Discussion

Our findings demonstrate that antibiotic prescribing for oral infections by VA dentists was mostly consistent with the ADA CPG prior to its release: 88.3% of IP and 82.6% of AP visits did not receive antibiotics and, thus, were consistent with the guidelines. However, 11.7% of IP visits and 17.4% of AP visits were associated with potentially unnecessary antibiotic prescriptions. Prolonged use of antibiotics was frequent for visits that were associated with antibiotic prescribing (42.5% of IP visits, 44.9% of AP visits, 49.4% of AAA visits received antibiotics for ≥8 days). Interestingly, the Western US had lower odds of antibiotic prescribing for IP and AP, whereas the Midwest was associated with antibiotic prescribing for IP. These geographic trends are consistent with studies conducted in primary care medical clinics,^{Reference Fleming-Dutra, Hersh and Shapiro12} but they are inconsistent with the highest prevalence of unnecessary antibiotic prophylaxis in the Western US.^{Reference Suda, Calip and Zhou13} Furthermore, specific dental procedure categories were associated with antibiotic prescribing for all infection types including oral maxillofacial surgery, tooth implants, and root canals (endodontic procedures). The guideline-discordant antibiotic prescribing observed within the study may represent unnecessary antibiotic prescribing placing patients at increased risk of adverse events without providing any meaningful benefit.^{Reference Thornhill, Dayer, Prendergast, Baddour, Jones and Lockhart14} Antimicrobial stewardship programs, which have reduced unnecessary dental antibiotics through employment of education and institutional guidelines, have been successfully implemented in other facilities, therefore showing potential for this prescribing group.^{Reference Gross, Hanna, Rowan, Bleasdale and Suda15}

Importantly, not all antibiotics used for acute oral infections carry the same risk of adverse events. Clindamycin, a lincosamide antibiotic, has a higher likelihood of causing CDIs relative to other antibiotic agents and carries a significantly higher rate of fatal adverse drug reactions compared to penicillins.^{Reference Gross, Suda and Zhou16–Reference Teng, Reveles, Obodozie-Ofoegbu and Frei19} CDI not only causes great economic losses to society but also imparts deep suffering for patients and their families.⁵ Efforts to reduce unnecessary clindamycin usage has been executed successfully, with significant decreases in CDI.^{Reference Cruz-Rodríguez, Hernández-García, Salinas-Caballero, Pérez-Rodríguez, Garza-González and Camacho-Ortiz20,Reference Climo, Israel, Wong, Williams, Coudron and Markowitz21} One teaching hospital observed a >80% reduction in CDI following the restriction of clindamycin in their orthopedic ward.^{Reference Cruz-Rodríguez, Hernández-García, Salinas-Caballero, Pérez-Rodríguez, Garza-González and Camacho-Ortiz20} With dentists currently being the top prescribers of clindamycin,^{Reference Suda, Roberts, Hunkler and Taylor22} reducing use of clindamycin by limiting treatment durations and promoting first-line use of amoxicillin or clindamycin alternatives in penicillin-allergic patients, for acute oral infections is critical. To achieve the latter, antimicrobial stewardship efforts geared toward thorough investigation of β-lactam allergy history can mitigate unnecessary clindamycin use.

With higher rates of adverse effects occurring with prolonged exposure to antibiotic therapies, using the shortest effective course is a prudent practice.^{Reference Tamma, Avdic, Li, Dzintars and Cosgrove23–Reference Royer, Demerle, Dickson and Prescott26} Antibiotic courses of 3–5 days for acute oral infections are likely as effective as 7-day courses.⁸ Although few studies support 3-day versus 7-day courses for acute oral infections, similar studies from medical literature support using the shortest effective treatment to maximize benefit while minimizing risk of antibiotic collateral damage.^{Reference Vaughn, Flanders and Snyder24–Reference Royer, Demerle, Dickson and Prescott26}

VA dental patients are more medically compromised than the general population with a higher rate of expected poor healing.^{Reference Jurasic, Gibson, Wehler, Orner and Jones27} Although in the full model the Charlson comorbidity index was not significant, when excluding those patients with probable premedication dosing or immunocompromised status, a higher Charlson comorbidity index did become significant. These patients may be considered more medically compromised than the general population described in the ADA guidelines.

Although most antimicrobial stewardship interventions have been conducted in inpatient settings, there has been an increasing focus on outpatient antimicrobial stewardship, including some successful dental stewardship interventions.^{Reference Spellberg25,Reference Royer, Demerle, Dickson and Prescott26} Previous studies have shown audit and feedback as well as education to be effective in curbing inappropriate dental antibiotic prescribing while also promoting accuracy in agent and dose selection.^{Reference Chate, White and Hale28,Reference Seager, Howell-jones, Dunstan, Lewis, Richmond and Thomas29} A recent study showed that implementation of a dental antibiotic stewardship program was successful in reducing antibiotic prescribing rates for acute oral infections by >70%. This stewardship program formulated an institutional guideline with recommendations largely consistent with the ADA CPG. As a behavioral “nudge,” recommendations in the institutional guideline were provided in a poster format and hung chairside (next to the dental chair). The results of the study confirm that implementing measures to promote adherence to the ADA CPG recommendations are a feasible and effective approach to curbing unnecessary antibiotic use.^{Reference Gross, Hanna, Rowan, Bleasdale and Suda15} Due to the association of antibiotic prescribing with oral and maxillofacial surgery, implant procedures, and endodontic procedures across all infection types observed in this study, it seems reasonable for a dental stewardship program to implement focused initiatives with the goal of preventing unnecessary antibiotic use for patients undergoing these specific procedures.

This study has several limitations. The results are dependent on the accuracy of ICD-10 coding in the dental setting; prescribing data suggest that antibiotics were prescribed for extra-oral infections (eg, nitrofurantoin, rifaximin). The study population includes only veterans treated within the Veterans’ Health Administration national healthcare system, which may not be representative of the general population. Dental visits associated with antibiotic courses <3 days and prescribed to patients with high risk cardiac conditions or receiving dental implants were not excluded in the primary analysis; thus, preprocedure prophylactic antibiotics may have been incorporated into the data. Patients without access to oral health care and those presenting to emergency or urgent care with oral conditions were not included. Thus, these results are not generalizable to antibiotic prescribing for oral infections in these groups. Additionally, the ADA guidelines provide recommendations for immunocompetent patients exclusively; however, immunocompromised patients were not excluded from the current study. However, sensitivity analyses were performed that excluded visits with patients with immunocompromising conditions. To account for the possibility that the antibiotic was prescribed for prophylaxis (vs treatment of an oral infection), the additional sensitivity analysis excluded patient visits that received antibiotic courses <3 days and patients with high-risk cardiac conditions. These 3 sensitivity analyses produced similar results to our primary analysis. Signs of systemic infection including extending infection, fever, and generalized malaise were not collected, making assessment of ADA guideline concordance for the treatment of AAA infeasible. Finally, these findings are based on data prior to publication of the ADA guidelines; thus, practice may have changed since the study period.

Regardless, the findings of this study suggest an opportunity for improving judicious antibiotic use in the outpatient dental setting by promoting concordance with the ADA acute oral infection guidelines. Due to ICD-10 codes not being commonly used in private-sector dental clinics, these unique VA data provide valuable insight into dental antibiotic prescribing practices for specific acute oral infections. These data provide additional evidence for antimicrobial stewardship initiatives targeting the reduction in the use of unnecessary antibiotics for IP and AP cases, especially for prolonged courses. Decreasing antibiotics prescribed for oral infections can lead to reductions in community-onset CDIs and infections caused by multidrug-resistant organisms.^{Reference Tampi, Pilcher and Urquhart30}

Prior to the release of the ADA guidelines, 88.3% and 82.6% of IP and AP visits in the VA setting were concordant with the ADA guidelines. Data from the current study identifies opportunities to improve acute oral infection antibiotic prescribing in outpatient dental settings. These results may serve as a benchmark for future outpatient antimicrobial stewardship efforts within the dental setting. Opportunities for future work in this area include the assessment of guideline concordance of antibiotic prescribing for AAA as well as acute oral infection antibiotic prescribing practices for both emergency department settings and for populations without access to health care.