High-Risk Medications in Pediatric Dentistry: Patterns, Outcomes, and Clinical Implications

Authors: Yilei Liu, Kawtar Zouaidi, Colin C. Hubbard, Muhammad F. Walji, Matthew R. Cooke, Donald B. Rindal, Kristin N. Ray, Elsbeth Kalenderian, Katie J. Suda

Abstract

Background

High-risk medications such as opioids and benzodiazepines are frequently prescribed in pediatric dental care, yet their prescribing patterns and associated adverse outcomes remain poorly understood. We aimed to determine the extent of such prescribing to pediatric patients and factors associated with adverse outcomes.

Methods

We analyzed MarketScan data for patients <18 years with a dental visit from 2014–2019. High-risk medications included benzodiazepines, barbiturates, or opioids. Composite outcomes included hospitalization, emergency department, or urgent care visit within 7 days post-visit. Opioid-attributable outcomes included opioid-related overdose within 7 days or persistent opioid use 4–365 days post-visit. Generalized estimating equations assessed the association between outcomes and patient- and visit-characteristics.

Results

Among pediatric dental visits, 0.72% (n=269,991) involved high-risk medications, with 4.3% experiencing a composite outcome. Higher odds were observed in ages 9–11 years (OR 1.56), males (OR 1.05), patients with complex chronic conditions (OR 2.22), and care delivered in hospital or ambulatory surgery settings (OR 2.20). Among dental visits with opioids, 10.1% had an opioid-attributable outcome, with the highest odds in patients 4–5 years (OR 1.48), females (OR 1.08), patients with complex chronic conditions (OR 1.22), and care delivered in outpatient clinics (OR 1.43).

Conclusion

One in ten pediatric dental visits involving opioids was associated with opioid-related overdose or persistent use with the highest odds in young children. These results highlight the need for caution in opioid prescribing and providing guideline-based non-opioid analgesia to children.

Practical Implications

Promote safer, equitable pediatric dental prescribing through guideline adherence, non-opioid pain management, and provider training.

Introduction

Dentists are prominent prescribers of opioids to manage post-operative pain,^1–3^ and dental visits are often children’s first exposure to opioids,^4,5^ with third molar (wisdom teeth) extractions as a common source in adolescents.^5,6^ Early opioid use has been associated with an increased risk of opioid misuse into adulthood.^7,8^ However, evidence on prescribing patterns and adverse events in pediatric dental practices is limited to medical clinics and hospitals,^9–12^ while research in outpatient dental settings is scarce.^13,14^ In addition, pediatric dental care often occurs in settings without integration of medical records, limiting dentists’ access to patients’ full medical histories, which is crucial for the safe administration and prescribing of medications.^15^

Due to fear or behavior management problems, some children are unable to cooperate with dental treatment under local anesthesia alone.^16^ Sedation, such as benzodiazepines, may be required to reduce anxiety and facilitating dental treatment.^17^ However, co-prescribing of opioids and benzodiazepines or other central nervous system (CNS) depressants is not recommended under the 2022 Centers for Disease Control and Prevention (CDC) guidelines and FDA-boxed warnings.^18,19^ Taken together, opioids and benzodiazepines increase the risk of respiratory depression, hospitalization, overdose, and death. From 2000 to 2019 in the US, opioids were involved in 84% of benzodiazepine-related overdose deaths,^20^ and in 2017, 21% of opioid overdose deaths involved benzodiazepines.^21^ Among adolescents, concurrent use of benzodiazepines with opioids was associated with an 83% increased risk of opioid-related overdose.^22^ However, the CDC guidelines are largely adult-focused and exclude children under 18 years,^23–25^ highlighting the need for pediatric-specific data.

Case reports and analyses of adverse outcomes after pediatric dental care highlight the urgent need to understand and apply appropriate safety practices for children in dental practices.^26–28^ However, US data on medication prescribing patterns of pediatrics by dentists and association with adverse events is limited. Therefore, we aim to assess pediatric outpatient claims for high-risk medications prescribed by dentists and identify factors associated with adverse outcomes.

Methods

Study design and data source

Using a unique patient identifier available across MarketScan files, we linked three claims databases (Commercial Claims and Encounters Database [CCAE], Dental Claims Database, and Multi-state Medicaid Database [MDCD]) to construct a national cohort of patients with commercial and Medicaid insurance. We conducted a longitudinal analysis of a retrospective cohort of patients under 18 years with at least one dental visit from 2014–2019.

Patient-level data included age, sex, race (only available for Medicaid beneficiaries), geographic region (only available for commercial beneficiaries), health plan, treatment settings, healthcare use one year prior to the dental visit (hospitalization, outpatient visits, or emergency department [ED] visits), concurrent medications, and dental procedures (per Comprehensive Dental Terminology [CDT] codes). Anticipated pain intensity from post-dental procedures was grouped as limited (e.g., diagnostics), mild (e.g., simple extraction), moderate (e.g., implant surgery), and severe (e.g., bony impaction surgery) based on CDT codes.^29,30^ Medical conditions 12 months prior to the dental visits were categorized as non-chronic, non-complex chronic, and complex chronic using the Pediatric Medical Complexity Algorithm (PMCA) algorithm.^31^ Neurological impairment was identified using high-intensity neurological impairment codes.^32,33^ Congenital malformations, (e.g., congenital facial malformations, congenital respiratory malformations, and other congenital malformations), were identified using ICD-9/10 and Current Procedural Terminology (CPT) codes.^34^

Study population

Among children with dental visits, we identified outpatient prescription claims for high-risk benzodiazepines, barbiturates, and opioids. Only one visit with barbiturates was identified. Thus, we excluded this visit and barbiturates from the cohort. Opioid prescriptions filled within 7 days before, on the day of, or after the visit were included to capture use for anticipated or post-procedural pain. Benzodiazepines filled within 7 days prior to the index visit reflected pre-procedural sedation or anxiety management.

To assess baseline conditions and outcomes, we included patients enrolled for at least one year prior and 6 months after the index dental visit. Because MarketScan does not identify the prescribing provider within the prescription claims data, we applied published criteria to reduce the likelihood that medications were unrelated to the dental visit.^29,35^ Patients were excluded if they (1) a hospice encounter in the year prior to the dental visit;^36,37^ (2) medical provider visits within 7 days prior; (3) chronic pain diagnosis within one year prior;^38^ (4) same-day diagnosis of oral pain;^38^ (5) sickle cell disease or seizure disorder;^39–41^ or (6) chemotherapy within 30 days prior.^42,43^ To include only incident opioid use, we excluded opioid fills indicating a refill or a renewal, defined as a 90-day opioid supply in the past year, or opioid prescriptions for the same active ingredient (with no dental visit) 8–35 days before the index date.^44,45^

Outcomes

The composite outcome included any hospitalization, ED, or urgent care (UC) visits within 7 days post-dental visit. No deaths were reported. Opioid-attributable outcomes included opioid-related overdose within 7 days or persistent opioid use (a new opioid claim between 4–365 days post-visit).^46,47^ Opioid-related overdose was defined as any occurrence of the following diagnoses based on Healthcare Common Procedure Coding System (HCPCS) or ICD-9/ICD-10 poisoning by opiates and related narcotics, alteration of consciousness, malaise, fatigue, or lethargy, respiratory failure, or naloxone administration.^48,49^

Statistical analysis

Given the large sample size and few visits per patient, within-patient correlation was ignored in descriptive analyses. Wilcoxon Rank Sum test and Chi-square test or Fisher’s exact test (with simulated p-values) were used to compare variables. Generalized estimating equations (GEEs) with an AR(1) correlation structure modeled associations between outcomes and characteristics, reporting odds ratios and 95% confidence intervals (CIs). Multicollinearity was assessed using variance inflation factor. Missing data were retained as unknown and modeled as a separate category. Analyses were conducted in R (4.4.0). Statistical significance was defined as p<0.001, using Bonferroni correction for up to 50 comparisons to control family-wise error.^50^

Multiple sensitivity analyses tested robustness. First, we included data from January 2020–June 2021, which were excluded from the primary analysis due to COVID-19 impacts in 2020 and partial data in 2021 (to allow for a 6-month follow-up). Second, we stratified by insurance type since race was only available for Medicaid and region for commercial beneficiaries due to limitations of the MarketScan datasets. Third, we narrowed the definition of persistent opioid use to 4–90 days post-visit (instead of 4–365 days).

Results

Patient characteristics

Among the 37,269,952 dental visits, 0.72% (n=269,991 visits for 204,224 unique patients) received a high-risk medication prescribed by a dentist (Figure 1). Over half of patients receiving high-risk medications were female (54%) with a median age of 15 years (interquartile range [IQR]=12–16) (Table 1). Dental visits associated with a high-risk medication decreased from 2014 (18.0%) to 2019 (12.8%). Over half (56.6%) involved oral and maxillofacial surgery procedures and 4.3% for patients with complex chronic conditions. Notably, 29.1% of high-risk medications were prescribed for dental visits associated with limited pain, which is higher than visits with mild (23.0%) or moderate (9.2%) pain. Almost all patient and visit characteristics significantly differed across medication groups (p<0.0001).

High-risk medications prescribed

Of the dental visits associated with high-risk medications, 91.2% involved opioids, 5.7% benzodiazepines, and 3.1% received opioids concurrent with benzodiazepines. Among overall dental visits associated with any opioid or benzodiazepine, hydrocodone (50.3%), codeine (30.0%), and oxycodone (12.9%) were the most frequent opioids, while diazepam (6.4%) was the most common benzodiazepine (Table 2). Codeine-containing products were more frequent in the opioid-only group, while oxycodone-containing products were more frequent in the concurrent opioid and benzodiazepine group. Among 66,636 visits (24.7% of all dental visits with opioids or benzodiazepines) for children under 12 years, 60.8% included codeine-containing products, and 0.2% involved tramadol-containing products.

Factors associated with adverse outcomes

Composite outcomes

Among all dental visits with high-risk medications, 11,731 (4.3%) experienced a subsequent composite outcome (Table 3). Composite outcomes were more frequent for the opioid-only visits (4.5%) than the other groups (3.5% for benzodiazepine-only, 2.4% for benzodiazepines and opioids).

Compared to dental visits involving benzodiazepines, opioid-only (OR 2.10, 95% CI 1.87–2.32, p<0.0001) and concomitant opioid and benzodiazepine visits (OR 1.69, 95% CI 1.39–1.98, p<0.0001) were more likely to experience composite outcomes (Figure 2). Children aged 9–11 years had the highest odds (OR 1.56, 95% CI 1.46–1.65, p<0.0001; Ref: 15–17 years), while females were 5% less likely than males (OR 0.95, 95% CI 0.92–0.99, p=0.0186). Compared to dental care delivered in outpatient clinics, hospitals or ambulatory surgery were associated with the composite outcome (OR 2.20, 95% CI 1.97–2.43, p<0.0001). Visits with invasive dental procedures (OR 0.38, 95% CI 0.37–0.40, p<0.0001) or care from dental specialists (OR 0.90, 95%CI 0.86–0.94, p<0.0001) were less likely to experience hospitalization or ED/UC visits. In contrast, patients with complex chronic conditions (OR 2.22, 95% CI 2.05–2.38, p<0.0001), prior history of ED/UC visits (OR 2.61, 95% CI 2.50–2.72, p<0.0001), and neurological impairment (OR 1.20, 95% CI 1.09–1.31, p=0.0001) were more likely to experience the composite outcome.

Opioid-attributable outcomes

Of the dental visits with opioids prescribed, 10.1% experienced an opioid-attributable outcome (10.3% in the opioid-only group, 5.7% for opioid + benzodiazepine group) (Table 4). Of the opioid-attributable outcomes, 71.3% were for persistent opioid use 4–90 days post-visit, 28.7% persistent use 91–365 days, and 0.3% associated with an opioid-related overdose.

Compared to opioid-only dental visits, opioid + benzodiazepine visits were not associated with higher odds of opioid-related overdose or persistent opioid use (Figure 3). Younger children (e.g., for aged 4–5 years, OR 1.48, 95% CI 1.38–1.57, p<0.0001; Ref: 15–17 years), females (OR 1.08, 95% CI 1.05–1.11, p<0.0001, Ref: males), patients with complex chronic conditions (OR 1.22, 95% CI 1.14–1.30, p<0.0001), and Medicaid beneficiaries (OR 1.29, 95% CI 1.25–1.33, p<0.0001) were more likely to experience opioid-attributable outcomes. Compared to dental care delivered in outpatient clinics, patients receiving treatments in hospital or ambulatory surgery settings were 30% less likely to experience opioid-attributable outcomes (OR 0.70, 95% CI 0.62–0.77, p<0.0001). Invasive dental procedures (OR 0.25, 95% CI 0.24–0.26, p<0.0001) or care from dental specialists (OR 0.85, 95%CI 0.83–0.88, p<0.0001) were associated with lower odds. Interestingly, opioid-attributable outcomes increased over time with the highest odds in 2018 (OR 1.20, 95% CI 1.14–1.26, p<0.0001; Ref: 2014).

Sensitivity analysis

Adding data from 2020 and the first 6 months of 2021 to the 2014–2019 cohort, we observed decreased likelihood of the composite outcome in 2020 compared to 2014, with 2021 returning to no significant difference (Supplement Figure 1). In contrast, the odds of opioid-attributable outcomes for each year from 2015–2021 were higher relative to the reference year of 2014 (Supplement Figure 2). We observed higher odds ratios for year when restricting opioid-attributable outcomes to persistent opioid use 4–90 days post-visit compared to 4–365 days post visit (e.g., for 2018, OR 1.54, 95% CI 1.45–1.63 vs. OR 1.20, 95% CI 1.14–1.26, Ref: 2014, Supplement Figure 3 and Figure 3). In the primary analysis, compared to outpatient clinics, patients receiving treatments in hospitals or ambulatory surgery settings were 120% more likely to have the composite outcome (Figure 2). However, the odds of patients receiving treatments in hospitals or ambulatory surgery settings were much higher within the commercial beneficiary cohort (OR 7.13, 95% CI 5.72–8.55, p<0.0001, Supplement Figure 4) and were 19% lower in the Medicaid beneficiary cohort (OR 0.81, 95% CI 0.68–0.94, p=0.0095, Supplement Figure 5).

Discussion

Among all dental visits from 2014 to 2019, 0.72% experienced a high-risk medication prescribed by a dentist, of which 4.3% experienced a hospitalization, emergency or urgent care visit within 7 days of the dental visit. Among children receiving opioids associated with dental visits, 10.1% had an opioid-attributable outcome, with the highest likelihood observed in younger patients. Across our outcomes, opioids were associated with increased odds relative to benzodiazepines. We observed generally increased odds of opioid-attributable outcomes from 2015–2018. These prescribing trends may be impacted by evolving coding practices, increased clinical awareness, and data completeness following the 2016 CDC opioid prescribing guidelines.^51,52^ Notably, the number of available opioid-related diagnosis codes increased dramatically from 20 ICD-9-CM codes to 100 ICD-10-CM codes,^53^ suggesting systematic differences in the overdose measurement.^54^ While our study adjusted for multiple covariates, we cannot fully separate these temporal influences from true changes.

Despite clinical warnings about the increased risk of co-using opioids and benzodiazepines,^18,19^ we found no significant difference in opioid overdose or persistent use between visits involving opioids alone versus those with concurrent benzodiazepines. This may reflect several factors. First, the low co-prescribing rate (3.1% of the cohort) may have limited statistical power. Second, dental prescriptions are typically short-term, potentially reducing risk differences. Third, dentists may be more cautious when co-prescribing, such as selecting lower-risk patients, offering additional monitoring, or staggering medication timing. Nonetheless, our results should not be interpreted as the combined use of opioids with benzodiazepines are without risk, and the results highlight the need to improve safety even when opioids are prescribed alone. Dental procedures categorized as mildly invasive or invasive, those performed by dental specialists (vs. by general dentists) and in hospital or ambulatory surgery (vs. outpatient clinics) were associated with lower odds of opioid-attributable outcomes. A study in Veterans similarly found that routine dental visits had a higher prevalence of opioid overprescribing compared with invasive visits.^55^ Our findings in a pediatric cohort may be due to more structured pain management guidelines,^18,56^ the use of multimodal analgesia,^57–59^ and closer monitoring and post-procedure care in these settings.^60^ However, these results raise concern that opioids may be unnecessarily prescribed for procedures with minimal anticipated pain where non-opioid multimodal analgesia is adequate. Our findings highlight the importance of standardized pain management across all treatment settings to ensure safe opioid prescribing practices, especially in routine dental visits.

Race and region are potential confounders in our analysis of composite outcomes. To address the large proportion of missing race and region data, we conducted subgroup analyses separately within Medicaid beneficiaries (with race data, Supplement Figure 5) and commercial beneficiaries (with region data, Supplement Figure 4). The qualitative and quantitative inconsistencies in odds ratios across subgroups (e.g., OR 2.20 overall, 0.81 in Medicaid, 7.13 in commercial beneficiaries for hospital or ambulatory surgery settings vs. outpatient clinics) suggest residual confounding and likely reflect differences in healthcare infrastructure, healthcare access, provider practices, or unmeasured social factors. The lower odds observed in non-White children and in the Southern/Western regions highlight the need to further examine whether these patterns reflect true risk differences or potential biases in care delivery and reporting. Future studies with more complete and harmonized data are needed to clarify whether the observed inconsistencies reflect true differences in risk or underlying biases in care delivery and reporting.

Medicaid beneficiaries receiving opioids had higher odds of opioid-attributable outcomes compared to commercial beneficiaries. Medicaid coverage, an indicator of lower socioeconomic status, serves children from families with incomes ≤133% of the federal poverty level.^61,62^ Our findings align with prior studies showing higher adverse event rates among Medicaid patients.^63–65^ Limited dentist participation in public insurance may push Medicaid dental beneficiaries toward safety-net settings with fewer providers.^66–68^ Few dental trainees anticipate practicing in these settings,^68^ raising concerns about provider capacity and care quality. Additionally, publicly insured patients are more likely to have limited English proficiency,^69^ which may further compromise medication safety. Our findings emphasize the need for oral health and pain management equity and are important for policy development to reduce opioid-related risks in vulnerable pediatric populations.

While most of the prescribing occurred among older children, younger children who did receive these medications were more likely to experience opioid-attributable outcomes; those aged 0–8 years experienced over 40% increased odds. This is concerning given that American Dental Association (ADA) guidelines recommend only non-opioid analgesics for children under 12 years.^70,71^ In addition, codeine- and tramadol-containing products are contraindicated in children <12 years,^72^ but comprised 70.0% of high-risk prescriptions to in this age group. Reducing such inappropriate prescribing is critical. The evidence for non-opioid multimodal analgesia (MMA),^73^ which maximizes the effects of non-steroidal anti-inflammatory drugs (NSAIDs) and acetaminophen, have demonstrated optimal pain control.^74^ In addition, providers are advised to adhere to the principle of minimum effective dosage of analgesic prescriptions to minimize adverse events.^71^ Our results support the guidelines which strongly recommend non-opioid MMA as first-line therapy and do not recommend opioids for pediatric pain management.^70,71^ Expanding provider training on non-opioid MMA and improving access to medical histories in dental settings may support safer prescribing.^75,76^ System-level support, including improved provider capacity and continuous education, may further promote guideline adherence.^77^

Limitations

This study has several limitations. First, race and region data were missing for almost half of the sample (45.2% and 57.3%, respectively), limiting the ability to assess their impact on prescribing and adverse outcomes. To address this, we conducted subgroup analyses within the Medicaid beneficiaries (with race data) and the commercial beneficiaries (with region data) to explore potential impacts of missingness. Second, findings are not fully representative of patients without dental insurance or access to oral care. While Medicaid includes oral care for children as part of the standard benefit, access remains limited,^78^ and commercial insurance often relies on separate insurance specific to dentistry.^79^ Third, as an observational study using claims data, we cannot account for all confounding factors or establish causality between high-risk medication use and adverse outcomes. Finally, we did not include a control group of dental visits without high-risk medications, so we did not directly evaluate the association between high-risk medications prescribed in pediatric dental visits with adverse outcomes. Future studies should incorporate a matched comparison group of pediatric dental visits without high-risk medications to better isolate the risks attributable to high-risk medications.

Conclusion

One in 10 pediatric dental visits involving a dental opioid was associated with opioid-related overdose or persistent use with the highest odds in young children. These results emphasize the need for caution in opioid prescribing and support clinical practice guidelines that exclusively recommend non-opioid analgesics and FDA warnings on use of codeine- and tramadol-containing products in children. To support adherence to these guidelines, additional evidence-based interventions are necessary to further reduce potential harm which may include education for providers, stricter prescribing policies, or non-opioid multimodal analgesia.

Supplementary Material