Dental Surgeries in Hospitals and Surgery Centers for Children with Developmental Disabilities

Abstract

Background and Objective:

Children with intellectual disabilities and related conditions (IDRC) and autism often require dental treatment in surgical settings. Studies of older adults suggest that ambulatory surgery centers (ASCs) can reduce costs without reducing quality, but it is unknown if this is true for children with autism and IDRC and dental surgeries. This study compared outcomes among children with IDRC and autism receiving dental caries-related surgeries at ASCs and hospital outpatient departments (HOPDs).

Methods:

This cross-sectional analysis used Medicaid data from 29 states (2016–2020) to examine 17,552 outpatient dental surgeries in ASCs and HOPDs for children with IDRC or autism. Regression models were used to examine if outcomes varied by setting of care (ASC vs. HOPD). Outcomes included: days from diagnosis to surgery, Medicaid payment, ED visits and hospitalization post-surgery.

Results:

Multivariate regression results indicate time from diagnosis to surgery was, on average, 8.7 days sooner for a child with autism and 13.2 days sooner for a child with IDRC if furnished in an ASC compared to a HOPD (P<0.001 for both). Average Medicaid payment was $419 and $363 less in an ASC than HOPD for a child with autism (P=0.003) or IDRC (P=0.02), respectively. The likelihood of an ED visit or hospitalization was lower in an ASC than HOPD for a child with IDRC (P<0.001), but not autism (P>0.05).

Conclusions:

Increasing use of ASCs for dental care may increase timely receipt of care and lower Medicaid costs without increasing adverse outcomes for children with autism and IDRC.

Introduction

Children with developmental disabilities, such as autism and intellectual disabilities and related conditions (IDRC), are at increased risk for dental caries due to many factors, including the sugar content and side effects of medications, poor dental hygiene related to oral sensory sensitivities, and limited cognitive abilities or manual dexterity.¹ Children with developmental disabilities also face barriers to dental care,² due in part to dentists’ inadequate training, low confidence caring for patients with developmental disabilities,² and disability-based bias and discrimination.^3,4

Children with Medicaid have the additional challenge of finding a dentist that accepts their insurance.^5,6 Many children with autism and IDRC receive treatment for dental caries under general anesthesia due to behavioral indications and burden of dental disease.^7,8 Surgical procedures for dental caries paid by Medicaid are typically provided in hospital outpatient departments (HOPDs) and ambulatory surgery centers (ASCs), with the rate of these procedures being performed in ASCs increasing since 2019.⁹ Both HOPDs and ASCs furnish surgical procedures that do not require overnight stays, with the latter being free-standing facilities providing only outpatient surgeries.

Access to HOPDs is a challenge for pediatric dentists,¹⁰ contributing to longer wait times, care delays, and more emergency department (ED) visits for dental caries.¹¹ As ASCs offer an alternative setting for dental surgery, they may lessen demand at HOPDs while offering cost advantages and maintaining comparable quality. ASCs generally cost 40–50% less than HOPDs,^12,13 and have similar rates of complications, patient anxiety, and patient satisfaction.^12,14 ASCs have lower costs due to their streamlined operations and lower overhead costs compared to HOPDs. In addition, procedures in ASCs take less time than HOPDs.¹⁵ However, most studies focus on adults and routine medical procedures and have not examined children with autism and IDRC. To our knowledge, three studies have examined pediatric dental surgeries, finding procedures in ASCs took less time and were less costly than HOPDs.^16–18 Two of these studies were conducted more than 10 years ago and none examined children with autism or IDRC.

This study compared timeliness of care, adverse events, and cost for pediatric dental surgeries performed in ASCs and HOPDs in children with autism and IDRC. As the autism community has diverse preferences regarding use of person-first and identity-first language,¹⁹ we use both throughout this paper and use “autism” to refer to diagnoses of “autism spectrum disorder” because some view the use of “disorder” as ableist.²⁰

Methods

Data

We obtained Transformed Medicaid Statistical Information System (T-MSIS) Analytic Files (TAF) for 2016–2020. The TAF Demographic and Eligibility File provides information about enrollee characteristics and benefits, the TAF Other Services File provides information about diagnosis of dental caries, outpatient surgical procedures, ED visits, and Medicaid payments, and the TAF Inpatient File provides information about inpatient hospitalizations.

Sample

We examined children and adolescents aged 18 years and younger with full Medicaid benefits and an outpatient surgical procedure in an ASC or HOPD for dental caries (International Classification of Diseases tenth revision code (ICD-10) K02). We required 8-months of continuous enrollment, six months prior to the surgical procedure and two months following the surgical procedure, to allow for examination of outcomes occurring before and after the procedure. We required children to have one or more diagnoses of intellectual disabilities and related conditions (IDRC) and/or autism during our study period, using condition definitions from the Centers for Medicare and Medicaid Services (CMS) Chronic Condition Warehouse.²¹ Due to concerns about the quality of TAF,^22–25 we excluded state-years that the CMS Data Quality Atlas flagged as “unusable” or of “high concern,”²³ as well as state-years with ≥20% missingness for place of service and ≥20% missingness for a dental caries procedure or diagnosis six months prior to the surgical procedure. We excluded procedures occurring on or after February 1, 2020 to exclude the emergence of the COVID-19 pandemic. We examined 17,552 surgical procedures in ASCs and HOPDs occurring July 1, 2016 to January 31, 2020 for children with IDRC and/or autism (Appendix Table A1) in 29 states that contributed one to five years of data (Appendix Table A2).

Exposure

We created a binary variable to identify if the outpatient surgical procedure was furnished in an ASC or HOPD, identified using the place of service variable. If missing, procedures were assigned to ASCs or HOPDs using the bill type code, billing provider type code, billing provider taxonomy code, and type of service code (Appendix Table A3).

Outcomes

We examined four outcomes: days from diagnosis to procedure, ED visits and hospitalizations after surgery, and Medicaid payment. We examined receipt of timely care by constructing a measure of days from diagnosis to surgical procedure. We used the date of surgery as an anchor and counted back the number of days from the procedure to the most recent encounter or claim with a diagnosis of dental caries, up to six months prior. Because dentists typically use procedure codes rather than diagnosis codes, we identified the most recent occurrence of a procedure code indicating evaluation and diagnosis (Current Dental Terminology codes D0120, D0140, D0145, D0150, D0160, D0190, D9310) or diagnosis of dental caries or diseases of pulp and periapical tissues (ICD-10 codes K02 and K04). To identify adverse events, we examined any ED visit within 7 days, excluding the day of the procedure to avoid visits that may precipitate a procedure, and any hospitalization within 30 days, including the day of the procedure. Total Medicaid payments were calculated by summing all fee-for-service (FFS) paid amounts on the date of the procedure. This was calculated only for procedures paid as FFS (42.4% of our sample) due to concerns about inaccurate and missing information on services paid via managed care. We adjusted payments to 2016 dollars using the Consumer Price Index for Medical Care and excluded extreme values (top and bottom one percent of the distribution of cost).

Analytic Approach

We analyzed children with IDRC and autism separately because although some dental care barriers are similar, many are distinct and vary in prevalence (e.g., sensory sensitivities and autism)^2,26 and children with IDRC are less likely than autistic children to use ASCs.²⁷ We reported descriptive statistics separately for children with autism and IDRC by place of service (ASC or HOPD) and tested for differences in case characteristics by place of service using t-tests for continuous variables and chi-squared tests for categorical variables. Children with both autism and IDRC were included in the sample of autistic children. We estimated regression models to determine the association between place of service and each outcome. Models controlled for race and ethnicity, age in years, sex, count of dental procedure codes billed with the procedure, a dichotomous measure of pediatric surgical risk, and state and year fixed effects. We included race and ethnicity, a social construct, due to differential use of ASCs and HOPDs.⁹ The count of dental procedure codes billed during the procedure serves as a proxy for disease severity. The pediatric surgical risk score was calculated using information on age, sex, and diagnosis codes, and the resulting values were categorized as low risk (score=0) or higher risk (score>0).²⁸ We estimated generalized linear models, with a log link and binomial distribution for ED visits and hospitalizations, an identity link and Gaussian distribution for days from diagnosis to surgery, and a log link and gamma distribution for FFS Medicaid payment. We present predicted probability by place of service and calculate marginal effects, with standard errors (SEs) clustered at the state-level. As controlling for state fixed effects resulted in several states being dropped from the sample, we also estimated regression models without state fixed effects to assess whether excluding these states influenced our results. Analysis was conducted using STATA/MP 18.0. A two-sided significance threshold was set at P<0.05. This study follows STROBE reporting guidelines for cross-sectional studies and was deemed exempt by RAND’s human subjects’ protection committee (category 4.²⁹

Results

Among autistic children who received dental surgery in an ASC or HOPD, 24.2% of procedures were performed in ASCs, and were more often furnished to children who were younger (7.5 vs. 8.0 years, p<0.001) and Hispanic (28.5% vs. 16.8%, p<0.001) and less often to Black (7.9% vs. 10.9%, p<0.001) and White (26.0% vs. 32.1%, p<0.001) children than procedures in HOPDs (Table 1). Among children with IDRC who received dental surgery in an ASC or HOPD, 17.3% of procedures were performed in ASCs, and were more often furnished to Hispanic children (32.4% vs. 22.4% p>0.001) and less often to White children (20.6% vs. 29.0%, p<0.001) than procedures in HOPDs. No other statistically significant sociodemographic differences were observed by place of service.

Table 1.

Sample characteristics by place of service for children with autism and intellectual disabilities and related conditions

	Autism (n = 12,944)		Intellectual Disabilities and Related Conditions (n = 4,643)
	HOPD	ASC	HOPD	ASC
	n (%)	n (%)	n (%)	n (%)
Observations, n1	9,813	3,131	3,840	803
Sex
Female	2,018 (20.56)	624 (19.93)	1,624 (42.29)	338 (41.09)
Male	7,795 (79.44)	2,507 (80.07)	2,216 (57.71)	465 (57.91)
Race and ethnicity
White, non-Hispanic	3,150 (32.10)	815 (26.03)**	1,115 (29.04)	165 (20.55)**
Black, non-Hispanic	1,065 (10.85)	248 (7.92)**	296 (7.71)	55 (6.85)
Hispanic	1,649 (16.80)	892 (28.50)**	859 (22.37)	260 (32.38)**
Other, non-Hispanic2	366 (4.07)	137 (4.38)	130 (3.39)	22 (2.74)
Missing	3,583 (36.51)	1,039 (33.18)**	1,440 (37.50)	301 (37.48)
Pediatric Surgical Risk3
No risk (score = 0)	9,775 (99.61)	3,113 (99.43)	3,798 (98.91)	NR
Higher risk (score ≥ 1)	38 (0.39)	18 (0.57)	42 (1.09)	NR
Count of dental procedures billed during procedure, mean (SD)	15.0 (8.8)	15.2 (7.7)	14.7 (9.2)	14.4 (8.4)

Abbreviations: ASC = Ambulatory Surgery Center, HOPD = Hospital Outpatient Department, NR = Not reported, FFS = Fee-for-service.

^*P<0.05,

^**P<0.001.

¹There are 35 cases with a diagnosis of both autism and IDRC. These cases are included in both the autism and IDRC columns.

²We used the race and ethnicity categories and variable labels provided in the Medicaid TAF data, combining groups to adhere to small cell size suppression policies in our data use agreement. Autism: Asian (n=191), American Indian/Alaska Native (n=180), Hawaiian/Pacific Islander (n=32) and Multiracial (n=100). Intellectual Disabilities and Related Conditions: Asian (n=61), American Indian/Alaska Native or Hawaiian/Pacific Islander (n=75, combined to censor small cell sizes), Multiracial (n=17).

³These results are not reported to adhere to small cell size suppression policies in our data use agreement.

Across all outcomes and models, unadjusted and regression-adjusted results were similar. The unadjusted days from diagnosis to procedure was 48.7 days in ASCs compared to 61.6 days in HOPDs for children with autism and 53.1 days in ASCs compared to 67.7 days in HOPDs for children with IDRC (Table 2). In regression-adjusted analyses, surgeries occurred 8.7 (95% confidence interval (CI) = −11.42 to −6.03, p<0.001) days sooner for an autistic child and 13.2 (95% CI = −16.39 to −9.95, p<0.001) days sooner for a child with IDRC in ASCs than HOPDs (Figure 1, Appendix Table A4, A5, and A6).

Table 2.

Unadjusted outcomes for ASC and HOPD use for pediatric dental surgical procedures among children with autism and intellectual disabilities and related conditions

	Days from diagnosis to surgery	Emergency department visit within 7 days of surgery2	Inpatient hospitalization within 30 days of surgery2	Medicaid Payment3
Autism1	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Probabilities
ASC	48.72 (42.12 to 55.33)	0.021 (0.015 to 0.026)	0.006 (0.004 to 0.009)	$1,614 ($1,406 to $1,822)
HOPD	61.56 (56.48 to 66.64)	0.023 (0.019 to 0.026)	0.007 (0.005 to 0.009)	$2,228 ($1,770 to $2,685)
Marginal effects
ASC vs. HOPD	−12.84** (−16.89 to −8.78)	−0.002 (−0.010 to 0.006)	−0.001 (−0.004 to 0.002)	−$614* (−$1,034 to −$194)
IDRC1	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Probabilities
ASC	53.09 (46.51 to 59.67)	0.017 (0.012 to 0.023)	0.010 (0.006 to 0.014)	$1,498 ($1,310 to $1686)
HOPD	67.70 (63.20 to 72.21)	0.028 (0.024 to 0.032)	0.022 (0.017 to 0.027)	$2,191 ($1,759 to $2622)
Marginal effects
ASC vs. HOPD	−14.62** (−18.63 to −10.60)	−0.010* (−0.017 to −0.004)	−0.012** (−0.017 to −0.006)	−$693* (−$1,133 to −$252)

¹For the 35 cases that had a diagnosis of both autism and IDRC, we included these cases in all models.

²Risk score was not included as a covariate in models for emergency department visits and hospitalizations because of high collinearity.

³Examined 5,288 procedures for children with autism and 1,974 procedures for children with IDRC paid fee for service.

Abbreviations: ASC = Ambulatory Surgery Center, HOPD = Hospital Outpatient Department.

^*P<0.05,

^**P<0.001

Figure 1.

Regression-adjusted predicted days from diagnosis to surgery by place of service for pediatric dental surgeries among children with autism and intellectual disabilities and related conditions

*P<0.05, **P<0.001. P-values represent within disability group comparisons of days from diagnosis to surgery between ASCs and HOPDs. Abbreviations: ASC = Ambulatory Surgery Center, HOPD = Hospital Outpatient Department, IDRC = Intellectual Disabilities and Related Conditions.

Unadjusted rates of ED visits within 7 days post-surgery and hospitalizations within 30 days post-surgery were similar for autistic children receiving care in ASCs and HOPDs. For children with IDRC, rates of ED visits and hospitalization post-surgery were lower in ASCs (1.7% and 1.0%, respectively) than HOPDs (2.8% and 2.2%, respectively) (Table 2). Results were similar in regression-adjusted analyses (Table 3). For a child with autism, the probabilities of a post-surgery ED visit or hospitalization were similar in ASCs and HOPDs. For a child with IDRC, the probability of an ED visit was 1.0 (95% CI = −1.6% to −0.4%, p<0.001) percentage point lower and the probability of hospitalization was 1.4 (95% CI = −1.8% to −0.9%, p<0.001) percentage points lower in ASCs than HOPDs.

Table 3.

Regression-adjusted outcomes for ASC and HOPD use for pediatric dental surgical procedures among children with autism and intellectual disabilities and related conditions

	Emergency department visit within 7 days of surgery2	Inpatient hospitalization within 30 days of surgery3
Autism1	Estimate (95% CI)	Estimate (95% CI)
Probabilities
ASC	0.019 (0.014 to 0.024)	0.007 (0.004 to 0.009)
HOPD	0.024 (0.022 to 0.025)	0.007 (0.006 to 0.008)
Marginal effects
ASC vs. HOPD	−0.005 (−0.011 to 0.002)	−0.001 (−0.004 to 0.003)
IDRC1	Estimate (95%CI)	Estimate (95%CI)
Probabilities
ASC	0.020 (0.016 to 0.025)	0.009 (0.006 to 0.013)
HOPD	0.030 (0.029 to 0.031)	0.023 (0.022 to 0.024)
Marginal effects
ASC vs. HOPD	−0.010 (−0.016 to −0.004)*	−0.014 (−0.018 to −0.009)*

^*P<0.001.

¹For the 35 cases that had a diagnosis of both autism and IDRC, we included these cases all both models.

²Examined 12,831 procedures for children with autism and 4,237 procedures for children with IDRC.

³Examined 12,410 procedures for children with autism and 4,495 procedures for children with IDRC.

Abbreviations: ASC = Ambulatory Surgery Center, HOPD = Hospital Outpatient Department.

Lower FFS Medicaid payment was observed for procedures in ASCs for both children with autism and IDRC. Unadjusted mean payment was $1,614 in ASCs and $2,228 in HOPDs for a child with autism and $1,498 in ASCs and $2,191 in HOPDs for a child with IDRC (Table 2). In regression-adjusted analyses, Medicaid payment in ASCs was $419 (95% CI = −$896 to −$141, p<0.003) less for an autistic child and $363 (95% CI = −$670 to −$57, p=0.02) less for a child with IDRC compared to HOPDs (Figure 2 & Appendix Table A4, A5, and A6). Using these estimates, Medicaid would be projected to save nearly $6.96 million if all pediatric dental surgeries for these children were furnished in ASCs. Results were similar in direction and statistical significance in models without state fixed effects (Appendix Table A7).

Figure 2.

Regression-adjusted predicted Medicaid payment by place of service for pediatric dental surgeries among children with autism and intellectual disabilities and related conditions

*P<0.05, **P<0.001. P-values represent within disability group comparisons of Medicaid payments between ASCs and HOPDs. Abbreviations: ASC = Ambulatory Surgery Center, HOPD = Hospital Outpatient Department, IDRC = Intellectual Disabilities and Related Conditions.

Discussion

Findings from this cross-sectional analysis of Medicaid data have important implications for Medicaid programs seeking to reduce costs and deliver timely care, as well as for hospitals seeking to address high demand for operating rooms.

Studies of ASCs have largely focused on adults and few examine dental procedures; however, our findings agree with extant studies showing ASCs provide care that is as safe and less costly than HOPDs.^12,13,15 Days from diagnosis to surgery was shorter in ASCs than HOPDs, by up to two weeks for children with IDRC, which is noteworthy given that some children in our study waited more than two months for surgery. Timely receipt of dental treatment is essential for preventing more serve dental disease and complications, reducing pain, and promoting quality of life, and a more positive patient experience.³⁰

Our finding that ASCs are cost-saving for Medicaid is important because Medicaid insures 42% of children with special health care needs,³¹ which includes children with autism and IDRC, and enormous cuts to Medicaid were signed into federal law in July 2025.³² As states struggle to pay for Medicaid, shifting more procedures to ASCs may help to balance their budgets.

Our study suggests that dental procedures performed in ASCs are as safe as those in HOPDs, with similar rates of ED visits and hospitalizations post-surgery for autistic children and lower rates for children with IDRC. Despite few adverse events for these procedures, children with autism and IDRC may have inequitable access to ASCs when compared with children without these conditions. Among pediatric dental procedures for all Medicaid-enrollees, about 30% occur in ASCs,³³ compared to 24% for autistic children and 17% of children with IDRC.²⁷ A similar pattern was observed in a study of patient medical records from a HOPD and an ASC exclusive to pediatric dentistry and oral surgery in Seattle between 2017–2018;ASCs were less likely to serve patients diagnosed with autism and developmental delays than HOPDs.³⁴ Provider disability bias, such as perceiving people with disabilities as unhealthy, more medically complex, or more difficult to provide care because of needed accommodations and lack of adequate training, may contribute to greater reluctance in treating these children in an ASC.^3,35 Some children with autism and IDRC may be better cared for in HOPDs (e.g., those at high risk for complications⁷), but these findings can help to shift the narrative that many children with autism and IDRC are not too medically complex for ASCs.

Although not the focus of this paper, we found differences in receipt of care by sex and by race and ethnicity. Among autistic children, nearly 80% of procedures were received by males, likely reflecting the higher prevalence of autism in males than females.³⁶ For both children with autism and IDRC, unadjusted rates of ASC use were highest among Hispanic children—although HOPD use was more common across all racial and ethnic groups. Future work should examine if surgery outcomes vary by place of service, sex, and race and ethnicity. Although equitable access to ASCs is important, receipt of regular dental visits can help to prevent dental care and the need for these procedures. Ensuring the availability of inclusive dental environments, appropriate accommodations such as sensory and behavioral supports, and training more providers with disability competency can help to promote dental visits for this population.^1,37

Despite the potential benefits of ASC use to patients and Medicaid programs, there are factors impeding increased use for dental surgery for children with autism and IDRC. First, few ASCs offer dental care. In 2024 only 7.6% of ASCs offered dental care.³⁸ Although the share of Medicaid-enrolled children receiving dental surgery in ASCs has increased since 2019,⁹ there are not yet enough ASCs to move all cases to ASCs. Second, even when a child may live in an area with both an ASC and HOPD, determining where the surgery occurs is a decision largely made by the dentist and anesthesia provider, which is likely driven by provider preferences and relationships, but may be susceptible to disability bias. Further, ASCs have been found to locate in affluent counties,³⁹ which may make them less accessible to children with disabilities, as they tend to have lower household incomes and less discretionary income because of disability-related costs in health care, transportation, and specialized aid.⁴⁰ Many children with IDRC and autism would benefit from similar accommodations that create a more inclusive dental environment, however accommodations likely differ according to their condition and should be tailored to their needs.^2,26 For example, an autistic child may need accommodations to mitigate sensory sensitivity (e.g., supports for bright lights and loud noises) whereas adaptive communication strategies may be needed for a child with IDRC. Finally, patient and caregiver preferences likely influence when and where the procedure occurs. Some caregivers may feel more comfortable with their child receiving care at HOPDs where there is easy access to emergency services. Other parental factors that facilitate dental care access for autistic children include an accessible dental environment, timely care and shorter waiting times, provider flexibility and tailored accommodations, and materials to better prepare parents their child for dental visits (e.g., social stories).²⁶ Whereas this study identifies many benefits to ASC use, a variety of factors, including supply limitations and patient and provider preferences, may help to explain differential rates of ASC use relative to HOPD use among children with autism and IDRC.

Limitations

First, this repeated cross-sectional study provides information about associations but not causation. While we attempted to control for factors that indicate care may be more appropriately delivered in HOPDs than ASCs, including higher surgical risk and the number of dental procedure codes billed during surgery, unmeasured differences could potentially bias our results. For example, we may observe higher rates of ED visits and hospitalizations for cases in HOPDs if these cases are more medically complex or challenging in ways that were not captured by our controls. Second, these findings may not be generalizable to all state Medicaid programs because we had to exclude some state-years due to concerns about TAF data quality. Third, we identified a diagnosis within six months of surgery for 81.5% and 77.2% of children with autism or IDRC, respectively. Due to concerns about data quality, we are uncertain if these diagnoses occurred earlier or are missing from our dataset. If the children missing this information are more complex or different in other ways from our sample our results may not be generalizable to them. Fourth, we examined only FFS Medicaid payments (42% of our sample) due to inaccurate and missing information for surgeries paid by Medicaid managed care. Because most children in most states are enrolled in Medicaid managed care,⁴¹ it is important to examine if cost-savings persist in managed care.

Conclusion

Analysis of Medicaid data from 29 states found that pediatric surgical procedures for dental caries in ASCs had a shorter time from diagnosis to treatment, were less expensive, and as safe as those performed in HOPDs for children with autism and IDRC. These findings highlight the role of ASCs in promoting timely and equitable care for children with developmental disabilities and for reducing costs for Medicaid. Additional research is needed to explore whether ASCs offer similar benefits for other outpatient procedures and if findings vary for subgroups defined by race and ethnicity, rurality, and insurance-type.

Abbreviations:

ASC

Ambulatory surgery centers

CMS

Centers for Medicare and Medicaid Services

ED

Emergency Department

FFS

fee for service

HOPD

Hospital outpatient department

ICD

International Classification of Diseases

IDRC

Intellectual disability and related conditions

TAF

T-MSIS Analytic Files

T-MSIS

Transformed Medicaid Statistical Information System