Dexmedetomidine as an adjunct to local anesthetics for impacted third molar surgery: a systematic review and meta-analysis

Abstract

Recent studies have investigated the use of dexmedetomidine (DEX) as an adjuvant to local anesthetic solutions. Building on this emerging evidence, this study aimed to assess the efficacy of adding DEX to local anesthetics in patients undergoing impacted third molar surgery. A systematic search was conducted across four electronic databases, PubMed, ScienceDirect, Web of Science, and the Cochrane Library, using a combination of Medical Subject Headings terms and free-text keywords related to impacted third molar surgery. Six studies comprising 266 patients were included in the qualitative systematic review, and only four of them, involving 200 patients, were included in the quantitative meta-analysis. The random-effects model demonstrated statistically significant reductions in anesthetic onset time (mean difference [MD] = −0.96 min; 95% confidence interval [CI]: −1.52 to −0.39; P = 0.0009), systolic blood pressure (MD = −3.58 mmHg; 95% CI: −6.58 to −0.58; P = 0.020), and pain at 6 and 12 h postoperatively (MD = −2.32, 95% CI: −2.85 to −1.78; P = 0.0001) in the local anesthesia + DEX group compared with local anesthesia alone. The findings indicated that incorporating DEX into local anesthesia for impacted third molar surgery resulted in statistically significant improvements in the onset of anesthetic effects and postoperative pain scores.

INTRODUCTION

Impacted third molars are among the most common clinical challenges in dental practice [1]. Patients undergoing surgical extraction of impacted third molars occasionally experience postoperative pain and discomfort [2,3].

Local anesthetics are fundamental in routine dental practice, providing pain control through the reversible blockade of nerve conduction [4,5]. The selection of an appropriate local anesthetic agent is primarily determined by its potency, onset of action, and duration of effect [6]. In addition to anesthetic agents, α2-adrenergic receptor agonists and vasoconstrictors such as adrenaline are commonly combined with lidocaine or other agents to enhance the depth and duration of anesthesia [7,8].

Dexmedetomidine (DEX) demonstrates an exceptional selectivity ratio of 1600:1 for α2- over α1-receptors. Its primary site of action is the locus coeruleus, where it induces sedation by promoting non-rapid eye movement sleep, thereby distinguishing it from conventional sedatives [9,10]. Through α2-adrenoceptor activation, DEX reduces sympathetic outflow, leading to decreased blood pressure (BP) and heart rate (HR) without causing significant cardiorespiratory depression. DEX has also been shown to provide neuroprotection and attenuate postoperative pain, further enhancing its clinical utility [10,11]. Currently, DEX is increasingly employed for premedication and procedural sedation, with multiple available routes of administration, including intramuscular, intravenous, oral, and mucosal delivery [12].

Recently, studies have explored the incorporation of DEX into local anesthetic solutions or its use as a substitute for adrenaline in nerve block techniques [8,13,14]. Although DEX has been extensively investigated as an adjuvant in regional anesthesia, its role in oral and maxillofacial surgery, particularly in third molar extraction, has not been systematically synthesized. Addressing this gap, this systematic review and meta-analysis aims to comprehensively evaluate the efficacy of adding DEX directly to the local anesthetic cartridge for patients undergoing impacted third molar surgery. Specifically, this review investigated whether the addition of DEX enhances intraoperative anesthetic performance, improves postoperative pain control, reduces perioperative anxiety, and lowers the need for supplementary analgesics compared with conventional anesthetic techniques without DEX supplementation.

METHODS

1. Protocol and registration

This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [15]. The methodological quality was assessed using AMSTAR 2 [16], and was registered in the Prospective Register of Systematic Reviews (CRD420251130854) (Supplementary table S1 and S2).

2. Population intervention comparison outcome framework and search strategy

Four electronic databases were used for the literature search: PubMed, ScienceDirect, Web of Science, and the Cochrane Central Register of Controlled Trials. The search strategy was based on the following research question: Among patients undergoing impacted third molar surgery (Population), does the addition of dexmedetomidine to local anesthetic agents for inferior alveolar nerve block (IANB) (Intervention), compared with plain local anesthetics or those containing adrenaline (Comparison), enhance anesthetic onset and duration, reduce postoperative analgesic consumption, and improve hemodynamic stability (Outcomes)? Medical Subject Headings terms and keywords were used, and the Supplementary table S3 provides the complete search strategy.

3. Eligibility criteria

Inclusion criteria were: (1) Adult patients (≥ 18 years) undergoing impacted third molar surgery; (2) use of local anesthesia combined with DEX (LA-DEX) compared to routine nerve block anesthesia; (3) randomized controlled trials (RCTs), non-RCTs, or observational studies; and (4) studies reporting perioperative outcomes.

Exclusion criteria were: (1) Case reports, case series, or review articles; and (2) articles evaluating procedures other than impacted third molar extraction.

4. Study selection and data extraction

Records retrieved from the initial search were systematically imported into a structured Excel spreadsheet designed to facilitate data management. The spreadsheet included key bibliographic and study-related information, such as the study identification number, authors’ names, year of publication, title, abstract, digital object identifier, and uniform resource locator. Each record in the title, abstract, and full text was independently screened for eligibility by two expert reviewers (M.A. and E.A.) in accordance with the predefined inclusion and exclusion criteria. In instances where uncertainty or disagreement arose regarding the inclusion of a study, the issue was addressed through discussion until a consensus was reached, thereby ensuring the reliability and rigor of the screening process. Furthermore, M.A. and E.A. independently performed data extraction and resolved discrepancies by consensus, ensuring reproducibility and minimizing extraction bias.

5. Risk of bias assessment

The risk of bias (RoB) was independently assessed by two reviewers (M.A. and E.A.). RCTs were evaluated using the Cochrane RoB-2 tool [17], which examines five domains: randomization, deviations from intended interventions, missing outcome data, outcome measurements, and selective reporting. Non-randomized studies were assessed using the risk of bias in non-randomized studies - of interventions (ROBINS-I) tool [18], covering seven domains, including confounding, participant selection, intervention classification, and missing data. Each domain was assigned a risk rating (low, some concerns, or high for RoB 2; and low, moderate, serious, or critical for ROBINS-I), leading to an overall judgment per study. Disagreements were resolved through discussions to reach a consensus, and a visual representation of the RoB results was conducted using the Robvis tool [19].

6. Certainty of evidence

The quality of evidence for each primary outcome was appraised using the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework [20]. This framework rates the certainty of evidence as very low, low, moderate, or high based on five key factors: RoB, inconsistency, indirectness, imprecision, and publication bias. A Summary of Findings table was generated using GRADEpro GDT software (McMaster University and Evidence Prime, 2025; gradepro.org) to provide a transparent overview of the evidence quality across outcomes.

7. Outcome measures

The primary outcome measure of this study included the onset of anesthesia, defined as the time interval between the administration of the local anesthetic and the establishment of a complete anesthetic effect. This was measured using both subjective signs, such as the patient’s reported loss of sensation, numbness, or tingling in the anesthetized region, and objective signs, including lack of response to gentle probing, absence of pain on needle prick, and absence of discomfort during the initiation of the dental procedure. The onset time was recorded in minutes from the completion of injection until the first confirmation of effective anesthesia, as determined by these combined criteria. Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were recorded at two time points, preoperatively (baseline, before local anesthetic administration) and postoperatively (following completion of the procedure). Changes in SBP and DBP were examined to determine the mean differences (MDs) between the groups. Pain intensity was measured using a visual analogue scale (VAS) at 6 and 12 h, which is a 10-point scale (0 = no pain, 10 = worst pain). The minimal clinically important difference (MCID) for the VAS is generally considered to be a reduction of approximately 10–20 mm (1–2 points on a 10-point scale). This threshold is supported by previous studies on clinical pain measurements [21]. HR, oxygen saturation (SpO₂), duration of postoperative analgesia, and the number of analgesics consumed postoperatively were the secondary outcomes, measured as the change in mean from baseline.

8. Data analysis

The characteristics of the included studies were systematically tabulated and summarized using descriptive statistics. Statistical analysis was performed using the “Meta” package in R (version 4.4.3; RStudio, Vienna, Austria). The effect estimates for all outcomes were calculated using a random-effects model. Continuous variables were reported as MDs with 95% confidence intervals (CIs). Statistical heterogeneity was categorized as follows: mild (I² < 50%), moderate (I² > 50% < 75%), and substantial (I² > 75% < 100%). Statistical significance was set at P < 0.05. Subgroup analysis included the type of local anesthetic agent used. Sensitivity analysis was also conducted to explore heterogeneity by systematically excluding one study. When outcome data were presented only in graphical form, values were extracted using WebPlotDigitizer software version 4.5 (https://wpd.starrydata2.org/), which enables accurate digitization of data points from published figures. For studies reporting continuous outcomes as medians with interquartile ranges, appropriate statistical conversions were applied to estimate the mean and standard deviation, ensuring comparability across trials. These conversions were performed using the MetaAccelerator tool [22].

Due to the small number of included studies (< 10 per outcome), publication-bias analysis was not feasible, in accordance with Cochrane guidelines.

RESULTS

1. Selection of studies

The electronic search strategy initially yielded 237 records. After removing duplicate entries, 217 unique studies remained and were screened by title and abstract. At this stage, irrelevant studies were excluded, and potentially eligible articles were retained for full-text review. Eight full-text records were assessed against the predefined inclusion and exclusion criteria to determine their eligibility for inclusion in the review. Six studies [23,24,25,26,27,28], comprising 266 patients, were included in the qualitative systematic review, and only four of these studies, involving 200 patients, were included in the quantitative meta-analysis (Fig. 1.

Fig. 1. Study selection flowchart.

2. Characteristics of the studies

The participants in the included studies ranged in age from 18 to 48 years. The sex distribution was balanced, with 133 males and 133 females (50% each). Five of the studies were randomized clinical trials [24,25,26,27,28], while one employed a non-randomized comparative study design [23]. Geographically, two studies were conducted in Iran, and the remaining four originated from India. All surgical interventions involved the extraction of impacted third molars.

Baseline health status was documented in four trials, with participants classified as American Society of Anesthesiologists (ASA) physical status I or II, indicating generally healthy individuals or those with only mild systemic disease. Local anesthetic selection was largely consistent, with five studies employing lidocaine (also known as lignocaine) [23,24,25,26,28] and one study utilizing levobupivacaine [27]. All trials uniformly adopted the IANB technique. Table 1 presents a detailed summary of the key characteristics of each study.

Table 1. Overview of the included studies.

Study, Year	Design	Country	Groups	Total number of patients	Surgical procedure	ASA	Age range	The aim of the study
Priyaranjan, 2022 [23]	comparative study	India	Lignocaine 2% + DEX 1 mcg/ml Lignocaine 2%+Adrenaline 1:80,000	40	Impacted third molar	N/A	20-40 years	This study primarily aimed to investigate whether the addition of DEX to lignocaine offers measurable advantages in third molar surgery, beyond those achieved with lignocaine alone.
Alizargar, 2021 [24]	RCT	Iran	Lidocaine + 0.15 ml of DEX Lidocaine	40	Impacted third molar	I/II	18-40 years	The primary objective of this study was to assess the effects of adding DEX to lidocaine in third molar surgery.
Nalawade, 2021 [25]	RCT	India	Lignocaine 2% + DEX 1 mcg/ml Lignocaine 2%+Adrenaline 1:200,000	80	Impacted third molar	N/A	18-35 years	To assess the effect of combining DEX with plain lignocaine for inferior alveolar and long buccal nerve blocks in patients undergoing surgical extraction of impacted mandibular third molars, and to compare it with 2% lignocaine containing 1:200000 adrenaline.
Etemadi, 2022 [26]	RCT	Iran	Lidocaine 2% + Adrenaline 1:80,000+DEX 0.1 ml Lidocaine 2%+Adrenaline 1:80,000	26	Impacted third molar	I/II	18-30 years	To assess the efficacy of DEX as an adjunct to LIDO in alleviating patients’ anxiety and enhancing cooperation during surgery.
Patil, 2022 [27]	RCT	India	Levobupivacaine 0.5% + DEX 0.2 ml Levobupivacaine 0.5%+Placebo 0.2 ml	50	Impacted third molar	I	22-48 years	To evaluate and compare the efficacy and safety of LB with or without DEX in LITM surgery, hypothesizing that adding DEX to LB may provide superior therapeutic benefit over LB alone for POA in the acute postoperative setting. The primary outcomes were POA duration and hemodynamic stability, while secondary outcomes included analgesic requirements and incidence of adverse effects.
Doshi, 2024 [28]	RCT	India	Lignocaine 2% + DEX 1 mcg/ml Lignocaine 2%+Adrenaline 1:100,000	30	Impacted third molar	I	21-30 years	To assess the efficacy of clonidine and DEX compared with adrenaline as adjuvants to local anesthesia, examining onset time, depth of anesthesia, hemodynamic parameters, and postoperative analgesic outcomes following extraction of impacted mandibular third molars.

Abbreviations: RCT, randomized controlled trial; DEX, dexmedetomidine; ASA, American Society of Anesthesiologists grade; LIDO, lidocaine; LB, levobupivacaine; LIMT, lower impacted third molar tooth; PAO, post-operative analgesia.

3. Risk of bias and certainty of evidence

The RoB in the included RCTs was assessed using the RoB-2 tool across five domains. All included studies were assessed as having “some concerns” regarding RoB. Specifically, Alizargar et al. [24] and Nalawade et al.’s [25] work was judged to have “some concerns” in the randomization process (Domain 1), deviations from intended interventions (Domain 2), and selection of reported results (Domain 5), resulting in an overall judgment of “some concerns.” Etemadi et al. [26] demonstrated low RoB across most domains, with some concerns noted in Domain 5 (selection of reported results). The study by Patil et al. [27] was assessed as having some concerns regarding the randomization process (Domain 1), while all other domains were rated low risk, resulting in an overall judgment of “some concerns.” Doshi et al.’s study [28] was rated low risk across most domains, with some concerns in Domain 2 (deviations from intended interventions). The RoB for the included non-randomized study by Priyaranjan et al. [23] was evaluated using the ROBINS-I tool, with an overall judgment of moderate risk (Fig. 2, 3; Supplementary table S4 and S5). The overall certainty of evidence for the primary outcomes ranged from low to moderate, according to the GRADE framework. The onset of anesthesia showed moderate-certainty evidence, downgraded for inconsistency due to substantial heterogeneity (I² = 92.2%). Pain reduction at 6–12 h also demonstrated moderate-certainty evidence, supported by a consistent direction of effect across studies. In contrast, outcomes related to SBP and DBP were rated as low certainty, primarily due to imprecision, with CIs of individual trials crossing the null. Table 2 summarizes the overview of the certainty of evidence for all outcomes.

Fig. 2. Risk of bias assessment for included randomized controlled trials using the RoB-2 tool.

Fig. 3. Risk of bias assessment for included non-randomized study using the ROBINS-I tool.

Table 2. Local anesthesia combined with dexmedetomidine compared with local anesthesia for impacted third molar surgery.

Certainty assessment									Summary of findings
									Anticipated absolute effects
Participants (studies)	Risk of bias	Inconsistency	Indirectness	Imprecision	Publication bias	Overall certainty of evidence	With Local anesthesia	With Local anesthesia with DEX	Risk difference with Local anesthesia with DEX
Onset of anesthesia (assessed with: time)
160 (3RCTs)	not serious	very seriousa	not serious	not serious	strong association	⊖⊖⊖◯ Moderatea	80	80	MD 0.96 lower (1.52 lower to 0.39 lower)
Systolic blood pressure (assessed with: mmHg)
120 (3RCTs)	not serious	not serious	not serious	very seriousb	none	⊖⊖◯◯ Lowb	60	60	MD 3.85 lower (6.58 lower to 0.58 lower)
Diastolic blood pressure (assessed with: mmHg)
120 (3RCTs)	not serious	not serious	not serious	very seriousb	none	⊖⊖◯◯ Lowb	60	60	MD 1.97 lower (4.29 lower to 0.35 higher)
Pain (12 hours assessed with: VAS)
90 (2RCTs)	not serious	not serious	not serious	not serious	all plausible residual confounding would reduce the demonstrated effectc	⊖⊖⊖◯ Moderate	45	45	MD 2.32 lower (2.85 lower to 1.78 lower)

CI, confidence interval; MD, mean difference; DEX, dexmedetomidine; RCT, randomized controlled trial; VAS, visual analog scale. Explanations: a. I² = 92.2%; b. All CIs of the studies crossed zero; c. differences in anesthetic agents, DEX doses, and patient characteristics.

4. Meta-analyses and synthesis of results

1) Onset of anesthesia

The pooled analysis of three studies [25,27,28] including 160 patients demonstrated a statistically significant reduction in onset time in the LA-DEX group compared with local anesthesia alone (MD = −0.96 min; 95% CI: −1.52 to −0.39; P = 0.0009). Heterogeneity was very high (I² = 92.2%, τ² = 0.2189; P = 0.0001). Individual study effects ranged from MD = −0.45 (Doshi et al. [28]) to MD = −1.40 (Patil et al. [27]), with all CIs excluding the null. A subgroup analysis by anesthetic agent revealed that in the lidocaine subgroup, DEX reduced onset time (MD = −0.69 min; 95% CI: −1.19 to −0.18; P = 0.0076) with moderate heterogeneity (I² = 65.6%, P = 0.088). In the levobupivacaine subgroup, heterogeneity was extreme (I² = 92.2%, P = 0.0001), and the effect size differed significantly from the lidocaine subgroup (subgroup differences: χ² = 7.36, P = 0.0067). Sensitivity analyses showed that excluding Nalawade et al. [25] yielded MD = −0.94; 95% CI: −1.87 to −0.01; P = 0.0476) with increased heterogeneity (I² = 95.8%). Excluding Patil et al. [27] reduced heterogeneity to I² = 65.6% (MD = −0.69; 95% CI: −1.19 to −0.18). Excluding Doshi et al. [28] strengthened the effect size (MD = −1.25; 95% CI: −1.66 to −0.83; P = 0.0001) with I² = 67.5% (Fig. 4A–C).

Fig. 4. (A) Forest plot for pooled results for onset of anesthesia. (B) Subgroup analysis by anesthetic type. (C) Leave-one-out sensitivity analysis. CI, confidence interval; DEX, dexmedetomidine; MD, mean difference;SD standard deviation.

2) Systolic and diastolic blood pressure

The pooled analysis of three studies [24,27,28] including 120 patients revealed a statistically significant reduction in SBP in the LA-DEX group, with no heterogeneity (MD = −3.58 mmHg; 95% CI: −6.58 to −0.58); P = 0.020). The largest effect was observed in Alizargar et al. [24] (MD = −4.00 mmHg), while Doshi et al. [28] reported the smallest effect (MD = −2.67 mmHg). However, no statistically significant difference was observed in DBP between the LA-DEX group and local anesthesia (MD = −1.41 mmHg; 95% CI: −4.29 to 0.35); P = 0.095), with negligible heterogeneity (I² = 0%, P = 0.560). Individual study results for DBP ranged from MD = −0.50 mmHg (Alizargar et al. [24]) to MD = −3.73 mmHg (Doshi et al. [28]), with all CIs crossing the null (Fig. 5A, B).

Fig. 5. (A) Forest plot for pooled results for systolic blood pressure (SBP). (B) Forest plot for pooled results for diastolic blood pressure (DBP). CI, confidence interval; DEX, dexmedetomidine; MD, mean difference;SD standard deviation.

3) Postoperative pain

The meta-analysis of VAS scores from two studies [24,27] including 90 patients after 6 and 12 h demonstrated a statistically significant reduction in postoperative pain in the LA-DEX group compared with local anesthesia (MD = −2.32; 95% CI: −2.85 to −1.78; P = 0.0001), with substantial heterogeneity (I² = 88.8%, τ² = 0.1789; P = 0.0001). A subgroup analysis by time point was conducted. At 6 h postoperatively, pain was significantly reduced (MD = −2.30; 95% CI: −3.53 to −1.07; P = 0.001) with moderate heterogeneity (I² = 69.3%, P = 0.071). At 12 h postoperatively, pain remained significantly lower (MD = −2.48; 95% CI: −2.65 to −2.30; P = 0.0001) with no heterogeneity observed (I² = 0%, P = 0.775) (Fig. 6).

Fig. 6. Forest plot for pooled results of pain scores with subgroup analysis by time point at 6 and 12 h. CI, confidence interval; DEX, dexmedetomidine; MD, mean difference; SD standard deviation.

4) Heart rate

The pooled analysis of three studies [24,27,28] including 120 patients showed no statistically significant difference in HR between the LA-DEX group and local anesthesia (MD = 2.12 bpm; 95% CI: −1.81 to 6.05; P = 0.291), despite substantial heterogeneity (I² = 65.4%, P = 0.056). Doshi et al. [28] reported the largest increase in HR (MD = 5.31 bpm), whereas Patil et al. [27] favored the control group (MD = −1.12 bpm). A subgroup analysis by anesthetic agent revealed a statistically significant HR increase with DEX in the lidocaine subgroup (MD = 4.23 bpm; 95% CI: 1.08 to 7.39; P = 0.0095) with no heterogeneity (I² = 0%, P = 0.417), with Doshi et al. [28] showing the largest effect (MD = 5.31 bpm; 95% CI: 1.23 to 9.39). In the levobupivacaine subgroup, no statistically significant trend was observed (MD = −1.12 bpm; 95% CI: −4.52 to 2.28). Furthermore, sensitivity analysis indicated that heterogeneity was resolved upon exclusion of Patil et al. [27] (I² = 0%) (Supplementary figure S1, S2, S3).

5) Oxygen saturation

The meta-analysis of two studies [24,27] including 90 patients revealed no statistically significant difference in SpO₂ between the LA-DEX group and local anesthesia (MD = 0.21%; 95% CI: −0.35 to 0.77; P = 0.455). Heterogeneity was negligible (I² = 0%, τ² = 0; P = 0.597). Notably, Alizargar et al. [24] reported no change in mean SpO₂ (MD = 0.00%; 95% CI: −0.97 to 0.97) (Supplementary Figure S4).

6) Postoperative analgesia

The duration of postoperative analgesia, evaluated in two studies [27,28] with 80 patients, was significantly prolonged in the LA-DEX group compared with local anesthesia (MD = 3.75 min; 95% CI: 1.66 to 5.85; P = 0.0005), with extreme heterogeneity observed (I² = 99.0%, τ²= 2.2673; P = 0.0001). Patil et al. [27] reported the largest effect (MD = 4.82 min; 95% CI: 4.58 to 5.06), while Doshi et al. [28] showed a smaller but still significant prolongation (MD = 2.68 min; 95% CI: 2.34 to 3.02) (Supplementary figure S5).

7) Postoperative analgesics consumption

The meta-analysis of postoperative analgesic consumption, based on two studies [24,27] including 90 patients, revealed no statistically significant reduction in analgesic use in the LA-DEX group compared with local anesthesia (MD = −1.10 tablets; 95% CI: −2.72 to 0.52; P = 0.183), with high heterogeneity (I² = 97.4%, τ² = 1.3308; P = 0.0001). Specifically, Patil et al. [27] reported a substantial reduction in analgesic use (MD = −1.92 tablets, 95% CI: −2.21 to −1.63, while Alizargar et al. [24] found a non-significant effect (MD = −0.27 tablets, 95% CI: −0.70 to 0.16) (Supplementary Figure S6).

Furthermore, some outcomes, particularly those related to patient cooperation, could not be incorporated into the statistical meta-analysis. Statistically significant differences were observed in Etemadi et al. [26] between the two groups (lidocaine + DEX and lidocaine alone). Specifically, patients in the lidocaine-alone group demonstrated greater interfering movements and more frequent verbal expressions of discomfort compared with the lidocaine + DEX group, as confirmed by the Wilcoxon test (P = 0.021). Nalawade et al. [25] reported that the mean duration of anesthesia was 76.71 min in the control group and 98.48 min in the study group, with a statistically significant difference of 21.76 min, favoring the study group (P = 0.001). Table 3 presents an overview of the overall findings.

Table 3. Summary of overall findings.

Outcome	No. of studies	Total No.	MD	95% CI	P-valuea	I2b	GRADE assessment
Onset of anesthesia	3	160	-0.96	[-1.52; -0.39]	0.0009	92.2%	⊕⊕⊕⊖ moderate
Systolic blood pressure	3	120	-3.58	[-6.58; -0.58]	0.0195	0%	⊕⊕⊖⊖ low
Diastolic blood pressure	3	120	-1.97	[-4.29; 0.35]	0.0954	0%	⊕⊕⊖⊖ low
Pain (VAS) 6 and 12h	4	180	-2.32	[-2.85; -1.78]	0.0001	88.8%	⊕⊕⊕⊖ moderate
Heart rate	3	120	2.12	[-1.81; 6.05]	0.2908	65.4%	⊕⊖⊖⊖ very low
SpO2	2	90	0.21	[-0.35; 0.77]	0.4546	0%	⊕⊖⊖⊖ very low
Duration of post-operative analgesia	2	80	3.75	[1.66; 5.85]	0.0005	99%	⊕⊖⊖⊖ very low
Analgesics consumption	2	90	-1.10	[-2.72; 0.52]	0.1826	97.4%	⊕⊖⊖⊖ very low

a. A p-value of <0.05 is statistically significant. b. Heterogeneity was assessed using the Q-test and I² statistics. Abbreviations: MD, mean difference; CI, confidence interval; GRADE, Grading of Recommendations, Assessment, Development and Evaluation; VAS, visual analog scale.

DISCUSSION

This systematic review and meta-analysis demonstrated that incorporating DEX with local anesthetics for IANB in impacted third molar surgery resulted in a statistically significant improvement in anesthetic onset and postoperative pain scores, and may also have prolonged the duration of postoperative analgesia compared with local anesthesia alone. These findings align with those of previous meta-analyses assessing DEX as an adjuvant for brachial plexus block [29,30,31] and paravertebral block [32]. Moreover, while the addition of DEX did not significantly increase the incidence of bradycardia (MD = 2.12 bpm; 95% CI: −1.81 to 6.05; P = 0.2908), its use was associated with a higher likelihood of hypotension, indicating a potential compromise in hemodynamic stability. The reduction in BP was primarily attributable to systolic values, whereas no statistically significant difference was observed in DBP. This suggests that the hypotensive effect of DEX during IANB is mild and predominantly systolic, which may be clinically relevant when monitoring patients with cardiovascular risk. However, some outcomes in this analysis were marked by substantial heterogeneity. To address this, subgroup and sensitivity analyses were performed to identify the potential sources of variability. Given the limited number of studies in the literature, our results and subgroup analysis should be interpreted with caution. Nevertheless, our findings provide preliminary evidence and underscore the need for future, more comprehensive evaluations of DEX in combination with local anesthetics for IANB, particularly considering the limited number of trials and their geographic concentration.

The improved clinical outcomes observed in the DEX group may be attributed to peripheral mechanisms or central effects arising from the absorption and systemic redistribution of perineurally administered DEX. Fritsch et al. [33] assessed plasma concentrations of DEX following perineural administration of 150 µg with ropivacaine in an interscalene nerve block and concluded that the prolongation of block duration was not due to systemic absorption. Evidence from two volunteer studies [34,35] and one animal experiment [36] suggests that perineural co-administration of DEX with local anesthetics significantly prolongs the duration of nerve block, an effect attributed primarily to peripheral rather than systemic mechanisms. The proposed peripheral analgesic action of DEX involves suppression of norepinephrine release and direct inhibition of nerve fiber action potentials mediated through α₂-adrenergic receptor activity [37]. Furthermore, a randomized clinical trial conducted to explore the enhancement of DEX in IANB in patients with irreversible pulpitis revealed that DEX increased the success rate of IANB to 72% (P = 0.0001) and enhanced the anesthetic effect of lidocaine [14].

Our results also showed that LA-DEX reduced pain scores at 6 and 12 h postoperatively (MD = −2.30 and −2.32, respectively). These results are consistent with those of previous meta-analyses [32,38], which reported pain reductions of 1.03 and 0.86, respectively. Despite conducting subgroup and sensitivity analyses to explore the sources of heterogeneity, their origins could not be fully determined. Similarly, our study demonstrated considerable heterogeneity in the primary outcomes, highlighting a persistent challenge in interpreting the pooled evidence. However, the subgroup analysis in this study indicated that the type of local anesthetic may account for the heterogeneity in postoperative pain scores.

Although our review searched the relevant databases and was not restricted to RCTs, many limitations should be acknowledged. First, the primary limitation of this study is the heterogeneity among the included trials, driven by differences in anesthetic agents—five trials used lidocaine, whereas only one trial used levobupivacaine—as well as variations in DEX dosage and study quality. These studies were pooled because they addressed the same clinical question—the adjuvant effect of DEX in IANB for impacted third molar surgery, regardless of the base anesthetic. A random-effects model was selected to account for the expected variability across anesthetic types, DEX doses, and study designs. Subgroup analyses were therefore conducted by anesthetic agent to assess potential effect modification.

Second, data were pooled despite high heterogeneity because all included trials evaluated the same intervention–outcome relationship using comparable methodologies. Heterogeneity was addressed through subgroup and sensitivity analyses. Even with substantial heterogeneity, the direction of the effect remained consistent, supporting the robustness of the pooled estimates. Additionally, although high heterogeneity downgraded the certainty of evidence from “high” to “moderate” or from “moderate” to “low,” the consistency in direction of effect and reproducible benefits across analyses strengthens the overall validity of the conclusions. All participants were ASA I–II adults, minimizing systemic bias; however, variations in surgical techniques and operator experience may also have contributed to the observed heterogeneity. Third, the included trials used DEX concentrations between 0.1–0.2 mL (approximately 1 µg/mL) for lidocaine or levobupivacaine cartridges. These variations may influence anesthetic onset and duration due to dose-dependent α₂-adrenergic receptor activation, which modulates sodium channel blockade and peripheral nerve hyperpolarization.

Fourth, a formal dose–response analysis was not feasible due to the limited number of trials and narrow dosing range (0.1–0.2 mL). However, this limitation has been explicitly acknowledged as a gap that warrants further investigation.

Fifth, although the reduction in onset time (approximately 1 min) and decrease in SBP (approximately 3 mmHg) reached statistical significance, their clinical importance is likely minimal. Conversely, the reduction in postoperative pain (approximately 2 VAS points) exceeded the MCID (1–2 points), indicating a meaningful clinical benefit. These variations necessitate cautious interpretation. Additionally, the overall number of studies and sample sizes was relatively small, with only six trials and a limited number of participants, which limits the generalizability of our findings. Furthermore, because DEX is currently approved by the Food and Drug Administration only for intravenous administration, all included trials were conducted in developing countries such as India and Iran, raising the possibility of publication bias.

Finally, although its efficacy is important, hemodynamic safety remains a critical consideration when determining whether DEX should be administered perineurally or systemically. Notably, perineural DEX use remains off-label, and evidence on its long-term neurotoxicity is limited, underscoring the need for further safety evaluations [8]. Therefore, future research should prioritize long-term safety outcomes and mechanistic investigations of perineural DEX administration for impacted third molar surgery. Future studies should also adopt standardized anesthetic protocols, larger sample sizes, uniform DEX dosing, patient-centered outcomes, comprehensive safety assessments, and higher methodological rigor to reduce bias and improve reliability.

In conclusion, within the limitations of this study, the addition of DEX to local anesthetics for impacted third molar surgery may enhance anesthetic quality by modestly improving onset time, postoperative pain control, and duration of analgesia. However, the overall certainty of the evidence remains limited. Therefore, although DEX shows potential as an adjuvant to local anesthesia, its routine use cannot yet be recommended. Instead, DEX may be considered selectively in appropriate clinical contexts, pending validation through larger, high-quality RCTs with standardized protocols and comprehensive long-term safety assessments.

DATA AVAILABILITY

All data are available in the manuscript and supplementary materials. Further details can be requested from the corresponding authors.

SUPPLEMENTARY MATERIALS

Supplementary Table S1

PRISMA Checklist.

Supplementary Table S2

AMSTAR2 assessment

Supplementary Table S3

Search strategy and results from databases.

Supplementary Table S4

Risk of bias assessment using the RoB2 Cochrane tool.

Supplementary Table S5

Risk of bias assessment using the ROBINS-I tool.

Supplementary Figure S1

Forest plot of heart rate.

Supplementary Figure S2

Forest plot of Heart rate divided by Anesthetic subgroup.

Supplementary Figure S3

Sensitivity analysis of heart rate.

Supplementary Figure S4

Forest plot of oxygen saturation (SpO2).

Supplementary Figure S5

Forest plot of Duration of postoperative analgesic effect.

Supplementary Figure S6

Forest plot of analgesics consumption after the surgery.