Investigating the efficacy of dexmedetomidine as an adjuvant to local anesthesia in dentistry: systematic review and meta-analysis

Abstract

Background

Dexmedetomidine (DEX) has been considered an effective adjunct to dental local anesthetics. Uncertainty about its efficacy and safety in dental use still remains.

Objectives

This study is primarily designed to evaluate the effectiveness and safety of DEX as an adjunct to local anesthesia in dental procedures.

Methods

This study followed PRISMA guidelines. The protocol was registered in PROSPERO (CRD420251035026). Outcomes were onset time, anesthesia duration, postoperative analgesia duration, post operative pain, sedation, and adverse events. Searches covered PubMed, Cochrane Library, Ovid, Embase, Web of Science, and Google Scholar. Risk of bias was assessed with Cochrane RoB 2 and ROBINS-I. RevMan 5.1 pooled mean differences and assessed heterogeneity. Sensitivity analyses and publication bias were examined in Comprehensive Meta-Analysis v3.7. Certainty of evidence was rated with GRADE.

Results

This meta-analysis encompassed sixteen controlled clinical trials comprising 658 participants. The addition of DEX decreased onset time by 48.78 s (95% CI: 64.22 to 33.35 s, p < 0.00001). This reduction is unlikely to be clinically significant in routine dental practice. DEX extended anesthesia duration by 31.10 min (95% CI: 13.77 to 48.43 min, p = 0.0004), and prolonged postoperative analgesic duration by 186.19 min (95% CI: 109.09 to 263.29 min, p < 0.00001), although both findings demonstrate substantial uncertainty due to unexplained heterogeneity (I² = 94% and 99%, respectively). Hemodynamic evaluation revealed no clinically significant alterations in heart rate, intraoperative systolic and diastolic pressure, and oxygen saturation (p > 0.05 for all parameters). Minor variations in heart rate and blood pressure did not advance to hypotension or bradycardia.

Conclusions

Dexmedetomidine as an adjunct to dental local anesthesia may reduce onset time, extend anesthesia duration, and postoperative analgesia with a favorable hemodynamic safety profile; however, substantial unexplained heterogeneity and modest sample size limit confidence in these estimates. Current evidence does not support routine clinical use; standardized, adequately powered trials are needed before definitive recommendations can be made.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13643-026-03202-0.

Introduction

Effective postoperative pain control is an integral part of patient care in dentistry, especially for procedures such as third molar extractions, which are known to be significantly painful postoperatively [1–3]. Local anesthesia forms the main modality of pain management in dental surgery; however, it suffers from drawbacks related to a very limited duration of action and systemic toxicity with larger doses of local anaesthetics [4]. Consequently, the search for adjuvants that enhance the quality and duration of anesthesia and analgesia with a reduction of adverse effects has been continued [4].

DEX is an α2-adrenergic receptor agonist that has demonstrated high selectivity and appeared as an effective adjuvant to extend the action of local anesthetics [5–10]. DEX possesses analgesic, anxiolytic, and sedative effects with minimal respiratory depression [5–7, 9, 11–14]. When used with local anesthetics, DEX can prolong analgesia duration, reduce postoperative pain scores, and reduce the need for rescue analgesics [5–12, 15, 16]. Numerous systematic reviews and meta-analyses have established DEX’s efficacy in regional anesthesia techniques, including peripheral nerve blocks, neuraxial blocks, and intravenous regional anesthesia [5–8, 10, 11, 17, 18]. The application of DEX as an adjuvant with local anesthesia has become a subject of interest in dentistry; however, there is still a need for an overview regarding efficacy and the profile of this drug’s safety. Notably, no meta-analysis has quantitatively synthesized the available evidence for DEX as an adjunct to dental local anesthesia, although individual dental trials have reported favorable outcomes [5, 6, 10, 15, 19, 20].

Beyond efficacy considerations, the safety profile of DEX warrants careful evaluation in dental settings. While generally considered safe, DEX can cause hemodynamic changes, including bradycardia and hypotension [3, 12, 15, 16]. These potential hemodynamic effects are particularly relevant in dentistry, where the majority of procedures are performed as day-care surgeries requiring rapid recovery with minimal adverse effects [3, 9]. Although several studies have investigated the hemodynamic impact of DEX as an adjuvant to dental local anesthesia, the findings have been conflicting [3, 4, 9, 13, 19–25], further underscoring the need for systematic synthesis of available evidence.

Therefore, the primary objective of our systematic review and meta-analysis is to assess whether DEX decreases the onset time and prolongs the duration of dental nerve blockade as an adjuvant to local anesthesia compared with the use of local anesthesia alone (control) in participants ≥ 10 years of age undergoing dental procedures. We further want to elaborate on the adverse events profile of DEX in order to understand its safety.

Methods

The PRISMA methodological approach was adopted in the present study to systematically conduct the review and meta-analysis. It, therefore, provided a systematic guide on the identification, evaluation, and synthesis of studies in a manner that made the review as transparent and reproducible as possible. In line with the PRISMA guidelines, we set out to provide a comprehensive overview of the available evidence with respect to the effectiveness and safety of DEX as an adjunct to local dental anesthesia [26]. Furthermore, the study protocol was registered with PROSPERO (CRD420251035026), and the complete PRISMA checklist is available in Additional File A.

Eligibility criteria

This systematic review applied PICOS criteria for study selection [27]. Studies were eligible if they involved participants ≥ 10 years receiving dental procedures where DEX was used as a local anesthetic adjuvant versus control groups (local anesthesia alone or with alternative adjuvants). Inclusion required reporting at least one of these outcomes: onset time, anesthesia duration, postoperative analgesia duration, pain scores, sedation, or adverse events (hemodynamic parameters). Only clinical trial articles published in English are included. Exclusion criteria included studies examining dental pulpal anesthesia, participants < 10 years, and publications lacking sufficient data for meta-analysis.

Search strategy

An experienced evidence-based medicine librarian (TW) developed a comprehensive electronic search strategy to identify both published and unpublished studies across multiple databases: MEDLINE (via PubMed), EMBASE, Cochrane Library, Ovid, Web of Science, and Google Scholar from January 2014 to August 10, 2025. The search strategy utilized a combination of keywords and index terms including: dexmedetomidine and its synonyms (DEX, Precedex, Dexdor, MPV-2426), adjuvant terms (adjuvant, additive, enhance, combined, added), anesthetic agents (anesthesia, lidocaine, lignocaine, levobupivacaine, ropivacaine, local anesthetic), and dental procedure terms (nerve block, dentistry, dental, oral surgery, local anesthesia). Boolean operators (AND, OR) were used to combine search concepts, with OR operators connecting synonymous terms within each concept and AND operators linking different concepts together. The search was limited to title and abstract fields, where possible, to maintain precision while ensuring comprehensive coverage. A complete electronic search strategy is provided in Additional File B (Table 1).

Study selection

The title of each search result was meticulously screened by two independent reviewers, SAA and WAA. First of all, screening was done in duplicate based on title and abstract. Following that, all the potentially qualified articles would undergo a full-text review to confirm their eligibility for inclusion. During this time, any discrepancies arising between these two reviewers were resolved through collaboration by discussion to a consensus. In case there was no consensus through discussion, disagreements were adjudicated by an independent third reviewer (NL). Lastly, the bibliographies and citations of all the included articles were carefully scanned to ensure the comprehensiveness of the search strategy regarding any missed studies.

Outcomes assessed

Primary outcomes were onset time and duration of dental local nerve blockade with DEX versus control. Onset time was defined as the interval from injection completion to effective sensory block (loss of response to pinprick or cold testing; measured in seconds). Duration of anesthesia was defined as the interval from full sensory block to return of sensation (measured in minutes). Secondary outcomes were postoperative pain at 12–24 h (validated scales, e.g., Visual Analog Scale (VAS) or Faces), duration of postoperative analgesia (minutes from procedure end to first rescue analgesic or pain threshold, typically VAS ≥ 3), sedation (trial-reported scales), and safety (heart rate, systolic/diastolic blood pressure, oxygen saturation measured intra- and postoperatively; thresholds per original trials).

Data management and extraction

A data extraction form was designed and then piloted by an independent reviewer (SAA). The data extraction form collected information regarding the study characteristics (first author, year, country/setting, design, sample size), clinical setting and demographics; interventions (local anesthetic type/concentration and DEX dose), comparator details; outcomes with definitions/units (e.g., anesthesia onset and duration) and adverse events; follow-up timing; and risk-of-bias judgments. The data were extracted from all included studies irrespective of risk-of-bias rating; risk of bias informed GRADE and sensitivity analyses. Two independent reviewers (SAA, WAA) extracted data to ensure accuracy and minimize the risk of error. In the case of a disagreement in data extraction, the two reviewers discussed until a consensus was reached. If a consensus still could not be reached, a third reviewer (NL) was tasked with making the final decision.

Methodological quality assessment process

The two reviewers (SAA and WAA) independently conducted the methodological quality assessment of the included studies. The Revised Cochrane Risk-of-Bias tool (RoB 2) was used to evaluate the 14 randomized controlled trials across five domains, with risk levels categorized as high (red), unclear (yellow), or low (green) [28]. The ROBINS-I tool assessed the two non-randomized studies across seven domains, with risk levels designated as serious (dark red), critical/high (red), moderate (yellow), or low (green) [29]. An unweighted kappa statistic was calculated to evaluate initial interrater agreement. Disagreements between reviewers were resolved through discussion, and when consensus couldn’t be reached, a third reviewer (LN) was consulted for final decision-making.

Data synthesis

This meta-analysis was carried out using Review-Manager software provided by the Cochrane Collaboration (RevMan V.5.1). Continuous outcomes were pooled as mean differences (MD) with 95% confidence intervals using the inverse-variance method. Heterogeneity was assessed with Cochran’s Q (chi-square; P < 0.10) and quantified with (I² > 50%) were considered substantial. A random-effects model (DerSimonian–Laird) was used when statistical or clinical heterogeneity was high; a fixed-effect model was used when heterogeneity was low (I² < 30%). A P value of ≤ 0.05 was established as the significance level for reporting the overall effect size. Studies were included in the meta-analysis only when reporting complete statistical data. This required a sample size and means with dispersion measures in compatible units for each arm. Studies lacking means, standard deviations, or other necessary statistics were excluded from pooled analyses. These studies were still included in the systematic review for qualitative assessment.

Sensitive analyses and meta-regression

Sensitivity analyses and random-effects meta-regression (REML) were conducted in Comprehensive Meta-Analysis v3.7 to assess the robustness of pooled effects. Sensitivity tests included leave-one-out deletion, fixed-effect versus random-effects model comparisons, and restriction to more homogeneous subsets (e.g., lidocaine-only trials and studies using 1 μg/mL dexmedetomidine), with stability of effect estimates and changes in heterogeneity evaluated across scenarios. Meta-regression examined whether dexmedetomidine dose and anesthetic formulation moderated onset time and duration (mean difference, minutes), using prespecified moderators: DEX dose (1 vs > 1) µg/mL; anesthetic type for onset (lidocaine vs non‑lidocaine); and lidocaine concentration for duration (2% vs 1%). Univariable and multivariable models reported moderator coefficients with 95% CIs and two‑sided p values, used the omnibus Qm test for joint significance, quantified heterogeneity with τ² and residual I², and summarized explained variance with the R² analog (α = 0.05).

Publication bias

Publication bias was assessed in Comprehensive Meta-Analysis v3.7 using funnel plots and formal tests: Egger’s regression (small-study effects), Begg–Mazumdar rank correlation, Duval and Tweedie’s trim-and-fill, and the classic fail-safe N. Analyses were restricted to outcomes with n ≥ 10. Secondary outcomes were not assessed for publication bias due to their exploratory nature and insufficient study numbers.

GRADE assessment of evidence certainty

Certainty of evidence was assessed with GRADE at the outcome level [30]. Two reviewers (SAA, WAA) independently evaluated risk of bias, inconsistency (I²/CI overlap), indirectness, imprecision (95% CI width and optimal information size), and publication bias, resolving disagreements by consensus or a third reviewer. Randomized evidence started high and nonrandomized low; we downgraded for serious/very serious concerns and upgraded for large effects, dose–response, or residual confounding. Final GRADE ratings (high, moderate, low, very low) were reported per outcome in the Summary of Findings.

Results

Study characteristics

At first, a literature search identified a total of 1806 articles. Ninety-nine articles were assessed as a primary result for initial screening. After screening the title and abstract, 24 articles were assessed for full-text eligibility. Of these articles, 16 articles were included in this meta-analysis, eight articles were excluded because of their age < 10 [31–34], article review [35, 36], and dental pulp studies [14, 37]. The full-flow PRISMA diagram of study inclusion is shown in Fig. 1. A comprehensive overview of all included studies is presented in Additional File C (Table 1). All studies were published between 2014 and 2024. Across all included studies, 658 participants (877 arch-level injections across the mandibular and maxillary arches) were assessed. Studies were conducted in four countries: India (n = 10), Japan (n = 3), Iran (n = 2), and Egypt (n = 1). Regarding local anesthetic techniques, 1 study used a submucosal injection [38], and 15 used nerve blocks [2–4, 13, 19–24, 39–43]. Of these 15, 11 used inferior alveolar nerve block (IANB) [2, 3, 13, 19–22, 39, 40, 42, 43], and 4 used infraorbital nerve block (IONB) in combination with IANB [4, 23, 24, 41]. DEX was used as an adjuvant to three different local anesthetics, which included lidocaine [2–4, 13, 19–24, 38, 40, 41, 43], levobupivacaine [39], and ropivacaine [42]. Across the studies, the dose of DEX ranged from 0.24 μg to 15 μg, and the dosages of local anesthetics varied among the studies.

Fig. 1.

PRISMA flow of study selection. Legend: Flow diagram showing identification of records via databases and trial registers (and other sources), removal of duplicates, title/abstract screening, full-text eligibility assessment with reasons for exclusion, and the final number of studies included in the qualitative synthesis and meta-analysis. Follows PRISMA 2020 (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines

Quality assessment

After bias assessment, the 14 RCTs assessed with RoB 2 mostly showed low bias (green), but some had high bias (red) in blinding [13, 21, 39] and unclear bias (yellow) in selective reporting or other domains. For the 2 non-randomized studies using ROBINS-I, Priyaranjan et al. (2020) had critically high bias (red) in confounding and selection, while Kosar et al. (2021) showed low bias overall with moderate bias (yellow) in selective reporting. Overall, the RCTs had fewer bias concerns than the non-randomized studies. For more information, the bias assessment of included studies was clarified in Fig. 2a–b.

Fig. 2.

Risk of bias summary. Legend: a ROBINS‑2, RCTs — Fourteen RCTs were mostly low risk (green); a few showed high risk in blinding; some had unclear risk (yellow) in selective reporting or other domains. b ROBINS‑I, non‑randomized — Priyaranjan 2020 had critical risk (red) in confounding and selection; Kosar 2021 was overall low risk, with moderate risk (yellow) for selective reporting

Onset of anesthesia

Of 16 studies, 12 assessed the onset time of anesthesia; however, eight studies (n = 560) had sufficient information to allow for pooling [4, 21–24, 39, 41, 42]. The addition of DEX significantly decreased the onset time of anesthesia of nerve blockade by an average of −48.78 s in comparison to the control (95% CI: −64.22 to −33.35; I = 95%), favoring DEX, with a highly significant overall effect (Z = 6.20, p < 0.00001) (Fig. 3a). Subgroup analyses were performed because the heterogeneity was above our predefined cutoff. Studies were grouped according to mandibular comparing to maxillary anesthesia (Fig. 3b), DEX concentration 1 µg/mL (Fig. 3c), and lidocaine 2% (Fig. 3d). The mandibular anesthesia meta-analysis mean difference was −47.94 s (95% CI: −74.01 to −21.86), showing a faster onset. Heterogeneity was high (I² = 98%). Maxillary Anesthesia mean difference was −49.88 s (95% CI: −63.76 to −35.99), with significant efficacy in reducing onset time and slightly high heterogeneity (I² = 62%). While the subgroup for one µg/mL DEX showed a mean difference of −43.72 s (95% CI: −59.69 to −27.76). This consistent improvement underlines the effective concentration used, though with high heterogeneity (I² = 94%). It was even stronger when put in relation to lidocaine, and the mean difference was −49.09 s (95% CI: −60.11, −38.08). The heterogeneity was lower than the others, with I² = 83%, indicating some variability, yet a strong general effect.

Fig. 3.

Forest plots of onset time for nerve blockade. Legend: (a) shows the overall mean difference (seconds) with 95% CI for dexmedetomidine versus control; (b) shows site-specific subgroups (maxillary and mandibular); (c) shows the subgroup using dexmedetomidine 1 µg/mL; and (d) compares dexmedetomidine adjunct with lidocaine. Negative mean differences indicate faster onset with dexmedetomidine. Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

Duration of anesthesia

Out of 16 studies, 13 revealed the duration time of anesthesia, but only seven studies had sufficient information to allow for pooling [2, 4, 20, 22, 24, 40, 41]. The meta-analysis compiled data from the seven studies involving 534 patients, with 267 assigned to the DEX group and 267 to the control group. The pooled results showed a mean difference of 31.10 min (95% confidence interval: 13.77 to 48.43) in favour of the DEX group. The findings were statistically significant, with a Z-value of 3.52 and a p-value of 0.0004. However, the meta-analysis showed significant heterogeneity among the included studies, with an I² value of 94% (Fig. 4). This high degree of heterogeneity suggests that there might be significant changes in methodology, study designs, and patient demographics—all elements that could affect results.

Fig. 4.

Forest plots of duration of nerve blockade in minutes. Legend: Dexmedetomidine vs control—pooled mean difference 31.1 min (95% CI 13.8–48.4; random-effects), favoring dexmedetomidine. Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

Analgesic duration

The meta-analysis of analgesia duration (time from procedure end to first request for rescue analgesia or pain score ≥ 3) pooled four studies [2, 21, 23, 39] involving 270 patients (135 DEX, 135 control). The mean difference in analgesia duration was 186.19 min (95% CI: 109.09 to 263.29; Z = 4.73, p < 0.00001), favoring DEX. However, high unexplained heterogeneity (I² = 99%) was present, and the pooled estimate should be interpreted with caution (Fig. 5).

Fig. 5.

Forest plots of Mean duration of postoperative analgesia. Legend: Dexmedetomidine vs control—pooled mean difference 186.19 min (95% CI: 109.09–263.29; random-effects), favoring dexmedetomidine; heterogeneity high (I² = 99%). Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

For postoperative analgesic requirement and consumption, seven studies [2, 3, 21, 23, 39, 43] reported that the addition of DEX significantly decreased postoperative analgesic requirement at12 to 24 h or more in comparison to control. The amount of postoperative analgesic drug consumed during 24 h of following by the patients exposed to DEX combined with local anesthesia was significantly less than standard local anesthesia with P < 0.05 [2, 23, 39]. Duration of postoperative analgesia was determined by measuring the time from the end of the procedure until the patient’s pain score reached a predefined threshold (e.g., VAS ≥ 3) or they requested rescue analgesia [21, 23]. Patients were commonly prescribed a combination of analgesics, including Paracetamol, Diclofenac Potassium, Enzoflam, Ketorolac Tromethamine, and pethidine, for pain relief [2, 21, 23, 43]. The studies also documented the use of rescue analgesics, typically intravenous Tramadol, for breakthrough pain when the VAS score was ≥ 4 [2, 23]. Because of the different analgesics used and the lack of sufficient reporting to allow for pooling, we could not estimate the effect of this outcome.

Postoperative pain

The included studies emphasized the importance of postoperative pain management and follow-up in dental procedures, particularly third molar extractions, which are known to induce moderate to severe pain [21, 39]. Varied pain scores and follow-up protocols were used within the first 12 to 24 h after surgery to evaluate the efficacy of various anesthetic procedures. The Visual Analogue Scale (VAS) was widely employed in seven included investigations [2, 3, 23, 24, 39, 42, 43], requiring patients to mark their pain level on a 10-cm line ranging from 10 (worst agony possible) to 0 (no pain). However, just one study used the Faces Pain Scale (FPS) [21], which depicts varying levels of pain through a sequence of facial emotions ranging from smiling to crying (Table 1). Pain intensity was measured regularly after the surgery to improve patient comfort and check analgesia effectiveness. The studies assessed pain at various time intervals following surgery, including 1, 3, 6, and 12 h [21, 39, 42]. This enabled researchers to monitor the progression of pain alleviation and identify potential rebound pain. Overall, most studies reported lower postoperative pain scores with DEX added to local anesthetic compared with control regimens (local anesthetic alone or with adrenaline) [2, 3, 23, 24, 39, 42, 43].

Table 1.

Individual postoperative pain outcomes at 12–24 h: dexmedetomidine vs control

Study ID	Dexmedetomidine	Control	P Value	Pain Scoring System Used
	Main (SD)	Main (SD)
Bhardwaj et al.,2024 [42]	2	5	p < 0.05	Visual Analog Scale (VAS), postoperative pain score at 12 h
Doshi et al., 2024 [21]	7.22 (0.56) h†	4.54 (0.38) h† 2.1 (0.38) h†*	p = 0.0001	Face Pain Scale (FPS), pain score reached a level of 3 or higher, the time was recorded, (†or † *) Another 2 groups
Patil et al., 2022 [39]	1	4	P = 0.0001	Visual Analog Scale (VAS), postoperative pain score at 24 h, Estimated from figure
Suryawanshi et al., 2022 [23]	443.60 99.28) min	332.50 (73.20) min‡	p = 0.000	Visual Analog Scale (VAS), ‡The time recorded from the point of completion of the extraction (In maxillary) to the time when the pain intensity was scored ≥ 3
Suryawanshi et al., 2022 [23]	446.12 (86.32) min‡	335.13 (69.82) min‡	p = 0.000	Visual Analog Scale (VAS), ‡The time recorded from the point of completion of the extraction (In mandibular) to the time when the pain intensity was scored ≥ 3
Elsawy et al., 2021 [2]	2 (2,3)*	3 (3,3)*	p < 0.05	Visual Analog Scale (VAS) pain score during 24 h of follow up, *Interquartile range
Priyaranjan et al., 2020 [3]	3	4	p < 0.05	Visual Analog Scale (VAS) pain score
Singh et al., 2017 [24]	Not specified	Not specified	Not specified	Visual Analog Scale (VAS)
Alizargar et al., 2021 [43]	1.60 (2.37)	4.00 (2.49)	p = 0.003	Visual Analog Scale (VAS), postoperative pain score at 24 h

Individual study results for postoperative pain at 12–24 h: across nine studies, dexmedetomidine generally reduced pain scores or prolonged time to reach the pain threshold versus control (p-values as reported)

Abbreviations: VAS Visual Analog Scale, FPS Face Pain Scale, SD standard deviation, IQR interquartile range, h hours, min minutes

^†or ^† Another 2 groups

^‡ The time recorded from the point of completion of the extraction to the time when the pain intensity was scored

^*Interquartile range

Sedation effect

A few studies specifically investigated the sedation side effects of DEX when used as an adjunct to local anesthesia. For instance, Etemadi et al. found that patients receiving DEX during third molar surgery felt significantly more relaxed than those in the control group [13]. Ouchi and Sugiyama found that increasing concentrations of DEX led to an increase in sedation scores [40]. However, it is essential to keep in mind that no participant in any of the studies reported experiencing profound sedation. The remaining studies acknowledged the positive properties of DEX as part of its mechanism of action but did not focus on this aspect as a primary outcome [19, 22].

Hemodynamic event

Most of the involved studies found no significant changes in heart rate, systolic blood pressure, diastolic blood pressure, or oxygen saturation when comparing a local anesthetic combined with DEX to a local anesthetic alone or a local anesthetic with adrenaline [2, 3, 19–24, 39–41]. While several studies noted a drop in either heart rate or blood pressure in the DEX group, these usually were slight and did not lead to hypotension or bradycardia [3, 4, 19, 20]. One study even noted that oxygen saturation was greater in the DEX group [19]. Regarding the meta-analysis of Intraoperative heart rate(30 min), among seven studies (350 patients) [21, 23, 24, 38, 39, 42, 43], the mean difference was 0.10 bpm (95% CI: −1.02 to 1.23), indicating no significant impact (p = 0.86). Heterogeneity was low (I² = 12%). Heart rate at 120 min, from three studies (150 patients) [21, 38, 39], the mean difference was 0.00 bpm (95% CI: −1.20 to 1.20), also showing no significant effect impact (p = 1.00). With no heterogeneity (I² = 0%). While heart rate at postoperative, from four studies (200 patients) [23, 24, 42, 43], the mean difference was 1.72 bpm (95% CI: −1.59 to 4.02), also showing no significant effect impact (p = 0.14), With low heterogeneity (I² = 20%) (Fig. 6).

Fig. 6.

Pooled heart rate (bpm) meta-analysis 95% confidence interval. Legend: Heart rate (dexmedetomidine vs control) Random-effects mean difference (bpm) shows no effect: 30 min 0.10 (95% CI: − 1.02 to 1.23; I² = 12%), 120 min 0.00 (95% CI: − 1.20 to 1.20; I² = 0%), postoperative 1.72 (95% CI: − 1.59 to 4.02; I² = 20%); all p > 0.05. Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

The meta-analysis evaluated the effect of DEX on systolic blood pressure intraoperatively (30 min), 120 min, and postoperatively. Intraoperative systolic pressure, among seven studies (350 patients)) [21, 23, 24, 38, 39, 42, 43], the mean difference was −0.13 mmHg (95% CI: −1.53 to 1.80), showing no significant impact (p = 0.88) with no heterogeneity (I² = 0%). Subgroup analysis of Systolic pressure at 120 min for three studies (150 patients) [21, 38, 39], the mean difference was −1.65 mmHg (95% CI: −0.68 to 3.98), also indicating no significant effect (p = 0.16) and no heterogeneity (I² = 0%). Whereas, postoperative systolic blood pressure, from four studies (200 patients) [23, 24, 42, 43], the mean difference was −2.15 mmHg (95% CI: −4.64 to 0.34), showing no significant effect impact (p = 0.09) with low heterogeneity (I² = 49%) (Fig. 7).

Fig. 7.

Systolic blood pressure (mmHg) meta-analysis with 95% confidence interval. Legend: Systolic blood pressure (dexmedetomidine vs control) No significant differences: MD (mmHg) at 30 min − 0.13 (95% CI: − 1.53 to 1.80; I² = 0%), at 120 min 1.65 (95% CI: − 0.68 to 3.98; I² = 0%), postoperative − 2.15 (95% CI: − 4.64 to 0.34; I² = 49%); all p > 0.05. Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

Our meta-analysis also clarified the impact of DEX on diastolic blood pressure during surgery (30 min), 120 min, and postoperatively. Intraoperative Diastolic Pressure, across seven studies (350 patients)) [21, 23, 24, 38, 39, 42, 43], the mean difference was −0.13 mmHg (95% CI: −1.53 to 1.80), with no significant effect (p = 0.88) and no heterogeneity (I² = 0%). Diastolic pressure at 120 min, among three studies (150 patients)) [21, 38, 39], the mean difference was 1.65 mmHg (95% CI: −0.68 to 3.98), again showing no significant impact (p = 0.16) and no heterogeneity (I² = 0%). In addition, in four studies (200 patients) at postoperative [23, 24, 42, 43], the mean difference was −2.15 mmHg (95% CI: −4.64 to 0.34), again showing no significant impact (p = 0.09) and low heterogeneity (I² = 49%) (Fig. 8). Furthermore, the oxygen saturation meta-analysis also performed with five studies included [23, 24, 39, 42, 43], the mean difference was −0.00 SpO₂ (95% CI: −0.11 to 0.10), with no significant effect (p = 0.94) with no heterogeneity (I² = 0%) (Fig. 9).

Fig. 8.

Diastolic blood pressure (mmHg) meta-analysis with 95% confidence interval. Legend: Diastolic blood pressure (dexmedetomidine vs control) No significant differences: MD (mmHg) at 30 min − 0.13 (95% CI: − 1.53 to 1.80; I² = 0%), at 120 min 1.65 (95% CI: − 0.68 to 3.98; I² = 0%), postoperative − 2.15 (95% CI: − 4.64 to 0.34; I² = 49%); all p > 0.05. Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

Fig. 9.

Oxygen saturation meta-analysis with 95% confidence interval. Legend: Oxygen saturation (SpO2, percentage points), dexmedetomidine vs control. No significant difference overall (MD 0.00; 95% CI − 0.11 to 0.10; 5 studies; I² = 0%). Abbreviations: MD, mean difference; CI, confidence interval; RE, random‑effects; I², heterogeneity; p, p‑value

In conclusion, While the pooled analyses showed no evidence of clinically significant hemodynamic changes, the total sample size (658 participants) provides insufficient statistical power to detect rare but potentially serious adverse events (e.g., severe bradycardia, hypotension requiring intervention, or respiratory compromise).

Sensitive analysis and meta-regression

Sensitivity analyses confirmed consistent, robust findings: switching between fixed- and random-effects models, or excluding any single study, did not materially change the pooled mean differences for onset time, anesthesia duration, or analgesia duration. Full results are provided in Additional File (D) —Fig. 1a–1b (onset time), 2a–2b (duration of anesthesia), and 3a–3b (analgesic duration).

In random‑effects meta‑regressions of studies, neither dexmedetomidine (DEX) dose nor anesthetic solution/concentration moderated effects on onset time or duration. For onset, dose (1 vs > 1) µg/mL and anesthetic type (non‑lidocaine vs lidocaine) were non‑significant in univariable and multivariable models (p ≥ 0.18), explained little variance (up to 25%; residual I² ≈ 92–96%), and the reference intercept (dose > 1 µg/mL with lidocaine) indicated a significantly faster onset (pooled MD − 67.0; p < 0.001). For duration, dose and lidocaine concentration (2% vs 1%) were also non‑significant (p ≥ 0.27; R² analog ≈0.22–0.29; residual I² ≈ 90–91%), with positive intercepts (~ 34–39 min). Overall, DEX shortened the onset and prolonged anesthesia, but the magnitude of these effects did not vary by the examined DEX or lidocaine dose categories. Full results are in Additional Ffile (D): Figs. 4–6 (onset) and 7–9 (duration).

Publication bias

Publication‑bias assessments for onset time and duration of anesthesia showed no material evidence of bias. Egger’s test was not significant for onset (p = 0.91) or duration (two‑tailed p = 0.077). Funnel plots were nearly symmetric for onset and showed only mild asymmetry for duration. Duval–Tweedie trim‑and‑fill imputed zero studies for both. Classic fail‑safe N values were large (onset 2,379; duration 986), supporting robustness. Detailed diagnostics are provided in Additional File (D) —Fig. 10a–10g (onset) and 11a–11g (duration).

Certainty of evidence (GRADE)

Certainty of evidence (GRADE) was moderate for primary efficacy outcomes: onset and duration of anesthesia, both downgraded for inconsistency (I² ≈ 95% and 94%, respectively), and duration of postoperative analgesia, downgraded for imprecision due to few studies/participants. Certainty was high for key safety outcomes, including heart rate, systolic blood pressure, diastolic blood pressure, and oxygen saturation, indicating consistent null effects. Certainty was low for postoperative pain at 12–24 h, and sedation to narrative synthesis, nonuniform measures, and sparse event data. Comprehensive GRADE certainty profiles are presented in Additional File (D) — (Table 1).

Discussion

Our systematic review and meta-analysis key findings include a reduction in anesthesia onset time, prolongation of anesthesia duration, and extension of postoperative analgesic effect. Patients receiving DEX showed better pain control with lower postoperative pain scores. From a safety perspective, hemodynamic analysis revealed no clinically significant changes in heart rate, blood pressure, or oxygen saturation, with only minor fluctuations that did not progress to hypotension or bradycardia. While some studies reported mild sedation effects, no profound sedation was observed in any participants. However, these findings are limited by high unexplained heterogeneity and a modest sample size.

Regarding anesthesia onset and duration, our findings support that DEX shortens the time to effective nerve blockade and extends the period of anesthesia. This makes the anesthetic more effective for prolonged surgeries and potentially decreases the need for additional injections, improving patient comfort. These findings agreed with other systematic reviews and meta-analyses describing DEX’s efficacy in many clinical practices. For instance, Hussain et al. investigated DEX, which significantly increased both the onset and duration of sensory and motor block durations in brachial plexus blocks [16]. Specifically, they observed a reduction in sensory block onset time by 3.19 min and a prolongation of sensory block duration by 261.41 minutes12. Similarly, motor block onset was reduced by 2.92 min, with duration prolonged by 200.9 min [16]. Further supporting the finding regarding the extended duration of anesthesia with DEX, Abdallah et al., in their systematic review and meta-analysis, found that dexmedetomidine used in neuraxial and peripheral nerve blocks prolonged sensory block duration by 150 min with intrathecal administration [12]. Another study focusing on DEX as an adjunct to local anesthetics in an erector spinae plane block (ESPB) observed a prolonged duration of anesthesia and reduced postoperative analgesic demand [17]. The studies mentioned above, and others focusing on upper extremity nerve blocks consistently demonstrate DEX’s benefits in enhancing onset times across different procedural contexts [5].

In terms of postoperative pain management, our meta-analysis showed a prolonged duration of analgesia and reduced postoperative analgesic demands with DEX. This also concurs with other studies in different clinical scenarios. For example, Wu et al., in their meta-analysis on neuraxial applications of DEX, reported significantly enhanced analgesic duration [44]. Similarly, a study by Zhao et al. on femoral nerve blocks further confirmed that DEX augmented postoperative pain control [11].

Our study also reported that from the safety perspective, no serious adverse changes in heart rate, blood pressure, and oxygen saturation were related to DEX. The minor fluctuations in heart rate or blood pressure occasionally did not progress to clinically significant hypotension or bradycardia, consistent with other meta-analyses where transient, easily managed bradycardia was noted [11, 16]. The low incidence of postoperative adverse events in dentistry in the current review compared to other systematic reviews and meta-analyses reporting some hypotension or bradycardia, may be explained by the low dose of DEX used in most of our included studies (1 μg/ml). This suggests that DEX is safe and effective for managing dental postoperative pain, especially in low dosages.

Limitations

The primary limitation is persistent, unexplained high heterogeneity for efficacy outcomes (I² = 94–99%) combined with modest sample size. Despite random effects modeling, subgroup analyses, meta-regression, and sensitivity analyses, heterogeneity sources remained unidentified, with no moderation by DEX dose, anesthetic agent, or concentration. Pooled estimates may substantially overestimate or underestimate true effects and should be interpreted as exploratory rather than definitive evidence. The safety analysis was underpowered to detect rare adverse events; larger prospective studies are needed.

Additional limitations include inconsistent outcome reporting, limited precision for pain outcomes, exclusion of studies with incomplete data, absence of sensitivity analysis for high-risk bias studies, and possible publication bias. Unmeasured factors including operator technique, injection site variability, and patient-specific characteristics may explain persistent heterogeneity. Future research requires standardized protocols with uniform DEX concentration, local anesthetic type, and outcome definitions to enable reliable clinical conclusions.

Clinical practice and research implications

Moderate‑certainty evidence indicates that adding dexmedetomidine (DEX) to dental local anesthetics probably shortens onset and prolongs block, with high‑certainty minimal hemodynamic effects. However, evidence for reduced postoperative analgesic consumption is inconsistent. Clinically, consider DEX selectively for longer or complex procedures rather than routine use; a low concentration (~ 1 μg/mL) is a pragmatic default since higher doses did not show added benefit in meta‑regression, and multimodal analgesia should be maintained. Research should prioritize adequately powered, preregistered trials that directly compare DEX doses, standardize co‑interventions and outcomes (including morphine‑equivalent consumption), and enroll diverse populations.

Conclusion

Dexmedetomidine as an adjunct to dental local anesthesia may reduce onset time and prolong anesthesia and analgesic duration, though substantial unexplained heterogeneity (I² = 94–99%) precludes confident estimation of effect magnitudes. The hemodynamic safety profile demonstrates consistent favorability. Current pooled evidence does not support routine clinical use; standardized, adequately powered trials with uniform protocols are essential before definitive recommendations can be established.

Abbreviations

bpm

Beats per minute

CI

Confidence interval

DEX

Dexmedetomidine

ESPB

Erector spinae plane block

FPS

Faces Pain Scale

GRADE

Grading of Recommendations Assessment, Development and Evaluation

I²

I-squared statistic (measure of heterogeneity)

IANB

Inferior alveolar nerve block

IONB

Infraorbital nerve block

MD

Mean difference

mL

Milliliter

mmHg

Millimeters of mercury

P

P value

PICOS

Population, Intervention, Comparison, Outcome, Study design

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta-Analyses

PROSPERO

International Prospective Register of Systematic Reviews

Qm

Omnibus Q statistic for moderators in meta-regression (test of moderators)

R²

Coefficient of determination; in meta-analysis, proportion of heterogeneity explained by moderators

RCTs

Randomized controlled trials

REML

Restricted maximum likelihood

RevMan

Review Manager (Cochrane Collaboration software)

ROBINS-I

Risk Of Bias In Non-randomized Studies of Interventions

RoB 2

Revised Cochrane Risk of Bias tool for randomized trials

SpO2

Peripheral capillary oxygen saturation

VAS

Visual Analog Scale

Z

Z value

µg

Microgram

µg/mL

Micrograms per milliliter

Authors’ contributions

SAA, WAA, TW, and NL contributed to the conceptualization and study design. SAA and WAA collected and extracted the data. SAA, WAA and LN performed data analysis and interpretation. All authors SAA, WAA, NL, TW, WW, MW, and YQH were involved in the writing or revision of the manuscript and approved the final version.

Funding

This work was supported by the Hunan Provincial Natural Science Foundation of China (Grant No. 2025JJ50539), the National Natural Sciences Foundation of China (Grant No. 82170974) and the Central South University Research Program of Advanced Interdisciplinary Studies (Grant No. 2023QYJC038).

Data availability

The original data presented in this article are available upon reasonable request from the corresponding author.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

All authors have given consent for publication.

Competing interests

There is no competing interest to declare.