Image

Annali di Stomatologia | 2026; 17(3): 641-654

ISSN 1971-1441 | DOI: 10.59987/ads/2026.3.641-654

Articles

Antibiotic use in third molar extraction: a systematic review and structured narrative synthesis

Gianluca Benincasa1, Ettore Lupi1, Alessandro Orsijena1, Alfredo De Rosa2, Marianna Chiacchia1, Giulia Caporro1, Antonio Capogreco1, Giulia Ciciarelli1

1Department of Life, Health and Environmental Sciences, University of L’Aquila, L’Aquila, Italy.

2Department of Life Sciences, Health and Health Professions, Link Campus University, Rome, Italy.

*Corresponding author: Gianluca Benincasa - gianluca.benincasa@student.univaq.it

Article History

Received: May 29, 2026

Accepted: July 19, 2026

Published: July 30, 2026

Abstract

Background

Antibiotic prescription following third molar extraction remains controversial because of variable clinical practices, uncertain indications, and concerns regarding antimicrobial resistance. This systematic review aimed to evaluate current evidence regarding the effectiveness and appropriateness of antibiotic use in healthy patients undergoing third molar extraction.

Methods

The review was prepared with reference to the Cochrane Handbook and PRISMA 2020. A PROSPERO identifier was reported in the submitted documentation (CRD420261405197). MEDLINE via PubMed, the Cochrane Library, and Google Scholar were searched for eligible studies published from 1 January 2016 to 31 March 2026. Thirteen studies were included and grouped by clinically distinct comparisons. Because interventions, comparators, designs, and outcome definitions were heterogeneous, findings were synthesized descriptively without meta-analysis. Formal GRADE assessment was not undertaken.

Results

The evidence comprised randomized and observational studies evaluating systemic prophylaxis or postoperative therapy, local antibiotic delivery, antibiotic-dose comparisons, and antibiotics versus non-antibiotic local strategies. Individual studies produced inconsistent and often imprecise estimates. Selected studies reported lower incidences of alveolar osteitis with local rifamycin (2/35 vs 6/35 sockets), an oxytetracycline-impregnated drain (5/100 vs 23/100 sockets), and adjunctive oxytetracycline-hydrocortisone gauze (2/215 vs 9/46 participants). The latter comparison was observational and all participants also received systemic amoxicillin. A placebo-controlled trial of preoperative amoxicillin reported postoperative infection in 2/77 versus 5/77 participants. Interpretation was limited by clinical heterogeneity, sparse events, incomplete methodological documentation, and the absence of a reproducible formal risk-of-bias assessment.

Conclusions

The included studies do not support indiscriminate antibiotic prescribing for every healthy patient undergoing uncomplicated third molar surgery, but they also do not establish absence of preventive benefit. Previous evidence syntheses suggest possible modest reductions in postoperative infection and alveolar osteitis, with low certainty and relatively high numbers needed to treat. Decisions should be individualized according to patient and procedural risk, adverse effects, and antimicrobial-stewardship principles.

Keywords: Molar, Third; Tooth Extraction; Antibiotic Prophylaxis; Alveolar Osteitis; Anti-microbial Stewardship.

1. Introduction

Surgical extraction of third molars, or wisdom teeth, is one of the most commonly performed procedures in dentistry and oral and maxillofacial surgery [1]. Third molars frequently remain impacted or partially impacted because of insufficient space in the dental arch or alterations in the eruption axis [1]. Their anatomical position, often associated with a partially erupted gingival flap, facilitates plaque accumulation and recurrent local infection, including pericoronitis [2].

Surgical removal, especially in cases of partial or complete bony impaction, may require elevation of a mucoperiosteal flap and osteotomy. This surgical trauma disrupts the local oral microbiota and exposes deeper tissues and blood vessels to opportunistic oral pathogens [1]. Consequently, the postoperative course is physiologically characterized by inflammatory symptoms, including pain, edema, and trismus, and may be complicated by superimposed infection [1].

Antibiotics have historically been widely prescribed in third molar surgery, although indications and prescribing patterns remain heterogeneous. Broad-spectrum agents — predominantly penicillins such as amoxicillin, sometimes combined with clavulanic acid — are used as perioperative prophylaxis or as postoperative courses lasting several days [3–5].

However, indiscriminate or defensive prescribing contributes to antimicrobial resistance (AMR), which is a major global public-health threat [6]. This context requires critical reassessment of the risk–benefit balance of antibiotic therapy and restriction of its use to evidence-based indications [3,5,7].

Among the most frequent and disabling local complications is alveolar osteitis (dry socket), a painful condition characterized by premature lysis or loss of the blood clot within the extraction socket, with exposure of the alveolar bone and nerve endings [8]. Its etiology is multifactorial and includes surgical trauma, smoking, oral contraceptive use, and impaired local hygiene; bacterial fibrinolytic activity may also contribute [8].

For decades, prevention of alveolar osteitis has been used to justify routine antibiotic prescription after third molar extraction. However, systemic antimicrobials do not eliminate the risk, and local prophylactic measures such as chlorhexidine gels or rinses may provide benefit without systemic drug exposure [5,9,28].

When assessing the appropriateness of antibiotic therapy, it is essential to distinguish prevention of local postoperative complications from prophylaxis against infective endocarditis. Infective endocarditis is a serious, potentially life-threatening condition caused by microbial colonization of the endocardium after bacteremia, which may occur following invasive dental procedures [10].

The American Heart Association recommends antibiotic prophylaxis before relevant dental procedures only for selected patients with cardiac conditions associated with the highest risk of adverse outcomes from infective endocarditis, such as prosthetic cardiac valves or a previous history of infective endocarditis [10]. In healthy patients without these conditions, antibiotic prescription should be evaluated on the basis of local surgical and infectious risk rather than endocarditis prevention [10].

Given the heterogeneity of antibiotic prescribing practices and increasing concerns regarding antimicrobial resistance, this systematic review aims to critically evaluate current guidelines and recent clinical evidence on antibiotic use in third molar extraction. The objective is to assess the actual effectiveness of antibiotic therapy in preventing postoperative infection and alveolar osteitis in healthy patients and to support evidence-based clinical decision-making.

2. Methods

The submitted documentation identified a PROSPERO registration number (CRD420261405197). The public record and timing of registration could not be independently verified during editorial finalization. The review was prepared with reference to the Cochrane Handbook for Systematic Reviews of Interventions [11].

Primary Objective

To evaluate the available evidence regarding the use of antibiotic therapy in patients undergoing third molar extraction.

Secondary Objectives

To investigate:

  1. clinical indications for antibiotic prescription;
  2. contraindications and safety considerations;
  3. antibiotics most commonly prescribed;
  4. dosage regimens;
  5. treatment duration;
  6. differences between simple and complex third molar extractions;
  7. effects on postoperative outcomes.

Research Question (PICO Framework)

Population (P)

Patients undergoing maxillary or mandibular third molar extraction, including erupted, partially impacted, and impacted teeth. The intended population was predominantly healthy patients or patients classified as ASA I–II; however, active local pathology was not consistently excluded across the included studies.

Intervention (I)

Antibiotic therapy including:

  • Preoperative prophylaxis;
  • Postoperative therapeutic administration;
  • Combined perioperative protocols.

Eligible interventions included systemic antibiotic prophylaxis, postoperative systemic therapy, and local antibiotic delivery. Non-antibiotic local interventions were eligible only when directly compared with an antibiotic regimen. For synthesis, studies were separated into: (1) systemic antibiotics versus placebo/no antibiotics; (2) systemic antibiotic regimen or dose comparisons; (3) local antibiotics versus local control; and (4) antibiotics versus non-antibiotic local strategies.

Comparison (C)

  • No antibiotic treatment;
  • Placebo;
  • Alternative antibiotic regimens;
  • Different doses and treatment durations.

Outcomes (O)

  • Primary Outcomes
    • postoperative infection;
    • alveolar osteitis (dry socket);
    • wound healing.
  • Secondary Outcomes
    • postoperative pain;
    • swelling;
    • trismus;
    • adverse events;
    • antimicrobial resistance concerns;
    • need for additional treatment or reintervention.

Search Strategy

The review was reported with reference to PRISMA 2020 [12] and the PRISMA-S extension for literature searches [29]. Searches were reported as performed in MEDLINE via PubMed, the Cochrane Library, and Google Scholar between November 2025 and May 2026. Eligibility was restricted to studies published from 1 January 2016 to 31 March 2026; the later search period could therefore retrieve records indexed after the publication cut-off. Exact source-specific search dates were not preserved in the available documentation.

The available documentation preserved the following PubMed strategy: ((“third molar” [Title/Abstract] OR “wisdom tooth” [Title/ Abstract] OR “third molar extraction” [Title/Abstract] OR “third molar surgery” [Title/Abstract]) AND (antibiotic* [Title/Abstract] OR “Anti-Bacterial Agents” [MeSH Terms] OR “antibiotic prophylaxis” [Title/Abstract] OR “antimicrobial prophylaxis” [Title/Abstract]) AND (extraction [Title/Abstract] OR surgery [Title/Abstract] OR postoperative [Title/Abstract])) AND (“2016/01/01” [Date - Publication]: “2026/03/31” [Date - Publication]). Complete source-specific strategies for the Cochrane Library and Google Scholar, including Google Scholar screening depth, ranking method, and stopping rule, were not retained. The search is therefore only partially reproducible.

Eligibility Criteria

Inclusion Criteria

  • studies involving human participants;
  • patients undergoing third molar extraction;
  • studies reporting antibiotic protocols;
  • randomized controlled trials (RCTs);
  • prospective studies;
  • retrospective studies;
  • systematic reviews used only for citation chasing and not included in the primary synthesis;
  • English main text;
  • studies published from 1 January 2016 to 31 March 2026;

Exclusion Criteria

  • case reports;
  • case series with <10 participants;
  • procedures unrelated to third molar extraction;
  • studies without clear information regarding antibiotic administration.

Study Selection and Data Extraction

Study selection was performed by two independent reviewers (G.B. and G.C.) through a two-stage screening process:

  • title and abstract screening;
  • full-text assessment of potentially eligible studies.

Disagreements during screening were resolved through discussion, with consultation of a third reviewer (A.O.) when consensus could not be reached. The record-management software, duplicate-detection procedure, and inter-reviewer agreement statistic were not documented in the available review files.

Data were extracted using a standardized data-collection form. Independent duplicate extraction, piloting of the form, and procedures for resolving extraction discrepancies were not documented. The following information was extracted:

  • Study characteristics (author, year of publication, country, study design);
  • Population characteristics (sample size, age, health status);
  • Type of third molar extraction (simple/non-surgical or complex/surgical);
  • Clinical indication for antibiotic administration;
  • Type of antibiotic prescribed;
  • Route and timing of administration (preoperative, postoperative, perioperative);
  • Dosage regimen;
  • Duration of antibiotic therapy;
  • Comparator (placebo, no antibiotic treatment, alternative regimen);
  • Clinical outcomes, including:
  • postoperative infection;
  • alveolar osteitis (dry socket);
  • pain;
  • swelling;
  • trismus;
  • adverse events;
  • need for additional intervention.

Methodological Appraisal

A formal risk-of-bias assessment could not be reported because domain-level judgments and supporting rationales were unavailable and the compiled plots did not consistently apply design-appropriate tools. In particular, split-mouth trials require the cross-over-trial version of RoB 2, whereas non-randomized intervention studies require ROBINS-I [13–14]. Unsupported traffic-light plots were therefore not retained.

Methodological limitations were considered qualitatively during synthesis, with attention to randomization and blinding, missing data, outcome measurement and reporting, sample size, sparse events, and confounding in observational comparisons. These considerations were not converted into formal risk-of-bias categories.

Data Synthesis and Statistical Analysis

A structured synthesis without meta-analysis was undertaken following SWiM reporting principles [30]. Studies were grouped by clinically coherent comparison and outcome. Dichotomous outcomes were summarized using event counts and risk ratios with 95% confidence intervals when these could be calculated from the reported data. Continuous outcomes were summarized narratively because scales, time points, and dispersion measures were inconsistent or incompletely reported. No statistical weighting or vote counting by significance was used.

Meta-analysis was not undertaken because few studies addressed the same intervention, comparator, outcome definition, and time point, and several reports did not provide compatible effect estimates or variance data. The certainty of evidence was not formally assessed using GRADE; conclusions were deliberately calibrated to avoid claims of equivalence, absence of effect, or high certainty.

3. Results

Search results and selection of included studies

A total of 276 records were identified through searches of MEDLINE (PubMed), the Cochrane Library, and Google Scholar. After removal of 53 records before screening, 223 records remained for title and abstract screening; 190 were excluded. Thirty-three reports were sought for retrieval, of which 2 were not retrieved. Thirty-one full-text reports were assessed for eligibility. Eighteen reports were excluded: 2 were not published in English, 14 did not include an eligible antibiotic intervention, and 2 did not evaluate antibiotic therapy or prophylaxis. Thirteen studies were included. The selection process is shown in Figure 1 [12].

Characteristics of Included Studies and Study Design

The 13 included studies were published between 2017 and 2025 [15–27]. Eleven were reported as randomized trials and two as observational studies. The evidence addressed four distinct questions: systemic antibiotics versus placebo/no antibiotics; comparisons among systemic regimens; local antibiotics versus local control; and antibiotics versus non-antibiotic local strategies. These evidence streams were not treated as a single homogeneous comparison. Detailed characteristics are presented in Table 1.

Methodological Appraisal

No formal study-level or outcome-level risk-of-bias judgments are reported because the necessary domain-level documentation could not be reconstructed from the available review files. Across the evidence base, interpretation was limited by small samples, sparse infectious events, incomplete reporting of randomization or blinding in some reports, heterogeneous outcome definitions and follow-up schedules, and confounding in observational comparisons.

Outcomes

The predefined primary outcomes were postoperative infection, alveolar osteitis, and wound healing. Postoperative infection and alveolar osteitis were analyzed separately because alveolar osteitis is not synonymous with infection. Wound healing was directly evaluated in only a minority of studies. Secondary outcomes included pain, swelling, trismus, rescue medication use, quality of life, adverse events, and patient satisfaction.

Outcome definitions and assessment schedules varied substantially [15–27]. Several studies focused principally on symptoms rather than infectious events, limiting their contribution to the primary clinical question.

Systemic antibiotics versus placebo or no antibiotics

Yanine et al. [19] reported postoperative infection in 2/77 patients receiving a single preoperative dose of amoxicillin and 5/77 receiving placebo (risk ratio [RR] 0.40, 95% confidence interval [CI] 0.08–2.00). The estimate was imprecise and compatible with substantial benefit or no benefit. Kirnbauer et al. [17] reported an overall surgical-site infection frequency of 11% and no statistically significant paired difference between perioperative amoxicillin and placebo. Mariscal-Cazalla et al. [24] reported lower pain and inflammation and less rescue medication with antibiotic regimens than placebo, but the abstract did not provide sufficient event data for an infection effect estimate. Rabi et al. [22] evaluated postoperative antimicrobial regimens primarily through pain and mouth opening and did not demonstrate a consistent clinically important advantage.

Local antibiotic interventions

Cigerim et al. [16] reported alveolitis in 2/35 rifamycin-treated sockets and 6/35 control sockets (RR 0.33, 95% CI 0.07–1.54). Øyri et al. [23] reported alveolar osteitis in 5/100 sockets treated with an oxytetracycline-impregnated drain and 23/100 control sockets (RR 0.22, 95% CI 0.09–0.55). Otake et al. [25] reported dry socket in 2/215 patients receiving adjunctive oxytetracycline–hydrocortisone gauze and 9/46 controls (RR 0.05, 95% CI 0.01–0.21); however, this was a retrospective observational comparison and all patients received systemic amoxicillin, so it does not establish that local therapy can replace systemic antibiotics.

Regimen comparisons and non-antibiotic comparators

Braimah et al. [21] and Janas-Naze et al. [26] compared antibiotic schedules or doses and therefore inform regimen selection rather than whether antibiotics are needed. Donmezer and Bilginaylar [20] compared local and systemic antibiotics combined with platelet-rich fibrin, but did not establish equivalence or non-inferiority. Chugh et al. [18], Pereira et al. [15], and Sivalingam et al. [27] compared antibiotics with chlorhexidine, photobiomodulation, or ozone; these studies address alternative strategies and cannot be interpreted as placebo-controlled evidence for or against prophylactic antibiotics.

Certainty and applicability

Because a reproducible formal risk-of-bias assessment and GRADE evaluation were not available, no high-certainty inference can be drawn from the included studies. The evidence is most applicable to young, generally healthy patients undergoing elective mandibular third molar surgery. It cannot be directly extrapolated to immunocompromised or medically complex patients, or to patients with established spreading odontogenic infection.

4. Discussion

This systematic review evaluated heterogeneous studies concerning antibiotic use in third molar surgery. The evidence base included prophylactic and postoperative systemic regimens, local antibiotic applications, dose-comparison studies, and trials in which antibiotics were compared with non-antibiotic local strategies. These comparisons address different clinical questions and should be interpreted separately. The present synthesis therefore supports caution against indiscriminate prescribing but cannot by itself establish that antibiotics provide no benefit in all healthy patients or procedures.

Systemic prophylaxis against placebo or no antibiotic provides the most direct evidence for the primary clinical question. Studies comparing two antibiotic doses or durations inform regimen optimization rather than the need for antibiotics. Studies comparing systemic antibiotics with chlorhexidine, photobiomodulation, ozone, or local delivery evaluate alternative strategies and should not be pooled conceptually with placebo-controlled prophylaxis trials.

Several individual trials did not report statistically significant reductions in infection or postoperative morbidity with systemic antibiotics [15,17,19,22,24]. Such nonsignificant findings, especially with uncommon events and small samples, do not establish absence of effect. The 2021 Cochrane review found low-certainty evidence that prophylactic antibiotics may reduce infection and alveolar osteitis, with approximately 19 patients treated to prevent one infection and 46 treated to prevent one dry socket [5]. A 2024 network meta-analysis similarly favored antibiotics for surgical-site infection and dry socket, with numbers needed to treat of 18 and 25, respectively [31]. These absolute benefits are modest and must be balanced against adverse effects and antimicrobial resistance. Some local strategies reported favorable outcomes for alveolar osteitis or postoperative discomfort [16,18,20,23,25,27]. These findings should be described as preliminary because many derive from single-center studies, small samples, subjective outcomes, or comparisons not designed for equivalence/non-inferiority. In Otake et al. [25], all patients also received systemic amoxicillin; the local intervention was adjunctive and cannot be presented as a systemic-antibiotic-sparing alternative. For Chugh et al. [18] and Donmezer and Bilginaylar [20], absence of statistically significant differences does not demonstrate equivalence unless a prespecified non-inferiority/equivalence margin and appropriate analysis were used.

The clinically defensible interpretation is that routine, indiscriminate antibiotic use is not justified solely by habit, while selective prophylaxis may confer a modest absolute benefit in some settings. This balance is consistent with antimicrobial-stewardship principles [2–6]. Infective-endocarditis prophylaxis is a separate indication restricted to selected high-risk cardiac conditions [10] and should not be conflated with prevention of local postoperative complications.

This review has important limitations. The search is only partially reproducible because complete source-specific strategies and exact execution dates were not retained. The eligibility framework combines distinct interventions, comparators, and patient populations; several studies emphasize symptoms rather than infectious events; outcome definitions and time points vary; compatible effect estimates are frequently unavailable; and no reproducible formal risk-of-bias or GRADE assessment could be completed from the submitted files. These limitations materially reduce confidence in definitive clinical statements and preclude claims of equivalence or absence of effect.

5. Conclusions

The evidence included in this review is insufficient to support indiscriminate routine antibiotic administration for every healthy patient undergoing uncomplicated third molar surgery. The evidence is also insufficient to conclude that antibiotics have no preventive effect. Existing systematic reviews suggest that prophylactic antibiotics may reduce postoperative infection and alveolar osteitis, but the certainty is low and the absolute benefit is modest, so many patients would need treatment to prevent one event.

Clinical decisions should therefore be individualized according to baseline infectious risk, surgical complexity, active local infection, relevant comorbidities, adverse-effect risk, and antimicrobial-stewardship considerations. In healthy patients undergoing uncomplicated elective surgery, the expected absolute preventive benefit appears modest; routine prescribing should not be automatic, while selective use may be reasonable when individual risk is higher.

Declarations

Protocol registration and deviations: The submitted documentation identified PROSPERO record CRD420261405197 and indicated that meta-analysis had originally been planned. The public record, registration timing, and protocol amendments could not be independently verified. Meta-analysis was not performed because the included studies were not sufficiently homogeneous and frequently lacked compatible effect data.

Author contributions

An author-specific CRediT contribution statement was not included in the submitted documentation.

Conflicts of interest and funding

Conflict-of-interest and funding declarations were not included in the submitted documentation.

Data availability:

No supplementary search logs, screening files, extraction dataset, or domain-level risk-of-bias judgments were supplied with the manuscript. The study characteristics and outcome data available for this review are summarized in Tables 1 and 2.

References

  • 1. Sayed N, Bakathir A, Pasha M, Al-Sudairy S. Complications of third molar extraction: a retrospective study from a tertiary healthcare centre in Oman. Sultan Qaboos Univ Med J. 2019;19(3):e230–e235. doi:10.18295/squmj.2019.19.03.009. https://doi.org/10.18295/squmj.2019.19.03.009
  • 2. Palmer N, editor. Antimicrobial Prescribing in Dentistry: Good Practice Guidelines. 3rd ed. London: Faculty of General Dental Practice (UK) and Faculty of Dental Surgery; 2020. ISBN: 978-1-8381964-2-4.
  • 3. Stein K, Farmer J, Singhal S, Marra F, Sutherland S, Quiñonez C. The use and misuse of antibiotics in dentistry: a scoping review. J Am Dent Assoc. 2018;149(10):869–884.e5. doi:10.1016/j.adaj.2018.05.034. https://doi.org/10.1016/j.adaj.2018.05.034
  • 4. Cervino G, Cicciù M, Biondi A, Bocchieri S, Herford AS, Laino L, et al. Antibiotic prophylaxis on third molar extraction: systematic review of recent data. Antibiotics (Basel). 2019;8(2):53. doi:10.3390/antibiotics8020053. https://doi.org/10.3390/antibiotics8020053
  • 5. Lodi G, Azzi L, Varoni EM, Pentenero M, Del Fabbro M, Carrassi A, et al. Antibiotics to prevent complications following tooth extractions. Cochrane Database Syst Rev. 2021;2(2):CD003811. doi:10.1002/14651858.CD003811.pub3. https://doi.org/10.1002/14651858.CD003811.pub3
  • 6. World Health Organization. Global antibiotic resistance surveillance report 2025. Geneva: World Health Organization; 2025. ISBN: 978-92-4-011633-7.
  • 7. Spellberg B, Bartlett JG, Gilbert DN. The future of antibiotics and resistance. N Engl J Med. 2013;368(4):299–302. doi:10.1056/NEJMp1215093. https://doi.org/10.1056/NEJMp1215093
  • 8. Mamoun J. Dry socket etiology, diagnosis, and clinical treatment techniques. J Korean Assoc Oral Maxillofac Surg. 2018;44(2):52–58. doi:10.5125/jkaoms.2018.44.2.52. https://doi.org/10.5125/jkaoms.2018.44.2.52
  • 9. Rodríguez Sánchez F, Rodríguez Andrés C, Arteagoitia Calvo I. Does chlorhexidine prevent alveolar osteitis after third molar extractions? Systematic review and meta-analysis. J Oral Maxillofac Surg. 2017;75(5):901–914. doi:10.1016/j.joms.2017.01.002. https://doi.org/10.1016/j.joms.2017.01.002
  • 10. Wilson WR, Gewitz M, Lockhart PB, Bolger AF, DeSimone DC, Kazi DS, et al. Prevention of viridans group streptococcal infective endocarditis: a scientific statement from the American Heart Association. Circulation. 2021;143(20):e963–e978. doi:10.1161/CIR.0000000000000969. https://doi.org/10.1161/CIR.0000000000000969
  • 11. Cumpston M, Li T, Page MJ, Chandler J, Welch VA, Higgins JPT, et al. Updated guidance for trusted systematic reviews: a new edition of the Cochrane Handbook for Systematic Reviews of Interventions. Cochrane Database Syst Rev. 2019;10(10):ED000142. doi:10.1002/14651858.ED000142. https://doi.org/10.1002/14651858.ED000142
  • 12. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71. doi:10.1136/bmj.n71. https://doi.org/10.1136/bmj.n71
  • 13. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019;366:l4898. doi:10.1136/bmj.l4898. https://doi.org/10.1136/bmj.l4898
  • 14. Sterne JAC, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions. BMJ. 2016;355:i4919. doi:10.1136/bmj.i4919. https://doi.org/10.1136/bmj.i4919
  • 15. Alves Pereira D, Silva Bonatto M, Souza Santos S, Gomes Junqueira Mendes P, Sales E Pessoa R, Pimentel Lopes de Oliveira GJ. Comparison of the effects of antibiotic therapy and photobiomodulation with red and infrared lasers on the healing of postextraction sockets of third molars: a randomised controlled trial. Lasers Med Sci. 2025;40(1):196. doi:10.1007/s10103-025-04445-4. https://doi.org/10.1007/s10103-025-04445-4
  • 16. Cigerim L, Orhan ZD, Kaplan V, Cigerim SC, Feslihan E. Evaluation of the efficacy of topical rifamycin application on postoperative complications after lower impacted wisdom teeth surgery. J Stomatol Oral Maxillofac Surg. 2024;125(4 Suppl):101501. doi:10.1016/j.jormas.2023.101501. https://doi.org/10.1016/j.jormas.2023.101501
  • 17. Kirnbauer B, Jakse N, Truschnegg A, Dzidic I, Mukaddam K, Payer M. Is perioperative antibiotic prophylaxis in the case of routine surgical removal of the third molar still justified? A randomized, double-blind, placebo-controlled clinical trial with a split-mouth design. Clin Oral Investig. 2022;26(10):6409–6421. doi:10.1007/s00784-022-04597-5. https://doi.org/10.1007/s00784-022-04597-5
  • 18. Chugh A, P.G G, Kumar P, Kaur A. Localised 0.2% chlorhexidine irrigation delivery system versus oral antibiotics in reducing postoperative complications in the surgical extraction of impacted mandibular third molar: a randomised controlled trial. Med Oral Patol Oral Cir Bucal. 2024;29(5):e690–e697. doi:10.4317/medoral.26676. https://doi.org/10.4317/medoral.26676
  • 19. Yanine N, Sabelle N, Vergara-Gárate V, Salazar J, Araya-Cabello I, Carrasco-Labra A, et al. Effect of antibiotic prophylaxis for preventing infectious complications following impacted mandibular third molar surgery: a randomized controlled trial. Med Oral Patol Oral Cir Bucal. 2021;26(6):e703–e710. doi:10.4317/medoral.24274. https://doi.org/10.4317/medoral.24274
  • 20. Donmezer CM, Bilginaylar K. Comparison of the postoperative effects of local antibiotic versus systemic antibiotic with the use of platelet-rich fibrin on impacted mandibular third molar surgery: a randomized split-mouth study. Biomed Res Int. 2021;2021:3040661. doi:10.1155/2021/3040661. https://doi.org/10.1155/2021/3040661
  • 21. Braimah RO, Ndukwe KC, Owotade JF, Aregbesola SB. Impact of oral antibiotics on health-related quality of life after mandibular third molar surgery: an observational study. Niger J Clin Pract. 2017;20(9):1189–1194. doi:10.4103/1119-3077.183235. https://doi.org/10.4103/1119-3077.183235
  • 22. Rabi A, Maheshwari R, Srinivasan B, Warad LP, Suvarna CC, Tank KS. Effectiveness of antimicrobial therapy after extraction of impacted mandibular third molar: a randomized clinical trial. J Contemp Dent Pract. 2018;19(1):81–85. doi:10.5005/jp-journals-10024-2215. https://doi.org/10.5005/jp-journals-10024-2215
  • 23. Øyri H, Jonsdottir O, Jensen JL, Bjørnland T. The use of a tetracycline drain reduces alveolar osteitis: a randomized prospective trial of third molar surgery under local anesthetics and without the use of systemic antibiotics. Oral Surg Oral Med Oral Pathol Oral Radiol. 2019;128(3):205–212. doi:10.1016/j.oooo.2018.12.004. https://doi.org/10.1016/j.oooo.2018.12.004
  • 24. Mariscal-Cazalla MDM, Manzano-Moreno FJ, García-Vázquez M, Vallecillo-Capilla MF, Olmedo-Gaya MV. Do perioperative antibiotics reduce complications of mandibular third molar removal? A double-blind randomized controlled clinical trial. Oral Surg Oral Med Oral Pathol Oral Radiol. 2021;131(3):286–294. doi:10.1016/j.oooo.2020.08.034. https://doi.org/10.1016/j.oooo.2020.08.034
  • 25. Otake H, Sato Y, Nakatani E, Hawke P, Takei S, Ogino A, et al. Oxytetracycline-hydrocortisone ointment reduces the occurrence of both dry socket and post-extraction pain after third molar extraction: an observational study. PLoS One. 2021;16(7):e0254221. doi:10.1371/journal.pone.0254221. https://doi.org/10.1371/journal.pone.0254221
  • 26. Janas-Naze A, Torbicka G, Chybicki D, Lipczyńska-Lewandowska M, Zhang W. Comparative efficacy of different oral doses of clindamycin in preventing post-operative sequelae of lower third molar surgery: a randomized, triple-blind study. Medicina (Kaunas). 2022;58(5):668. doi:10.3390/medicina58050668. https://doi.org/10.3390/medicina58050668
  • 27. Sivalingam VP, Panneerselvam E, Raja KV, Gopi G. Does topical ozone therapy improve patient comfort after surgical removal of impacted mandibular third molar? A randomized controlled trial. J Oral Maxillofac Surg. 2017;75(1):51.e1–51.e9. doi:10.1016/j.joms.2016.09.014. https://doi.org/10.1016/j.joms.2016.09.014
  • 28. Ramos E, Santamaría J, Santamaría G, Barbier L, Arteagoitia I. Do systemic antibiotics prevent dry socket and infection after third molar extraction? A systematic review and meta-analysis. Oral Surg Oral Med Oral Pathol Oral Radiol. 2016;122(4):403–425. doi:10.1016/j.oooo.2016.04.016. https://doi.org/10.1016/j.oooo.2016.04.016
  • 29. Rethlefsen ML, Kirtley S, Waffenschmidt S, Ayala AP, Moher D, Page MJ, et al. PRISMA-S: an extension to the PRISMA Statement for reporting literature searches in systematic reviews. Syst Rev. 2021;10(1):39. doi:10.1186/s13643-020-01542-z. https://doi.org/10.1186/s13643-020-01542-z
  • 30. Campbell M, McKenzie JE, Sowden A, Katikireddi SV, Brennan SE, Ellis S, et al. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ. 2020;368:l6890. doi:10.1136/bmj.l6890. https://doi.org/10.1136/bmj.l6890
  • 31. Camps-Font O, Sábado-Bundó H, Toledano-Serrabona J, Valmaseda-de-la-Rosa N, Figueiredo R, Valmaseda-Castellón E. Antibiotic prophylaxis in the prevention of dry socket and surgical site infection after lower third molar extraction: a network meta-analysis. Int J Oral Maxillofac Surg. 2024;53(1):57–67. doi:10.1016/j.ijom.2023.08.001. https://doi.org/10.1016/j.ijom.2023.08.001
image
Figure 1. PRISMA 2020 flow diagram for identification, screening, eligibility assessment, and inclusion of studies [12].

Tables

Table 1. Characteristics of the included studies and authors’ reported conclusions.
Name of the study (year) Country Study design Population Groups Antibiotic regimen Follow-up Outcomes Authors’ reported conclusions
Test Control
Pereira et al. (2025) Brazil Randomized controlled clinical trial 60 patients undergoing extraction of all four third molars ATB3: amoxicillin × 3 days; ATB7: amoxicillin × 7 days; PBMT Control: analgesics + anti-inflammatory only Amoxicillin 500 mg PO q8h × 3 days (ATB3); Amoxicillin 500 mg PO q8h × 7 days (ATB7) POD 3, 7, 14, 30, 90 Pain, oedema, mouth opening, soft tissue healing, bone density No clinical benefit of postoperative antibiotics; PBMT improved postoperative recovery
Cigerim et al. (2024) Turkey Prospective randomized double-blind split-mouth controlled clinical study 35 healthy patients (18–35 years) undergoing bilateral impacted lower third molar extraction (split-mouth design) Topical rifamycin application Saline irrigation (control) Rifamycin topical irrigation: 3 mL / 250 mg intraoperative application (including 1.2 mL/100 mg into extraction socket for 5 min + 1.8 mL/150 mg on flap and exposed bone); preoperative rinse with 3 mL/250 mg POD 2 and 7 (pain recorded daily for 7 days) Alveolar osteitis (dry socket), postoperative infection, pain (VAS), trismus, edema, rescue analgesic use Topical rifamycin reduced postoperative pain and showed lower incidence of alveolitis; no significant differences for edema or trismus. Authors proposed topical rifamycin as a possible alternative to systemic antibiotics
Kirnbauer et al. (2022) Austria Randomized, double-blind, placebo-controlled split-mouth clinical trial 50 healthy patients (≥16 years; ASA I) undergoing routine surgical removal of four impacted or partially impacted non-inflamed third molars (100 interventions) Perioperative antibiotic prophylaxis Placebo Amoxicillin 2 g PO 1 h preoperatively + 1.5 g/ day PO (500 mg q8h) for 3 postoperative days POD 1 and 7 (minimum 3-month interval between surgeries) Surgical site infection (SSI), swelling (analog and digital), trismus, bleeding, pain, analgesic intake Perioperative antibiotic prophylaxis did not significantly reduce postoperative complications or improve patient-reported outcomes compared with placebo; routine use in non-inflamed third molar surgery appears unnecessary
Chugh et al. (2024) India Single-center, double-blind, prospective randomized controlled trial (parallel-group) 84 patients, ASA I–II, 18–50 years, undergoing surgical extraction of impacted mandibular third molars Localized chlorhexidine irrigation (0.2%) via curved-tip delivery system without systemic antibiotics Standard systemic oral antibiotic protocol Control: Amoxicillin 500 mg + clavulanic acid 125 mg PO 1 h preoperatively + amoxicillin/clavulanic acid 500/125 mg PO TID for 3 postoperative days. Test: No systemic antibiotics; chlorhexidine 0.2% local irrigation twice daily postoperatively POD 1, 3, 5, and 7 Pain, trismus (mouth opening), swelling, infection, rescue analgesic use, additional antibiotic use, dry socket (alveolar osteitis), wound dehiscence, adverse events, patient satisfaction Localized chlorhexidine achieved outcomes comparable to systemic antibiotics, with fewer adverse events and lower dry socket incidence, supporting local delivery as an alternative to routine systemic antibiotic use after impacted third molar extraction
Yanine et al. (2021) Chile Parallel-group, randomized, blinded, placebo-controlled clinical trial 154 healthy patients aged 15–35 years undergoing extraction of at least one impacted mandibular third molar (Pell and Gregory class II, position B) Preoperative antibiotic prophylaxis Placebo Amoxicillin 2 g PO, single dose, 1 h before surgery POD 3, 7, and 30 Postoperative infection (alveolar osteitis, SSI), rescue analgesia use, adverse drug reactions Preoperative amoxicillin did not significantly reduce postoperative infectious complications compared with placebo, although it reduced the need for rescue analgesia; routine prophylaxis was not justified in healthy patients
Donmezer & Bilginaylar (2021) Turkey Prospective, randomized, split-mouth study 75 healthy patients (18–40 years) undergoing surgical extraction of impacted mandibular third molars Group 1: PRF + systemic amoxicillin/ clavulanic acid; Group 2: PRF + local amoxicillin/ clavulanic acid; Group 3: PRF + systemic clindamycin; Group 4: PRF + local clindamycin PRF alone (no antibiotics) Systemic: amoxicillin/ clavulanic acid 875/125 mg PO BID × 5 days or clindamycin 300 mg PO TID × 5 days. Local: PRF + 0.5 mL amoxicillin/ clavulanic acid (1000/200 mg) or PRF + 0.5 mL clindamycin (600 mg/4 mL) intra-alveolar POD 1, 2, 3 and 7 Pain (VAS), analgesic intake, trismus, swelling Both systemic and local antibiotic administration combined with PRF reduced postoperative pain and analgesic intake compared with PRF alone; local antibiotic delivery showed similar effects to systemic administration and may reduce exposure to systemic antibiotics
Braimah et al. (2017) Nigeria Observational comparative study with allocated treatment groups 135 healthy adults (18–35 years) undergoing impacted mandibular third molar extraction; indications mainly pericoronitis and apical periodontitis Group A: extended amoxicillin/ clavulanic acid; Group B: single-dose amoxicillin/ clavulanic acid Group C: singledose levofloxacin Group A: amoxicillin/ clavulanic acid 875/125 mg PO 1 h pre-op + 500/125 mg PO BID × 5 days; Group B: amoxicillin/ clavulanic acid 875/125 mg single pre-op dose; Group C: levofloxacin 1000 mg single pre-op dose POD 1, 3, 5, 7 and14 Oral health-related quality of life (UKOHRQoL) QoL worsened after surgery in all groups; extended amoxicillin/ clavulanate showed slightly better recovery, but differences were not statistically significant
Rabi et al. (2018) India Randomized controlled trial 60 healthy adults (20–35 years) undergoing surgical extraction of impacted mandibular third molars Group I: amoxicillin/ clavulanic acid; Group II: amoxicillin/ clavulanic acid + metronidazole Group III: no antimicrobial treatment Group I: amoxicillin/ clavulanic acid 625 mg PO for 5 days. Group II: amoxicillin/clavulanic acid 625 mg + metronidazole 400 mg PO for 5 days. Control: no antibiotics. Analgesics prescribed in all groups 1, 3, 5 and 7 Pain (VAS), mouth opening/interincisal distance (trismus), patient satisfaction Postoperative antimicrobial administration did not produce significant differences in postoperative sequelae after impacted mandibular third molar extraction; routine use showed no clear advantage
Øyri et al. (2019) Norway Single-blind randomized controlled trial 200 adults (ASA I–II) undergoing mandibular third molar surgery Oxytetracycline-impregnated gauze drain placed in extraction socket Sham drain placement (no antibiotic); no systemic antibiotics Local antibiotic: Terramycin-Polymyxin B (oxytetracycline + polymyxin B) applied on intra-alveolar gauze immediately after extraction. No systemic antibiotics used 7 days Primary: alveolar osteitis. Secondary: postoperative pain (NRS-11), analgesic use, absence from work/school, postoperative morbidity Local antibiotic drain significantly reduced alveolar osteitis incidence and postoperative morbidity without systemic antibiotic exposure
Mariscal-Cazalla et al. (2021) Spain Single-center, double-blind randomized controlled trial 92 ASA I patients (18–63 years) undergoing impacted mandibular third molar extraction Group 1: pre- and postoperative amoxicillin; Group 2: postoperative amoxicillin only Group 3: placebo Group 1: amoxicillin 750 mg PO q8h for 2 days pre-op + 5 days post-op. Group 2: amoxicillin 750 mg PO q8h for 5 days post-op. Group 3: placebo pre- and post-op. Rescue antibiotic if infection: amoxicillin/ clavulanic acid 875/125 mg q8h × 7 days 7 days Primary: postoperative infection, need for rescue antibiotics. Secondary: pain (VAS), swelling, jaw opening (trismus), adverse events, rescue analgesia Antibiotic prophylaxis reduced postoperative pain and inflammation compared with placebo; no advantage was observed for adding preoperative administration to postoperative treatment alone
Otake et al. (2021) Japan Retrospective observational study 261 patients undergoing lower third molar extraction Gauze coated with oxytetracycline-hydrocortisone ointment inserted into extraction socket No gauze insertion Systemic (all patients): amoxicillin 1 g PO pre-op + 1500 mg/day PO for 2 postoperative days. Test group additionally received local oxytetracycline-hydrocortisone ointment intra-alveolarly POD 0, 1, 3 and7 Dry socket (AO), pain (VAS), analgesic consumption Local oxytetracycline-hydrocortisone application reduced dry socket occurrence and postoperative pain
Janas-Naze et al. (2022) Poland Randomized, triple-blind, controlled clinical trial 278 patients undergoing surgical extraction of impacted mandibular third molars Group I: clindamycin 150 mg q8h; Group II: clindamycin 300 mg q8h Group III: clindamycin 600 mg q12h Group I: clindamycin 150 mg PO q8h × 5 days; Group II: clindamycin 300 mg PO q8h × 5 days; Group III: clindamycin 600 mg PO q12h × 5 days 5 days Infection, pain (VAS), rescue analgesia, trismus, edema, dysphagia, lymphadenopathy, salivary clindamycin concentration Lower-dose regimens showed similar or better outcomes; shorter and lower-dose administration was suggested
Sivalingam et al. (2017) India Randomized controlled trial, split-mouth, single-blind 33 ASA I patients requiring bilateral impacted mandibular third molar extraction Topical ozone gel (2 min, BID ×3 days), no systemic antibiotics Systemic antibiotics + plain gauze Control: amoxicillin 500 mg + metronidazole 400 mg PO TID ×5 days POD 1, 3, 7 Pain (VAS), swelling, mouth opening, analgesic use Ozone improved postoperative comfort and was proposed as an alternative to postoperative systemic antibiotics

Abbreviations: ATB3 = antibiotic therapy for 3 days; ATB7 = antibiotic therapy for 7 days; AO = alveolar osteitis; ASA = American Society of Anesthesiologists physical status; BID = twice daily; CHX = chlorhexidine; CTR = control group; mg = milligrams; mL = milliliters; NRS-11 = 11-point Numeric Rating Scale; OHRQoL = oral health-related quality of life; PBMT = photobiomodulation therapy; PO = oral administration; POD = postoperative day; PRF = platelet-rich fibrin; q8h = every 8 hours; q12h = every 12 hours; QoL = quality of life; SSI = surgical-site infection; TID = three times daily; VAS = Visual Analogue Scale.

Table 2. Reported clinical outcomes by intervention comparison and review-level interpretation.
Study Administration Reported outcome data Review-level interpretation
Kirnbauer et al. (2022) Systemic only SSI (11% overall), swelling (NS), trismus (NS), pain (NS), bleeding (NS) No statistically significant paired difference was reported. Group-specific event counts were unavailable for calculation of an effect estimate.
Yanine et al. (2021) Systemic only Postoperative infection: 2/77 with amoxicillin vs 5/77 with placebo; RR 0.40 (95% CI 0.08–2.00). Rescue analgesic use was lower with amoxicillin. The estimate is imprecise and does not exclude either clinically important benefit or no benefit.
Mariscal-Cazalla et al. (2021) Systemic only Higher postoperative pain (NS), inflammation (NS), and greater rescue medication use in placebo group Antibiotics were associated with lower pain and inflammation and less rescue medication; infection-specific event counts and confidence intervals were unavailable in the extracted data.
Rabi et al. (2018) Systemic only Pain (NS), trismus (NS), slightly higher patient satisfaction in combined antibiotic group No clinically meaningful reduction in postoperative complications.
Braimah et al. (2017) Systemic only Temporary deterioration of OHRQoL after surgery, especially POD1–3 (NS between regimens) This observational regimen-comparison study does not answer whether antibiotics are needed and is susceptible to confounding.
Janas-Naze et al. (2022) Systemic only Infection (NS), edema (NS), trismus (NS), dysphagia (NS), lymphadenopathy (NS) A dose-comparison trial; similar outcomes across doses do not establish that antibiotic therapy is necessary.
Øyri et al. (2019) Local only Alveolar osteitis: 5/100 tetracycline-drain sockets vs 23/100 control sockets; RR 0.22 (95% CI 0.09–0.55). A substantial reduction was reported in this single trial; replication and safety assessment are needed.
Otake et al. (2021) Combined systemic + local (same patients) Dry socket: 2/215 treated vs 9/46 control; RR 0.05 (95% CI 0.01–0.21). Pain and analgesic use were lower in the treated group. Retrospective observational evidence with strong potential for confounding; all patients received systemic amoxicillin, so the local ointment was adjunctive.
Donmezer & Bilginaylar (2021) Parallel local vs systemic (all groups received PRF) Lower pain and lower analgesic intake in local groups; swelling (NS); trismus (NS) Local antibiotic + PRF showed similar or lower symptom scores than systemic antibiotic + PRF in this study. Equivalence/non-inferiority was not established, and infectious outcomes were not the principal basis of the comparison.
Cigerim et al. (2024) Local only Alveolitis: 2/35 rifamycin-treated sockets vs 6/35 control sockets; RR 0.33 (95% CI 0.07–1.54). Pain was lower on selected postoperative days; edema and trismus were not significantly different. The point estimate favors topical rifamycin, but the confidence interval is wide and the trial is small.
Chugh et al. (2024) Parallel local antiseptic vs systemic antibiotic Pain (NS), trismus (NS), lower swelling only on POD3, significant difference in dry socket occurrence, fewer gastrointestinal adverse events Several outcomes were not statistically different and gastrointestinal adverse events were fewer with local CHX. This does not establish equivalence or non-inferiority unless a prespecified margin and appropriate analysis were used.
Sivalingam et al. (2017) Parallel local non-antibiotic vs systemic antibiotic (split-mouth) Greater pain, swelling, trismus, and higher analgesic requirement in systemic antibiotic sessions Topical ozone improved patient-comfort outcomes in a small split-mouth trial; the study does not directly establish prevention of postoperative infection.
Pereira et al. (2025) Systemic vs non-antibiotic comparator Pain, oedema, reduced mouth opening; ATB groups comparable to control Antibiotics did not improve the reported healing/symptom outcomes in this trial; the study was not primarily an infection-prevention trial.

Abbreviations: AO = alveolar osteitis (dry socket); ATB = antibiotic therapy; CHX = chlorhexidine; NS = not statistically significant; OHRQoL = oral health-related quality of life; PBMT = photobiomodulation therapy; POD = postoperative day; PRF = platelet-rich fibrin; QoL = quality of life; SSI = surgical-site infection.

Administration categories:

• Systemic only = antibiotic administered systemically only

• Local only = local administration without systemic antibiotics

• Combined systemic + local (same patients) = the same patients received both systemic and local administration

• Parallel local vs systemic = different study arms received local or systemic administration

• Systemic vs non-antibiotic comparator = systemic antibiotics compared with a non-antibiotic intervention

Table 3. Review-level interpretation by clinical comparison.
Clinical comparison Included evidence Main inference Confidence and limitations
Systemic prophylaxis or therapy vs placebo/no antibiotic Yanine; Kirnbauer; Mariscal-Cazalla; Rabi Individual estimates are inconsistent and often imprecise. A modest preventive benefit cannot be excluded. Limited confidence: small event numbers, heterogeneous regimens and outcomes, incomplete effect data, and no reproducible formal risk-of-bias assessment.
Local antibiotic vs local control Cigerim; Øyri; Otake Some studies reported lower alveolar-osteitis incidence. Limited confidence: single-center trials, one observational study, clinical heterogeneity, and adjunctive systemic antibiotic use in Otake.
Comparison among antibiotic regimens/doses Braimah; Janas-Naze May inform dose or duration but does not determine whether antibiotics are indicated. Very limited confidence for determining whether antibiotics are necessary because these studies address dose or regimen selection.
Antibiotic vs non-antibiotic local strategy Pereira; Chugh; Sivalingam Selected outcomes were similar or favored local strategies, but equivalence was not established. Very limited confidence: heterogeneous comparators, subjective outcomes, small studies, and no prespecified equivalence margins.

Interpretation note: A non-significant between-group difference is not evidence of equivalence. Claims of comparable efficacy require a prespecified equivalence or non-inferiority design, a clinically justified margin, and confidence-interval analysis.