primary outcomes for dexamethasone versus placebo in patients following hysterectomy for pain scores. Subgroup analysis forest plot of pain scores based on dose of dexamethasone (A). Subgroup analysis forest plot of pain scores based on timings of dexamethasone administration (B). Subgroup analysis forest plot of pain scores based on route of dexamethasone administration: for 24-hour pain score (C), for 8–12-hour pain score (D), for 4-hour pain score (E). Subgroup analysis forest plot of pain scores based on type of hysterectomy procedure: For 24-hour pain score (F), for 8–12-hour pain score, (G) for 4-hour pain score (6 hours).

Timing of dexamethasone administration

The subgroup analysis of dexamethasone on the basis of timings of administration for postoperative pain score revealed a highly significant overall effect whether administered before induction (RR: −0.24, 95% CI −0.45 to −0.02, p=0.03, I²=0%) or after (RR: −0.23, 95% CI −0.40 to −0.05, p=0.01, I²=0%), suggesting that both approaches offer a consistent analgesic effectiveness regardless of timing and providing clinicians flexibility in timing of administration. The test for subgroup differences was non-significant, indicating that dexamethasone is similarly effective in reducing pain scores regardless of the timings of administration (figure 6B).

Route of dexamethasone administration

At 24 hours postoperatively, pain reduction was significant in the intravenous group (MD –0.20, 95% CI –0.38 to –0.03, p=0.02), but not in the perineural group (MD –0.18, 95% CI –0.46 to 0.09, p=0.19). The overall effect favoured dexamethasone (MD –0.20, 95% CI –0.35 to –0.05, p=0.009), with no significant subgroup difference, suggesting consistent efficacy across both routes. Heterogeneity was low (I²=0%) for all subgroups (figure 6C).

At 8–12 hours, both routes showed significant pain reduction more pronounced with perineural administration (MD –0.94, 95% CI –1.23 to –0.66, p<0.00001) compared with intravenous (MD –0.33, 95% CI –0.66 to –0.00, p=0.05). A significant subgroup difference (p=0.006) indicates a potentially stronger effect with the perineural route. Heterogeneity was low (I²=0% for all subgroups) (figure 6D).

At 4 hours, only the perineural group showed significant pain score reduction (MD –1.43, 95% CI –2.40 to –0.47, p=0.004, I²=86%), while the intravenous group did not (MD –0.09, 95% CI –0.29 to 0.12, p=0.41, I²=0%). The overall effect was non-significant (p=0.19), but subgroup differences were significant (p=0.008), highlighting a possible route-specific variation in early analgesic efficacy (figure 6E).

Type of procedure

At 24 hours postoperatively, pain reduction was significant in the open hysterectomy group (MD −0.20, 95% CI −0.35 to −0.05, p=0.009), but not in the laparoscopic hysterectomy group (MD −0.06, 95% CI −1.52 to 1.40, p=0.94). The overall effect favoured dexamethasone efficacy for both procedures (MD −0.20, 95% CI −0.35 to −0.05, p=0.009), with no significant subgroup difference (p=0.85). Heterogeneity was low (I²=0% for all subgroups) (figure 6F).

At 8–12 hours, both routes showed significant pain reduction, more pronounced for the open hysterectomy group (MD −0.61, 95% CI −0.91 to −0.31, p<0.0001) compared with the laparoscopic hysterectomy group (MD 0.09, 95% CI −1.64 to 1.83, p=0.92). The overall effect favoured dexamethasone efficacy for both procedures (MD −0.60, 95% CI −0.88 to −0.31, p<0.0001), with no significant subgroup difference (p=0.43). Heterogeneity was low for all subgroups (figure 6G).

At 4 hours postoperatively, neither the open hysterectomy group (MD –0.48, 95% CI –1.14 to 0.19, p=0.16, I²=92%) nor the laparoscopic group (MD 0.14, 95% CI –1.62 to 1.91, p=0.87, I²=91%) showed a statistically significant reduction in pain scores. Both the overall effect (p=0.19) and the test for subgroup differences (p=0.52) were non-significant (figure 5H).

Publication bias

We assessed publication bias for outcomes reported in ten or more studies by visually inspecting their funnel plots.⁴¹ The funnel plot for the incidence of PONV showed no asymmetry (online supplemental figure S2A), which was further supported by Egger’s test indicating no small-study effects (p=0.4140). The funnel plot for all time points of the pain score shows asymmetry, indicating potential publication bias (online supplemental figure S2B). The funnel plots for the secondary outcomes are represented in online supplemental figures S2C-F.

Results of certainty of evidence assessment

Table 2 summarises the main findings, and table 3 presents the quality of available evidence from RCTs. These findings indicate that dexamethasone is particularly effective in reducing PONV, rescue antiemetic use and postoperative pain, with moderate to high confidence in the evidence for most outcomes. However, its impact on opioid consumption and hospital stay remains less certain due to heterogeneity and imprecision.

Table 2. Summary of findings (GRADE Evidence).

Outcomes	Mean (SD) or n/N in dexamethasone group	Mean (SD) or n/N in placebo group	SMD, mean difference or RR (95% CI)	Number of participants (studies)	Quality or certainty of the evidence (GRADE)	Comments
Incidence of postoperative nausea and vomiting	212/767	327/595	RR 0.53 (0.47 to 0.61)	1362 (15 studies)	⨁⨁⨁⨁High*†‡§	No risk of bias, low heterogeneity and no evidence of downgrading detected at any stage. Thus, this outcome is rated as high certainty.
24-hour postoperative pain score scale from 0 to 10	1.95 (1.44)	2.13 (1.77)	SMD −0.21 (−0.37 to −0.06)	678 (8 RCTs)	⨁⨁⨁◯Moderate*†‡¶	Downgraded for imprecision. Thus, this outcome is rated as moderate certainty.
8−12-hour postoperative pain score scale from 0 to 10	2.16 (1.55)	2.51 (1.77)	SMD −0.39 (−0.74 to −0.04	572 (7 RCTs)	⨁⨁⨁◯Moderate‡¶**††	Downgraded for imprecision but no other evidence of downgrading. Hence rated as moderate.
4-hour postoperative pain score scale from 0 to 10	2.42 (2.02)	2.69 (1.95)	SMD −0.66 (−0.28 to −0.05	682 (8 RCTs)	⨁⨁⨁◯Moderate‡**‡‡§§	Downgrading for inconsistency; Rated as moderate-quality evidence.
Incidence of rescue analgesia	199/448	166/299	RR 0.89 (0.77 to 1.03)	747 (7 RCTs)	⨁⨁⨁◯High†‡§¶¶	No evidence of downgrading; hence rated as high.
Incidence of rescue antiemetics	101/485	157/398	RR 0.57 (0.43 to 0.75)	883 (9 RCTs)	⨁⨁⨁⨁High*†‡***	Rated high-quality evidence due to non-detection of downgrading.
Postoperative opioid consumption	7.17 (16.58)	14.14 (20.26)	SMD −0.48 (−0.90 to −0.05)	391 (5 RCTs)	⨁⨁◯◯Low‡¶‡‡	Downgraded for inconsistency and imprecision; Hence rated as low quality evidence.
Length of hospital stay	5.9 (3.42)	5.42 (3.53)	SMD −0.27 (−0.60 to −0.06)	405 (4 RCTs)	⨁⨁⨁◯Moderate‡¶**	Downgraded for imprecision; hence rated as moderate quality evidence.

High certainty: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.

^*Some concern with allocation concealment and outcome measurement, but not rated down for risk of bias.

^†Statistical analysis showed no substantial heterogeneity resulting in narrow CIs.

^‡No indirectness was detected as the studies directly answered our question.

^§Narrow CIs with no significant heterogeneity.

^¶Downgraded for imprecision because the effect estimate did not cross the threshold for minimal important difference (MID) of 0.5 points lower.

^**Some concern with allocation concealment but not rated down for risk of bias.

^††Substantial heterogeneity resulting in wide CIs but not downgraded for inconsistency.

^‡‡Very substantial heterogeneity resulting in wide CIs hence downgraded for inconsistency.

^§§Not downgraded for imprecision because the effect estimate did cross the threshold for MID of 0.5 points lower.

^¶¶Some concern with allocation concealment and weight of studies, but not rated down for risk of bias.

^***Narrow CIs with moderate heterogeneity; hence not downgraded for imprecision.

GRADE, Grading of Recommendations Assessment, Development and Evaluation; n, is the number of participants with the event and N is the total number in the group; NPRS, Numeric Pain Rating Scale; RCTs, randomised controlled trials; RR, risk ratio; SMD, standardised mean difference.

Table 3. Evidence profile.

Outcomes	Limitations	Inconsistency	Indirectness	Imprecision	Publication bias	SMD or RR (95% CI)	Number of participants (studies)	Quality or certainty of the evidence (GRADE)
Incidence of postoperative nausea and vomiting	Low risk of bias; No serious limitations	Low test for heterogeneity (I2=0%); not downgraded	No indirectness detected	Not detected	Not detected	RR 0.53 (0.47 to 0.61)	1362 (15 RCTs)	⨁⨁⨁⨁ High
24-hour postoperative pain score scale from 0 to 10	Low risk of bias; No serious limitations	Low test for heterogeneity (I2=0%); not downgraded	No indirectness detected	Imprecision detected	Not detected	SMD −0.21 (-0.37 to -0.06)	678 (8 RCTs)	⨁⨁⨁◯Moderate
8−12-hour postoperative pain score scale from 0 to 10	Low risk of bias; No serious limitations	Substantial heterogeneity (I2=72%); but not downgraded for inconsistency	No indirectness detected	Imprecision detected	Not detected	SMD −0.39 (-0.74 to -0.04	572 (7 RCTs)	⨁⨁⨁◯Moderate
4-hour postoperative pain score scale from 0 to 10	Low risk of bias; No serious limitations	Very substantial heterogeneity (I2=93%); downgraded	No indirectness detected	Not detected	Not detected	SMD −0.66 (−0.28 to −0.05	682 (8 RCTs)	⨁⨁⨁◯Moderate
Incidence of rescue analgesia	Low risk of bias; No serious limitations	Low test for heterogeneity (I2=11%); Not downgraded	No indirectness detected	Not detected	Not detected	RR 0.89 (0.77 to 1.03)	747 (7 RCTs)	⨁⨁⨁⨁ High
Incidence of rescue antiemetics	Low risk of bias; No serious limitations	Moderate heterogeneity (I2=39%); Not downgraded	No indirectness detected	Not detected	Not detected	RR 0.57 (0.43 to 0.75)	883 (9 RCTs)	⨁⨁⨁⨁ High
Postoperative opioid consumption	Low risk of bias; No serious limitations	High heterogeneity detected (I2=74%); downgraded	No indirectness detected	Imprecision detected	Not detected	SMD −0.48 (−0.90 to −0.05)	391 (5 RCTs)	⨁⨁◯◯Low
Length of hospital stay	Low risk of bias; No serious limitations	Moderate heterogeneity (I2=58%); not downgraded	No indirectness detected	Imprecision detected	Not detected	SMD −0.27 (−0.60 to −0.06)	405 (4 RCTs)	⨁⨁⨁◯Moderate

GRADE, Grading of Recommendations Assessment, Development and Evaluation; RCTs, randomised controlled trials; RR, risk ratio; SMD, standardised mean difference.

Discussion

Our meta-analysis of 15 RCTs demonstrates that dexamethasone significantly reduces the incidence of PONV and postoperative pain score at 24 hours and 8–12 hours in patients undergoing hysterectomy. While dexamethasone significantly reduced the need for rescue antiemetics and demonstrated a statistically significant reduction in postoperative opioid consumption, it did not significantly affect the overall need for rescue analgesia, or the length of hospital stay. This distinction underscores the complexity of assessing its analgesic efficacy, where certain pain management measures showed benefit while others did not. Subgroup analyses showed that dexamethasone significantly reduced the incidence of PONV across various doses (4–10 mg), timings (before or after induction), routes (intravenous, perineural, oral) and surgical approaches (open or laparoscopic), with no meaningful subgroup differences, indicating consistent antiemetic efficacy. For postoperative pain, dexamethasone showed a significant reduction primarily at 24 and 8–12 hours, especially with higher doses (8–10 mg) and perineural administration. At 4 hours, analgesic effects were only significant in the perineural group. Subgroup differences were observed at 4 and 8–12 hours by route, suggesting possible time-dependent variation in analgesic efficacy.

Dexamethasone is a well-established prophylactic antiemetic in children and adults receiving general anaesthesia.^{42 43} Many previous studies support our study’s findings about the effectiveness of dexamethasone in lowering PONV. A meta-analysis on the impact of dexamethasone on antiemesis in gynaecologic surgeries yielded similar results.⁴⁴ Xu et al also reported its effectiveness in preventing PONV.⁵ A 2012 study provided strong evidence that a single IV dose of 5–10 mg dexamethasone is effective for reducing PONV in women receiving neuraxial morphine for caesarean delivery or abdominal hysterectomy.⁴⁵ The findings of the previous literature report favourable outcomes of dexamethasone in reducing postoperative pain. It is indicated that dexamethasone significantly reduces postoperative pain scores for the first 24 hours after surgery, enhancing patient comfort.⁴⁶ Patients treated with dexamethasone required less opioid use and experienced longer times to their first dose of analgesia, indicating its potential to reduce opioid-related adverse effects and support enhanced recovery. However, this finding is based on low-certainty evidence due to heterogeneity and imprecision and should be interpreted with caution. Additionally, dexamethasone led to shorter stays in the post-anaesthesia care unit, without increasing the risk of infection or delayed wound healing,^{47 48} a finding that may have practical implications for resource use.

Interestingly, while dexamethasone effectively reduced postoperative pain scores, it did not significantly decrease the need for rescue analgesia. This discrepancy can partly be explained by the results of 4-hour pain scores, which are non-significant. This aligns with the hypothesis that dexamethasone’s anti-inflammatory properties alleviate some pain, but not enough to eliminate the need for additional analgesics, especially for acute pain requiring immediate relief. For example, Waldron et al⁴⁹ noted that although dexamethasone lowered pain scores, it had a limited impact on reducing supplementary analgesic requirements.⁴⁹ Similarly, Mitchell et al found that although dexamethasone reduced pain-related inflammation and decreased pain scores, its effects on demand for additional analgesics were inconsistent.¹⁰ This supports the idea that dexamethasone may be effective in reducing general pain levels but is less effective in managing acute pain episodes that necessitate rescue analgesia.

The subgroup analysis indicates that dexamethasone doses between 4 and 10 mg are highly effective in enhancing postoperative recovery, whereas doses of 2.5 mg and 15 mg did not yield significant results. This suggests an optimal dose range of 4–10 mg for achieving antiemetic efficacy. The lack of added benefit with the 15 mg dose may reflect a dose-response plateau, where higher doses offer no additional advantage and may introduce adverse effects or receptor saturation, negating potential efficacy gains. These findings are consistent with studies such as De Oliveira et al,⁵⁰ which noted a dose-dependent effect of dexamethasone on PONV that plateaued around 8 mg.⁵⁰ Additionally, an RCT by Toroghi et al supports that higher doses (8 mg three times a day) do not enhance antiemetic efficacy and may increase adverse effects without added benefits.⁵¹ Similarly, another study by Coloma et al⁵² demonstrated that a single dose of dexamethasone (4 mg intravenous) decreased the time to ‘home readiness’ without increasing the incidence of postoperative wound complications in an outpatient population undergoing anorectal surgery.⁵² For analgesic effects, dexamethasone demonstrates a dose-dependent relationship. Quan et al⁵³ showed that patients receiving 0.2 mg/kg had significantly lower VAS scores 2 hours postoperation compared with those on 0.1–0.15 mg/kg, supporting 4–10 mg as the optimal analgesic dose range.⁵³ These findings support prioritising doses between 4 and 10 mg to maximise therapeutic benefit.

Our analysis further shows that dexamethasone is effective and safe for enhanced postoperative recovery regardless of whether it is administered before or after anaesthesia induction, indicating flexibility in timing for clinical practice. However, these findings differ from Wang et al,³¹ who observed that administering dexamethasone immediately before anaesthesia induction provided a more robust antiemetic effect compared with administration at the end of anaesthesia.³¹ A recent study by Yurkonis et al⁵⁴ also suggests standardising dexamethasone dosing at 8–10 mg and timing it at anaesthesia induction or before the first incision to reduce care variability, enhance patient satisfaction and simplify clinical decision-making and answer the research questions regarding timings and dosage.⁵⁴

Furthermore, this meta-analysis and subgroup analyses revealed that for PONV, results remained consistent, but for the analgesic efficacy of dexamethasone, it varied according to both the route of administration and the surgical approach for hysterectomy. At 24 hours postoperatively, intravenous dexamethasone significantly reduced pain scores, while perineural administration did not yield a statistically significant effect, although the overall subgroup difference was non-significant. This suggests a consistent late analgesic benefit across routes, aligning with prior studies that have demonstrated that the two modalities show equivalent effect as adjuvants.⁵⁵ In contrast, at earlier points (4–2 hours), perineural administration provided more pronounced analgesia, particularly at 4 hours, where the intravenous route showed no significant benefit. This temporal pattern supports the hypothesis that perineural dexamethasone, when used as an adjuvant to regional blocks, may exert a local anti-nociceptive effect that enhances early analgesia.^{56 57} Regarding surgical technique, dexamethasone was associated with significant pain reduction at 24 hours in open but not laparoscopic hysterectomy. The absence of subgroup differences implies that dexamethasone remains effective regardless of surgical invasiveness, although the benefits appeared more pronounced in open procedures at 8–12 hours. This observation may be due to higher baseline pain intensity in open surgery, offering a greater window for analgesic effect.⁵⁸ The lack of significant pain reduction at 4 hours may reflect the delayed onset of dexamethasone’s genomic effects, which typically require 1–2 hours to manifest. Prior studies suggest administering dexamethasone at least 60 min before incision enhances its analgesic efficacy by aligning peak activity with the surgical inflammatory response.⁵⁹

This review adds value to the existing body of evidence by focusing specifically on hysterectomy patients, a population often grouped with other surgical types in previous analyses. Unlike broader meta-analyses published in 2012 and 2013,^{44 49} our study stratifies postoperative pain scores by specific time intervals, incorporates the latest trial data, and evaluates both dose-response and timing effects of dexamethasone, as well as surgical approach and route of administration, explored for optimum function of dexamethasone. Furthermore, we applied the ROB two tool and GRADE framework to ensure methodological rigour. Lastly, although the most recent included trial was published in 2021, the research question remains clinically relevant due to the continued use of dexamethasone and the absence of newer high-quality studies. This gap underscores the value of our meta-analysis in offering an up-to-date synthesis of the best available evidence, supporting current hysterectomy protocols and identifying key areas for future research.

Our meta-analysis has several limitations, most of which are addressed through extensive subgroup analysis and sensitivity analysis. Significant variability existed across the included studies in terms of patient populations, surgical techniques (eg, open vs laparoscopic hysterectomy), and anaesthesia protocols. Additionally, many of the studies were conducted in low-income countries, where cultural differences and patient expectations may influence both the perception and reporting of PONV and pain. For instance, baseline PONV incidence in control groups varied widely, from 18% in some settings²⁰ to 27% and 65% in others,^{19 25} highlighting the influence of cultural and contextual factors that may limit the generalisability of our findings to higher-income settings or populations with lower baseline PONV risk. Protocols for dexamethasone administration varied across studies, with differences in dosage (ranging from 2.5 mg to 15 mg), timing (before vs after anaesthesia induction), and routes of administration. While most studies used the intravenous route, two studies employed perineural administration,^{18 24} and one used the oral route.²⁶ Furthermore, the included studies primarily focused on healthy patients undergoing hysterectomy, excluding high-risk groups such as pregnant women, obese individuals and those with diabetes or a high risk of PONV, leaving the efficacy and safety of dexamethasone in these populations unclear. While dexamethasone was generally well-tolerated, the included studies did not consistently report on its potential side effects, such as hyperglycaemia, delayed wound healing and psychological effects (eg, mood changes), underscoring the need for more comprehensive safety evaluations. Most studies concentrated on short-term postoperative outcomes, such as pain scores and PONV incidence, without exploring long-term effects, such as wound healing complications or persistent pain, limiting the ability to assess the complete safety of dexamethasone over extended periods. Additionally, there is a lack of studies evaluating dexamethasone’s efficacy as part of combination therapies with other antiemetics (eg, ondansetron, granisetron) or its use in neuraxial anaesthesia, and questions remain about the optimal timing of dexamethasone administration (eg, before vs after anaesthesia) to maximise its benefits. These gaps highlight the need for future high-quality, well-designed studies to better understand the efficacy, safety and optimal use of dexamethasone in diverse patient populations and clinical settings. Lastly, despite employing a robust methodology, we could not perform Egger’s or Begg’s test for the assessment of publication bias for certain outcomes due to the limited number of studies. However, we have still visually represented the small study effects by reporting the funnel plots, in accordance with the Cochrane guidelines.

Conclusions

This meta-analysis provides moderate to high-quality evidence that dexamethasone is an effective adjunct for improving postoperative recovery in hysterectomy patients, significantly reducing PONV and pain at key time points, especially 8–12 and 24 hours postsurgery. Its benefits are consistent across doses (4–10 mg), timing, surgical approaches and administration routes, with early analgesic effects more notable after perineural use. Although dexamethasone reduces opioid consumption, its limited effect on rescue analgesia suggests it works best as part of a multimodal analgesic approach. While these results support dexamethasone’s role in perioperative care, safety data remain limited and inconsistently reported. Therefore, conclusions about safety should be cautious, and further well-designed studies with standardised adverse event reporting are needed, particularly in high-risk groups and in combination with other agents, to optimise patient outcomes.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. Extracted data are available on request to the corresponding author.