This randomized, prospective, blinded study compared pain in children following dental treatment under general anesthesia (GA) using 1 of 2 established analgesia methods.
Patients age 4 to 7 years were randomly assigned to a control group (intravenous [IV] analgesics) or experimental group (IV analgesics and intrapapillary local anesthetic infiltrations) between July 2017 and February 2018. During recovery from surgery, Faces, Legs, Activity, Cry, and Consolability (FLACC) scores were recorded upon regaining consciousness and reassessed every 15 minutes until discharge. Overall pain occurrence (FLACC ≥1) and moderate/severe pain occurrence (FLACC ≥4) were analyzed using mixed effects logistic regression (N = 88).
The experimental group had a 17% lower overall pain occurrence than the control group (16 vs 33%; p = .02). Moderate/severe pain occurrence between the groups was not significant (9 vs 22%; p = .23). The dental treatment subjects received (number of completed stainless steel crowns, extractions, and/or pulpotomies) did not significantly affect pain occurrence.
Local anesthesia intrapapillary infiltrations around stainless steel crowns decrease overall pain occurrence but not moderate/severe pain occurrence following dental treatment under GA in pediatric patients.
Pediatric pain can be inadequately assessed and often goes untreated.1–3 Many studies address adult dental pain but few focus on children. Clinicians often assume that children do not perceive pain to the same magnitude as adults.1 Clinical trials have demonstrated that parents underestimate their child's pain when compared with the child's self-reported pain.4,5 Additional barriers to appropriate pediatric pain management include false assumptions that: (1) analgesics are not needed, (2) children do not remember painful experiences, and (3) prescription pain medications pose a heightened risk for respiratory depression and addiction in children.1,2 As a result, postoperative pediatric pain management is often inadequate.6
Invasive dental procedures, such as extractions and crowns, can cause postoperative discomfort. One-third of children undergoing such dental procedures reported postoperative pain,7 and 31% of children required analgesics afterwards.8 After general anesthesia (GA), children reported pain from dental restorations, stainless steel crowns (SSCs), pulpotomies, and extractions.7–10 Postoperative pain also reportedly was more intense as the number of dental procedures increased; children undergoing ≥12 procedures reported higher postoperative pain scores than children undergoing <12.9 GA itself can be an additional source of postoperative discomfort, with intubation causing sore throats in 27% of pediatric patients.9,11 Nasotracheal intubation is often utilized and can pose a greater risk of trauma than orotracheal intubation,12 and increased pain with traumatic nasal intubation has been reported in postanesthesia care units (PACUs).13 Due to pain from dental procedures and discomfort from GA, most patients (68–74%) reported discomfort,8 and half (48%) reported moderate-to-severe pain following dental surgery.14 As such, pediatric dental providers share responsibility for anticipating and appropriately managing dental pain,15 so optimal pain management strategies should be studied.
Local anesthesia (LA) is commonly utilized to provide perioperative analgesia, although studies and anecdotal observations have shown use during GA for pediatric dentistry can be uncomfortable, distressing, and predispose children to self-trauma to the oral soft tissues.16,17 Thus, some have advocated for intrapapillary LA infiltration as it produces less soft tissue numbness compared with vestibular infiltrations or blocks.6 Additionally, PACU nurses in 1 study consistently stated patients who received intrapapillary LA appeared to recover from anesthesia more comfortably than those who received LA using other techniques.18 This study investigated postoperative discomfort in children undergoing GA for dental rehabilitation. We hypothesized that intrapapillary LA infiltrations could improve postoperative pain by providing gingival analgesia without associated lip and cheek numbness.
This institutional review board-approved randomized, prospective, blinded study evaluated effects of 2 analgesic techniques on postoperative discomfort following dental rehabilitation under GA. Subjects receiving intraoperative intravenous (IV) analgesics were randomly assigned into 2 groups: an experimental group receiving 2% lidocaine with 1:100,000 epinephrine for intrapapillary infiltrations, and a control group receiving no LA. Subjects in this study consisted of patients who presented to the Dental Surgery Center at Nationwide Children's Hospital following referral for dental treatment under GA as they were unable to tolerate treatment in the office setting. Potential subjects were approached between July 10, 2017 and February 28, 2018 who met inclusion criteria consisting of age 4 to 7 years, American Society of Anesthesiologists class I or II, planned GA with nasal intubation, and treatment in all quadrants on primary teeth including at least 1 SSC. Exclusion criteria included neurological deficits making communicating/reporting pain difficult, legal guardian not speaking English or Spanish, use of oral analgesics (ibuprofen, acetaminophen, etc) the day of treatment,19 preoperative Wong-Baker Faces Pain Rating Score (WBFPRS) >4,19 procedure times >90 minutes, or a medical contraindication to LA use.
Before dental surgery, consent for study participation was obtained from the parent/legal guardian, and the subject completed a WBFPRS, a preoperative measure of self-rated pain and distress that has been validated for children ages 4 and older (Figure 1).20
GA was administered by 1 of 14 unblinded anesthesiologists participating in the study. None of the subjects required preoperative sedation. Standard American Society of Anesthesiologists monitors were placed, and subjects were mask induced with 8% sevoflurane and oxygen per hospital protocol. IV access was established, propofol (at least 1 mg/kg) and dexamethasone (0.1 mg/kg) were administered, and nasotracheal intubation was completed. Anesthesia was maintained with 2 to 2.5% isoflurane and oxygen. IV analgesics (ketorolac 0.5–1 mg/kg; fentanyl 2 mcg/kg; or ketorolac 0.5–1 mg/kg and morphine 0.1 mg/kg) were administered intraoperatively at the discretion of the anesthesiologist. Approximately 15 minutes before completion of the dental procedures, the isoflurane was discontinued, and GA was maintained with intermittent propofol boluses (1 mg/kg). Subjects underwent deep extubation and were transferred to the PACU while asleep with supplemental oxygen after ensuring adequate maintenance of oxygenation and ventilation.
All dental treatment was completed by 1 of 13 operators who was either a pediatric dentist or second-year pediatric dental resident. Before completion of surgery, subjects were randomly allocated to 1 of the 2 treatment groups using sequentially numbered, sealed, opaque envelopes. The experimental group received LA (2% lidocaine with 1:100,000 epinephrine) administered by the unblinded dentist immediately following completion of the dental procedures but prior to throat pack removal. LA injections (0.06 mL) were administered buccally and lingually into the interproximal papillae for each tooth that received a SSC using a Paroject periodontal ligament (PDL) syringe (Septodont, Inc, Saint-Maur-des-Fosses, France). The total LA dose varied based on the number of SSCs a patient received and ranged from 0.24 to 1.44 mL. All dentists were trained to use the PDL syringe prior to their participation. No LA was delivered at extraction sites as it was postulated that intrapapillary infiltrations would not provide sufficient alveolar anesthesia to reduce postoperative pain at that site. The control group was not given any LA.
Postoperative pain was assessed by 1 of 14 blinded and trained PACU nurses using the Faces, Legs, Activity, Cry, and Consolability (FLACC) scale (Figure 2), a validated measure of pain in children between 2 months and 7 years of age that has been successfully used in nonverbal patients or those awakening from GA.21 Scores from each of the 5 categories are summed for a final score totaling 0 to 10. In the PACU, nurses completed a FLACC score upon the subject regaining consciousness and repeated the scoring every 15 minutes until the patient met discharge criteria. FLACC scores over each subject's entire PACU stay were used to determine “pain occurrence” and “moderate/severe pain occurrence” as described below.
Primary outcome measures were pain occurrence and moderate/severe pain occurrence. “Pain occurrence” was defined as FLACC ≥ 1 at any point during PACU stay, and “no pain occurrence” was defined as FLACC = 0 throughout the entire PACU stay. “Moderate/severe pain occurrence” was defined as FLACC ≥ 4 at any point during PACU stay and “no/mild pain occurrence” as FLACC < 4 over entire PACU stay.22 Potential covariates included subject's age (years), sex, preoperative WBFPRS (0–10), dentist, dental procedures (type and number completed), procedure length (PL; throat pack placement to removal), anesthesiologist, IV analgesic (type and quantity administered), time to regain consciousness (from GA per hospital discharge criteria), length of PACU stay, and nasal intubation (difficult or easy).
Statistical analysis was performed with R Version 3.5.1 and JMP (SAS Institute, Cary, NC). Independent t tests and Fisher exact test analyzed similarities in the experimental and control groups with respect to demographics and covariates. Multivariable mixed effects logistic regression methods assessed treatment significance and covariate association with outcome measures. The experimental group assignment (LA or no LA) was designated as 0 or 1 in the data sheet to ensure the data analyst was blinded to the group assignment. All analyses utilized P < .05 for statistical significance. Because the initial sample size power analysis was computed under a different statistical paradigm, a posthoc power analysis was also conducted for overall pain occurrence.
Of the 141 subjects who met selection criteria, 104 consented to participate and 37 declined. There were 16 subjects later excluded due to not meeting treatment inclusion criteria. Of those excluded, 4 had preoperative WBFPRS > 4, 2 did not receive SSCs, 3 did not have 4 quadrants of dental treatment completed, 2 had treatment on permanent teeth, 4 had procedure times > 90 minutes, and 1 was orally intubated. A total of 88 subjects were included in this study.
Study Group Comparison
Of the included subjects, 43 were assigned to the experimental group and 45 to the control group. No significant differences in gender, age, type/number of dental treatments, IV analgesic doses, time to regain consciousness, or length of PACU stay were noted between groups (Table 1). The pediatric dentists and dental residents completed treatment at about the same rate as internal quality assessment studies, and differences in PLs between faculty and residents were not significant (faculty PL = 52 ± 17 minutes, resident PL = 61 ± 25 minutes; P = .06). However, mean PL were significantly different between the experimental (61 ± 22 minutes) and control groups (52 ± 20 minutes; P = .04).
Overall Pain Occurrence
Potential covariates were examined individually for relationship with overall pain occurrence and moderate/severe pain occurrence. Five covariates had P < .2 univariate associations with overall pain occurrence: (1) ketorolac dose, (2) morphine dose, (3) fentanyl dose, (4) time to regain consciousness, and (5) experimental group assignment. These 5 covariates and the “treating dentist” covariate were included in the final analysis for overall pain occurrence. Treating dentist was maintained as a random effect while all other covariates were fixed (Table 2).
All subjects had a minimum of 2 recorded FLACC scores while some had a maximum of 4. Postoperative FLACC scores ranged from 0 to 10 (Figure 3). Of the 22 subjects that had occurrence of pain (FLACC ≥1 at any point during PACU stay), 7 were in the experimental group and 15 in the control group (Figure 4). Experimental group assignment was found to be significant—a 17% lower overall pain occurrence (16 vs 33%; P = .02). The only other covariate found to have a correlation with overall pain occurrence was time to regain consciousness with the experimental group being 2 minutes faster (28 vs 30 minutes; P = .03). Ketorolac, morphine, and fentanyl dosing did not have any significant associations with pain occurrence (Table 2).
Adjusted odds ratios were calculated using the notable covariates, and the 95% confidence interval (0.048–0.734) further supports significantly lower overall pain occurrence for the experimental group (Table 3). A posthoc power analysis performed for 88 subjects with a unidirectional α of .05 to detect a decrease in overall pain occurrence from 33 to 16% (odds ratio, 2.6) yielded a power of 57.8%.
Moderate/Severe Pain Occurrence
Four covariates had P < .2 univariate associations with moderate/severe pain occurrence: (1) time to regain consciousness, (2) PL, (3) length of PACU stay, and (4) experimental group assignment. These 4 covariates and the treating dentist covariate were included in the final analysis for moderate/severe pain occurrence. Treating dentist was maintained as a random effect while all other covariates were fixed (Table 3).
Of the 14 subjects that had moderate/severe pain (FLACC ≥4 at any point during PACU stay), 4 were in the experimental group and 10 in the control group (Figure 4). The 13% lower occurrence of moderate/severe pain in the experimental group lacked significance (9 vs 22%; P = .23). Time to regain consciousness neared significance with the experimental group being 2 minutes faster (28 vs 30 minutes; P = .06). No other covariables were found to have significant associations with moderate/severe pain occurrence (Table 2).
Adjusted odds ratios were calculated using the notable covariables, and the 95% confidence interval (0.045–1.606) further supports this lack of statistical significance (Table 3). The random effect of 2 treating dentists was significantly greater than zero, indicating 2 dentists had significantly higher odds of moderate/severe pain. A posthoc power analysis performed for 88 subjects with a unidirectional α of .05 to detect a decrease in overall pain occurrence from 22 to 9% (odds ratio, 2.8) yielded a power of 49.2%.
After weighing risks and benefits, pediatric patients are commonly referred for dental treatment under GA when unable to tolerate treatment in an office setting. All dental needs are typically addressed in a single event during GA. Due to pain associated with dental treatment, the number of dental procedures completed, and the potential discomfort inherent to GA, it is imperative that providers exercise optimal pain management strategies.
Postoperative pain is reduced when LA is administered intraoperatively during medical procedures.23 Tonsillectomy, appendectomy, orchidopexy, and hypospadias repair have all shown significant improvement in postoperative comfort when intraoperative LA is used.23 In addition, the majority of dentist anesthesiologists (90%) prefer the use of LA during dental rehabilitation.24 A survey of the American Society of Dentist Anesthesiologists members reported common reasons favoring intraoperative LA were as follows: (1) decreased need for intraoperative anesthesiologist intervention (92.9%); (2) improved patient recovery (82.1%); (3) avoidance of deep pain pathway activation (60.7%); and (4) shortened recovery (57.1%).24 Anesthesiologists who did not prefer use of LA (10%) cited the risk of postoperative lip/cheek biting (85.7%) and unnecessary LA pain control if the patient already received systemic analgesics (57.1%).24,25
Despite support by anesthesiologists, little evidence exists that LA improves postoperative comfort following pediatric dental rehabilitation under GA. A handful of studies have shown that LA improves postoperative comfort,6,9,26,27 whereas conflicting studies have suggested LA has no effect15,18 or could even increase postoperative distress.28 Because patients are typically given perioperative analgesics, and because systematic reviews have failed to demonstrate that LA decreases postoperative pain, it remains unclear whether LA is needed to improve postoperative comfort.15 As a result, no established evidenced-based guidelines exist for perioperative LA during dental rehabilitation under GA to prevent pain associated with GA and dental treatment.16
Results of this study support findings of other studies where LA reduced postoperative pain.6,9,26,27 However, this study is unique in the method of LA administration—intrapapillary infiltrations. Previous pediatric studies have investigated LA blocks,29,30 vestibular infiltrations,6,9,15,26 periodontal ligament injections,6,15 and topical anesthetics,31,32 but this is the first study to the authors' knowledge that investigated intrapapillary infiltrations. Our results demonstrated that LA intrapapillary infiltrations are statistically and clinically associated with decreased overall postoperative pain occurrence but not moderate/severe pain occurrence.
Random effects model was chosen to treat the treating dentists in the study as a random sample of the population of all dentists. The random effect for overall pain occurrence was shown to be nonsignificantly different from zero, indicating the treating dentist effect is not significant. Though clinicians assume that dental procedures (eg, posterior tooth extractions and back-to-back SSCs) and IV analgesic dosages (eg, morphine, ketorolac, and fentanyl) affect the occurrence of postoperative pain, our results suggest the use of LA via intrapapillary injections may be more significant. Though the time to regain consciousness was statistically significant for overall pain occurrence, the finding is not clinically significant as the experimental and control groups only differed by ∼2 minutes.
Our results support the use of intrapapillary infiltrations during dental treatment under GA as they provide gingival LA, reduce postoperative pain occurrence, and do not cause discomfort from lip/cheek numbness. Study strengths include stringent selection criteria and the prospective, randomized design with blinded subjects and PACU raters. Although many early studies of pediatric postoperative dental pain examined extractions only, this study included comprehensive dental treatment. It is also the first study to examine LA intrapapillary infiltrations following pediatric dental treatment under GA.
This study also had several limitations. Several dentists, anesthesiologists, and nurses participated in this study. Although multiple providers allocated equally among experimental and control subjects could improve generalizability, they also could introduce variability. For example, the timing and selection of administered IV analgesics varied amongst anesthesia providers. Finally, differences in individual pain thresholds, temperament, emergence delirium phenomena, separation from parents, LA metabolism, LA administration with Paroject syringe, and other factors not measured may have contributed to findings.
A potential criticism of our findings is the short duration of anesthesia from intrapapillary infiltrations (20–60 minutes). Another concern is that some patients did not regain consciousness for some time, which may have affected postoperative pain and assessment of the LA effects. However, no subject in this study reported experiencing pain 30 minutes after regaining consciousness in the PACU (Figure 3), suggesting that the most important pain management period may be within this timeframe, when intrapapillary infiltrations are effective.
An additional consideration is that the experimental design did not control for emergence delirium phenomena, and thereby did not separate that phenomena from pain occurrence. Although most subjects in this study experienced little to no postoperative pain, and our analyses failed to detect significance for moderate/severe pain, our results are clinically relevant if the clinician aims to optimize occurrence of painless recovery. It should be noted that the study was underpowered; however, that does not affect our finding relative to LA's impact on overall pain occurrence. Further, failure to find a significant difference for moderate/severe pain occurrence could be credited to the underpowered study and the relatively rare occurrence of moderate to severe pain. A follow-up study is indicated to increase sample size and improve power. It should also be noted that the random effect for moderate/severe pain occurrence of 2 dentists in this study were significantly greater than zero, indicating 2 dentists had significantly higher odds of moderate/severe pain.
Future studies could investigate at-home follow-up to determine if the gingival trauma from intrapapillary injection affects postoperative pain after LA effects wane. Additional studies could also examine pre-emptive analgesia, which suggests that preventing nociception, even during GA, inhibits future neural hypersensitivity.33 This theory suggests that LA should be administered prior to the onset of the surgical insult (ie, the beginning of the dental surgery), rather than the end of the case.
Local anesthetic intrapapillary infiltrations around SSCs decrease the overall pain occurrence but not the moderate/severe pain occurrence following dental treatment under GA in pediatric patients.
This study was financially supported by the Delta Dental-Dental Master's Thesis Award and was awarded the Ralph E. McDonald Award at the AAPD 2018 Annual Session in Honolulu, HI.