An Observational Retrospective Study of Adverse Events and Behavioral Outcomes during Pediatric Dental Sedation

Abstract

Purpose:

Electronic health records (EHRs) were used to identify sedation-related adverse events (AEs) and assess patients’ behavioral outcomes during routine pediatric dental sedations (PDSs) in a dental school clinic.

Methods:

A database was screened for patients younger than 18 years old who had a dental sedation during 2019. The qualifiying EHRs were then accessed, and sedations were reviewed for AEs. AEs were categorized using a 12-point classification system and the Tracking and Reporting Outcomes of Procedural Sedation tool. Patients behaviors were assessed using providers progress notes and categorized as presence/absence of agitation.

Results:

A total of 690 sedations were reviewed, yielding 28 AEs. Emesis was the most common AE observed in 1.3% of sedations. Respiratory and cardiovascular AEs were observed in 0.7% and 0.6% of sedations, respectively. Agitation was identified in 47.5% of sedations. 34.1% of agitations resulted in documented suspension of dental treatment. Agitation was mainly observed for Nitrous Oxide and oral sedation resulting in 1 failed sedation out of 5 sedations for each method.

Conclusions:

Potentially serious AEs were identified during PDSs, but their incidence was low. A significant proportion of the sedated children experienced agitation resulting in some sedation failures. Such events need to be tracked and examined for risk assessment reduction and quality of-care improvement.

INTRODUCTION

Pediatric dental sedation (PDS) is common in the US. An estimated 100,000 to 250,000 sedations for dental care of children are performed yearly.^{1, 2} In 2001, it was estimated that approximately one million US children required some form of pharmacological behavior management to facilitate dental treatment.³ A 25-year follow-up survey conducted among members of the American Academy of Pediatric Dentistry (AAPD) to monitor changes in sedation use in the US, revealed consistent increase since the 90’s.⁴ Children aged 3 to 5 years old comprise the highest percentage of children sedated for dental care.^{4, 5} Younger children often lack coping abilities necessary for a dental experience and are more prone to manifest fear and anxiety in dental settings. Dentists also report children’s behavior is becoming more challenging to manage and parents are increasingly sensitive to their child’s dental experience.⁶

Sedation is a continuum where typically a more potent medication or higher dose deepens the level of sedation from mild, to deep and possible general anesthesia (GA). At profound levels, child will become unresponsive and unable to breathe on their own or maintain cardiovascular functioning.⁶ If respiratory compromise is not addressed quickly, respiratory arrest then causes cardiac arrest. Children are more susceptible than adults to respiratory causes of cardiac arrest due to age related differences in airway anatomy, and respiratory and cardiac physiology.^{7, 8} Sedation-related in-office deaths are almost always due to the inability to resuscitate the child that has suffered cardiac arrest following respiratory compromise.^{9, 10}

During the last decade, office-based PDSs primarily have received attention from mainstream media because of several life-threatening complications and deaths.^{2, 9–11} Despite the rise in PDSs performed in dental offices, little has been documented regarding their safety and overall effectiveness.^{3, 5, 10, 12} Studies exploring the extent and/or nature of sedation-related adverse events (AEs) have examined small sample sizes, considered a single regimen and/or medication. This safety and effectiveness gap has also been referenced in multiple systematic reviews due to data paucity and inconsistency.^{13, 14} No national requirement for reporting outcomes nor mandatory registry for morbidity and mortality related to PDS exists in the US.^{5, 15} Without a national database or consistent reporting researchers often rely on media reports and lawsuits as primary sources and indicators of sedation-related AEs. Lee et al. used media reports from 1980 to 2011 to understand sedation-related morbidity and mortality among US children.¹⁰ The authors discovered 44 deaths reported for those younger than 21, with half being 2 to 5 years old. Additionally, 45% of deaths were reported for sedations considered moderate depth. Similarly, the authors found that among the 14 closed malpractice insurance claims involving PDSs between 1993 and 2007, 6 involved deaths and 1 involved neurologic injury.¹⁶ While malpractice claims are filed in order to seek financial compensation for harm resulting from medical care, they also serve as data source to investigate factors associated with medical care-related AEs. However, malpractice claims analysis are limited by selection bias, as not all AEs result in malpractice claims.¹⁶ Casamassimo articulated in a 2018 editorial advocating for safety in pediatric dental practice – “We just don’t know what we don’t know!”.¹⁵

Besides safety in pediatric dental sedation, dental sedation effectiveness in children is poorly reported.¹⁷ Studies investigating the efficacy and effectiveness of PDSs mainly rely on the ease of treatment delivery and other treatment completion indicators.^{17, 18} Patient and provider satisfaction, reduction in anxiety or other patient centred measures are rarely used. ¹⁷Mittal et al (2015) consider sedation as successful when the planned treatment is not only completed without any intraoperative AEs but also without any patient interference, movement or crying.¹⁸ Thus, a more patient-centered approach should prioritize the experience and potential psychosocial impact on patient, parents, and providers. Sedation has enabled dental treatment of anxious children who otherwise may not have been easily treated using conventional methods. However, sedation safety and quality is essential to the delivery of high quality dental care.

Electronic Health Records (EHRs), and chart reviews, have been successfully used to investigate quality of care and identify and quantify AEs.^{19, 20} This study examined a university-based dental EHR database in order to better understand the safety and quality of PDSs.

METHODS:

This study was a retrospective analysis of PDSs performed at the UTHealth School of Dentistry in Houston, Texas. The study was approved by its IRB (HSC-DB-12–0662: Developing a Patient Safety System for Dentistry).

The BigMouth²¹ Database, a multi-center dental database repository, was screened to identify pertinent patient records at one dental institution. Inclusion criteria were children younger than 18 years old who had undergone oral, parenteral sedation, and GA during 2019. Patient EHRs were then accessed, and all dental sedations performed on the identified patients were reviewed, including sedations undergone by patients prior to and after 2019. All GAs were excluded from the analysis during patient chart reviews. Patient chart review and data collection were performed between February and July 2020. Both structured and unstructured relevant data were retrieved in a de-identified manner, captured and managed using REDCap²² secure electronic data capture tools. Sedation-related AEs were categorized using a novel 12-point classification system available in a check box format with narrative free text/comments previously developed by Kalenderian et al.²⁰ The Tracking and Reporting Outcomes of Procedural Sedation (TROOPS) Comprehensive Research Tool developed by the International Committee for the Advancement of Procedural Sedation was also used to further stratify the AEs recorded.²³ This latter tool is an intervention-oriented tool that only records AEs associated with unplanned interventions and/or undesirable outcomes. The TROOPS is a revision of the International Sedation Task Force of the World Society of Intravenous Anesthesia which is a reporting tool and itself an update of the Quebec Guidelines tool for standardizing and reporting sedation-related AEs.^{24, 25} With both tools, all events were reported separately. Of note, patients may have encountered more than one category of AE, as well as more than one type of AE within a given category.

Data were exported from RedCap to Microsoft Excel™ for analysis. A descriptive analysis was performed to include: (i) Demographics including age, gender, and race/ethnicity, The American Society of Anesthesiologists (ASA) physical status classification, and insurance status (ii) Sedation mode and regimen, (iii) Sedation-related AEs, and (iv) Patient behavioral outcomes during sedation. The primary outcome analysis of this study was the incidence and nature of AEs and observed behaviors during dental sedation and through recovery prior to discharge. Sedation-related AEs were defined as any “unexpected and undesirable response(s) to medication(s) and medical intervention used to facilitate procedural sedation and analgesia that threaten or cause patient injury or discomfort”.²⁴ Patients behaviors were categorized by the presence/absence of agitation after administration of sedatives. In this study, patients’ agitation was defined as any manifestation of discomfort, stress, and anxiety including but not limited to mood changes, irritability, restlessness, excessive cognitive and motor activity, and combativeness occurying during sedation. During patient chart review and data collection, agitation was considered present when providers’ progress notes in the EHR mentioned patient was whining, crying, screaming, moving head or body, or refusing to cooperate. The inability to complete the planned dental procedure under sedation due to patient’s agitation was considerd a failed sedation. See Table 1 for comprehensive list of measures and variables used.

Table 1.

Variables and outcomes

Variable/Outcome	Description
Sedation-related adverse event	“Unexpected and undesirable response(s) to medication(s) and medical intervention used to facilitate procedural sedation and analgesia that threaten or cause patient injury or discomfort” (World SIVA)24
Agitation	Any manifestation of discomfort, stress, and anxiety including but not limited to mood changes, irritability, restlessness, excessive cognitive and motor activity, aggressivity, and combativness
Failed sedation	Inability to initiate or complete the planned procedure due to patient’s agitation
Ineffective sedation	Difficulty to provide dental care due to child’s agitation. Although dental treatment was provided, sedation was considered ineffective by the dentist provider.

RESULTS

Demographics

A total of 363 patient records were reviewed. Twelve cases were excluded due to missing sedation data; leaving 351 patient records available for analysis. Of these 351 children, 150 (42.7%) were below the age of 6, 133 (37.9%) were between 6 and 12 years old, and 68 (19.4%) were 12 years old and above (Table 2). The mean age of the patients was 7.4 years old. Gender distribution was as follows; 169 females (48.1%) and 182 males (51.9%). All were classified ASA I or II (Table 2). Race/ethnicity of the study population were categorized as follows: 170 (48.5%) Hispanic or Latinx, 61 (17.4%) Black or African-American, 46 (13.1%) White, 13 (3.7%) Asian, 1 (0.3%) American Indian or Native, 8 (2.3%) mixed-race, and 52 (14.2%) unknown (Table 2). Furthermore, most patients had a form of government insurance; 248 (70.7%) patients had Medicaid or Children’s Health Insurance Program (CHIP), 32 (9.1%) patients had a private insurance, and the remaining 71 (20.2%) patients did not have any form of dental insurance (Table 2).

Table 2.

Patient Demographics

Variables	Values N= 351
Age, yrs. (mean ± SD)	7.4 ± 4.1
Age, yrs.
Min	2
Max	17
Age, yrs. n (%)
< 6	150 (42.7)
6 – 12	133 (37.9)
≥ 12	68 (19.4)
Gender, n (%)
Female	169 (48.1)
Male	182 (51.9)
Race/Ethnicity, n (%)
White	46 (13.1)
Black or African American	61 (17.4)
Hispanic	170 (48.5)
Asian	13 (3.7)
American Indian and Alaska Native	1 (0.3)
Two or more races	8 (2.3)
Not mentioned	52 (14.2)
ASA*, n (%)
I-II	351(100)
III and above	0 (0)
Insurance, n (%)
Private	32 (9.1)
Public	248 (70.7)
None	71 (20.2)

^*ASA: The American Society of Anesthesiologists (ASA) physical status classification

Sedation regimen and pharmacologic agents

Of the 351 patient records investigated, a total of 690 dental sedations were reviewed indicating the use of multiple sedation visits for some children to complete their dental treatment. 153 (43.6%) children received at least one sedation, 111 (31.6%) received two, 49 (14%) received three, and 38 (10.8%) received between four and eight sedations over time. On average, each patient underwent two sedation visits over time. Three different administration routes were used to sedate patients: 130 (18.2%) sedations were completed with nitrous oxide (N₂O) inhalation alone, 360 (50.5%) with oral sedatives, and 200 (28.0%) with parenteral sedatives. Regarding oral sedation, different sedative agents were used alone or in combination. Four different oral sedation drug regimens were employed: 40% of oral sedations were performed using meperidine and hydroxyzine, 27% using midazolam and hydroxyzine, 19.4% using meperidine, midazolam, and hydroxyzine, 7% using diazepam and hydroxyzine, 3% using midazolam alone, 1.6% with hydroxyzine alone, 1.1% using diazepam alone, and 1% using midazolam and meperidine. Regarding parenteral sedation, multiple drug combinations were used depending on children’s behavior, type and length of dental procedure. Ketamine and midazolam were the primary agents used: IM Ketamine was used in 42% of parenteral sedations and IM midazolam in 16%.

Sedation-related Adverse Events:

Sedation-related adverse events were found to be low in this study.

Overall, 28 AEs were identified. Emesis was the most common AE and was observed in 9 (1.3%) sedations, followed by over-sedation in 6 (0.9%), blood pressure fluctuation in 3 (0.4%), low oxygen saturation in 3 (0.4%), excessive lung mucus secretion/production in 2 (0.3%), hypothermia in 2 (0.3%), allergy in 1 (0.1%), partial laryngospasm in 1 (0.1%), and transient ST-elevation in 1 (0.1%). The majority of AEs occurred during oral sedation. 19(68%) AEs were observed during oral sedation, 6 (21.4%) during parenteral sedation, and 3(10.7%) during N₂O. See Table 4 for overview. Among the 19 AEs identified during oral sedation, 11 (58%) were identified when Midazolam was used either alone or in combination.

Table 4.

Outcomes of Patients’ Agitation by Sedation Mode^*

		Sedation mode			Total
	Outcomes	N2O	Oral	Parenteral
Agitation	Failed sedation, n(%)	25(22.3)	82(73.2)	5(4.5)	112
	Ineffective sedation, n(%)	1(1.5)	51(77.3)	14(21.2)	66
	No undesirable outcome recorded, n(%)	40(26.7)	95(63.3)	15 (10)	150
Total		66(20.1)	228(69.5)	34(10.4)	328

^*The percentages presented in Table 4 show the proportions of agitation outcomes for each sedation mode to the total number of that specific outcome. For example, failed sedations in oral sedation represent 22.3% of the total failed sedations, i.e. 82/112.

Of the 28 AEs recorded, 8(28.6%) resulted in one or more unplanned interventions and/or undesirable outcomes. Four case of emesis resulted in escalation of care with suctioning and positive pressure oxygen supplementation. Three cases of over sedation resulted in escalation of care where patients stayed longer in the dental clinic under surveillance. The partial laryngospasm case was associated with a parenteral sedation and resulted in positive pressure oxygen supplementation and a longer stay under surveillance before dismissal. No undesirable outcome nor intervention were recorded and identified by the reviewer in the remaining 20 (71.4%) AEs. Overall, of the 28 sedation-related AEs recorded using the 12-point classification system, only the 8 AEs associated with unplanned interventions and/or undesirable outcomes were captured by the TROOPS Comprehensive Research Tool.

Behavioral Outcomes:

Agitation was identified in 328 (47.5%) sedations. Patients were found to have a documented record of agitation in 228 (63.3%) oral sedations, 66 (50.7%) N₂O, and 34 (17%) parenteral sedations. Agitation resulted in failed sedation in 112(34%) cases, and ineffective sedation in 66(20%) cases. Most failed sedations were recorded during oral sedations: 73.2% of failed sedations were recorded in oral sedation, 22.3% in N₂O, and 4.5% in parenteral sedation. Additionaly, 77.3% of ineffective sedations were recorded in oral sedation, 21.2% in parenteral sedation, and 1.5% in N₂O sedation. Table 5 shows agitation analyzed by outcome type. In contrast, Table 6 presents agitation analyzed by sedation mode. When considering the total number of oral sedations, 1 out of 5 oral sedations was a failed sedation, and 1 out of 7 was considered ineffective. Similarly, 1 out of 5 N₂O sedations was a failed sedation (Table 6).

Table 5.

Behavioral Outcome (presense/absence of agitation) by Sedation Mode ^*

	With Agitation			Without Agitation
			No undesirable outcome recorded
	Failed sedation	Ineffective sedation
Sedation mode
N2O (n=130)	25(19.2)	1(0.8)	40(30.8)	64(49.2)
Oral (n=360)	82(22.8)	51(14.1)	95(26.4)	132 (36.7)
Parenteral (n=200)	5(2.5)	14(7)	15(7.5)	166(83)
Total (N=690)	112(16.2)	66(9.6)	150(21.7)	362(52.5)

^*The percentages presented in Table 5 show the proportions of agitation and non-agitation cases for each sedation mode to the total number of cases of that sedation mode. For instance, agitation resulting in failed sedation represents 22.8% of the total oral sedation cases, i.e. 82/360.

DISCUSSION

The most common AE was emesis identified in 1.3% of all sedations and representing 32.1% of all identified AEs. Of the nine emesis cases, five cases resolved spontaneously with no intervention while four cases needed positive pressure oxygen supplementation and suctioning. Respiratory AEs were identified in 0.7% of all sedations representing 17.6% of all identified AEs. Three low oxygen saturation cases and one case of excessive lung secretions recovered spontaneously, one laryngospasm case needed positive pressure oxygen supplementation and was kept under surveillance before discharge. Cardiovascular AEs were identified in 0.6% of all sedations representing 14.3% of all identified AEs. All three cases of blood pressure fluctuation and one case of transient ST elevation recovered spontaneously without escalation of care. Numerous studies have shown that a low, but consistent, incidence of respiratory and cardiovascular AEs persists during pediatric sedation.^{5, 26, 27} This study corroborates these findings. However, just because these AEs are rare does not mean they are not important. If an estimated of 100,000 to 250,000 dental sedations are performed yearly in the US², and assuming perfect inference, this could translate to 1300–3000 respiratory and cardiovascular AEs yearly. Similar to all rare events, the statistical power of the current investigation on AEs remains limited and studies involving large samples from different delivery settings are needed, hopefully enabled by a unified reporting tool and repository of all sedation-related AEs across different sites.

In this study, two distinct tools were used to asses the safety of PDSs. While one tool, the 12-point classification system, was event-oriented and permitted identifications of all AEs independently of their clinical outcomes, the other (TROOPS) was intervention-oriented and captured only AEs associated with unplanned interventions and/or indesirable outcomes. The methodology not only identified AEs related to PDSs but also compared both approaches and their suitability to dental settings. Although the TROOPS tool was developed to standardize the reporting of sedation-related AEs for research purposes, it generates numerous controversial questions(e.gs., in order to count an event, does it have to reach the patient? Is an event that resolves spontaneously less significant thus not worth reporting?)^{12, 28}

Patients’ agitation was identified in 47.5% of all sedations. Documentation in the EHRs indicated patients were agitated during 63.3% of oral sedations, 50.8% of N₂O, and 17% of parenteral sedations. In the literature, agitation may include a broad spectrum of behavioral reactions ranging from whines, cries, screams, and restlessness to mood changes, hallucinations, and physical combativeness.^29–31 The Richmond Agitation-Sedation Scale (RASS) has been found to reliably assess agitation in critically ill children and adults under sedation,^{32, 33} as well as in adult patients about to undergo major surgery.³⁴ Agitation was defined in the population under review, as any manifestation of discomfort, stress, struggle, anxiety and non-cooperation observed after administration of sedatives. This project was a secondary analysis of EHR data, and as such we relied on provider progress notes to assess patients behaviors and levels of agitation after the administration of sedation. The cases of agitation if scaled would likely fall somewhere between +1 and +4 on the (RASS),³² although it is acknowledged that the RASS has neither been validated for outpatient surgery nor pediatric dental surgery. In this study, agitation resulted in both failed and ineffective sedation. Agitation was also confirmed to occur independently from both, suggesting that a great number of dental treatments under sedation continued to be performed despite patient’s agitation.

Failed sedation for this study purposes was defined as the inability to initiate or complete a planned dental procedure due to patient’s agitation. Ineffective sedation was defined as the provider’s perceived difficulty to provide dental treatment due to patient’s agitation although dental treatment was completed. The results revealed that 1 out of 5 oral sedations was a failed sedation and 1 out of 7 oral sedations was considered ineffective by the dentist provider. Huang et al (2015), investigated pediatric oral sedation AEs and found that an oral route was ineffective in 23.9% of the cases.²⁹ Similarly, Olacke et al (2018) found that 26.4% of oral sedations completed among 427 two-to 12-year-old children were unsuccessful due to patients lack of cooperation during sedation.³⁵ Moreover, the results of this study showed that that 1 out of 7 N₂O cases was considered a failure because of patient’s agitation during dental treatment and is consistent with prior literature.^36–39 Nevertheless, assessing sedation success only on the basis of the provider’s ability to complete treatment is limited and noninclusive of the patient perspective. Hence, treatment completion is only a component of the interpretation of the success of sedation and the patient’s level of discomfort and stress during sedation should not be neglected. Also, a provider’s threshold to proceed with dental treatment when a patient is agitated is likely to vary depending on providers’ demographics, psychosocial factors (e.g., provider’s self-efficacy, beliefs…), past experiences, nature of dental treatment, urgency of care, and skill level. This study showed that in 64% of sedations where agitation was observed, dental treatment was completed despite patient’s discomfort or stress.¹⁸ Agitation is often expected during pediatric dental sedation procedures.^{30, 31} The question is to which degree is agitation actually expected and where is the limit between an expected and an unexpected agitation? Also, how does agitation and at what level of agitation does it become an issue and for whom? This study showed that agitation interferes with the performance of dental procedures and can result in dental treatment cancelations. These cancellations have wide ranging implications such as financial and time costs for children, parents, medical institutions, and third-party payers.⁴⁰ Most likely failed procedures are either rescheduled with different medication, deeper sedation, or general anesthesia. Delaying care further burdens the family by lost productivity and more travel time. Additionally, the disease process predisposes the patient to greater risk of discomfort. Ironically, studies have shown previous negative dental sedation experiences are more likely to reduce the success of subsequent sedation appointments.⁴¹ Not surprisingly, parental satisfaction about dental care is positively correlated with the child’s cooperative behavior during dental treatment. In fact, the longer the period of quiet behavior during dental treatment is, the higher the parental satisfaction.⁴² Lastly, failed sedations are stressful for the dental staff involved. Uncooperative behavior in children has been linked to low satisfaction and stress among dentists.^{42, 43} This is another reason to identify and capture the underlying factors behind failed and ineffective sedations.

LIMITATIONS

This is an observational retrospective single academic clinical site study. Thus, generalization should be performed with caution. Overall, as the study was conducted at only one site, the number of PDSs cases were limited and a similar analysis using data from multiple institutions may reveal different findings. Due to the overall low incidence of AEs found, it is also difficult to account for the impact of multiple sedation approaches, medications used, various dental treatments delivered, and different providers’ profiles involved in the sedation cases investigated. Furthermore, a retrospective review of records has other inherent limitations. One major limitation is missing data in patients EHRs. Depth of sedation and some patients’ documentation (i.e. informed consents, vitals) were missing for some sedation cases, posing the issue of data completeness. Additionally, much of the data was unstructured and was in text form in providers’ electronic notes and handwritten notes on scanned sedation reports. The process of data extraction was difficult, making an in-depth chart review process not only more complex and time-consuming but also more prone to transcription error.

CONCLUSIONS

Based on this study’s results, the following conclusions can be made:

1. Potentially serious AEs were identified during pediatric dental sedation, but their incidence was low and they were managed without any complications.

2. A significant proportion of children experienced agitation while sedated. Further research should define, capture, measure, and uncover the occurrence of agitation during PDS to better understand its impact on the safety and quality of the care delivered.

3. Such events need to be continuously tracked and examined for risk assessment reduction, patient safety, and quality of-care improvement.

Table 3.

Adverse events identified with the 12-point classification system ^*

		Sedation Mode			Total
		N2O	Oral	Parenteral
Adverse event category	Adverse event
Allergy/Toxicity/Foreign Body Response	Allergic reaction to surgical tape used for monitors	0	1	0	1
Systemic Harm/Complications	Hypothermia	0	2	0	2
	Excessive lung secretions	0	0	1	1
	Blood pressure fluctuation	0	2	1	3
	Emesis	2	5	2	9
	Low oxygen saturation	0	3	0	3
	Transient ST elevation	0	1	0	1
	Partial Laryngospam	0	0	1	1
Other Harm/Complications	Over sedation	1	4	1	6
	Other	0	1	0	1
Total,-n(%)		3(10.7%)	19(67.9%)	6(21.4%)	28(100%)

^*The 12-point classification system used to identifie adverse events was previously developed and published by Kalenderian et al (Journal of Patient Safety 2021, 17(6): e540–e556). This classification system uses the word “harm” to define and categorize adverse events.