Study design

The study was designed as a randomized, double-blind, cross-over clinical trial (registered in www.clinicaltrial.gov with the code NCT01118884) on 18 children (totally 36 clinical sessions) aged 3–6 years who were rated as Category 1 or 2 on the Frankl behavior rating scale, which means showing negative or definitely negative behavior. ² , ¹⁹ Sample size calculation was performed by power analysis software (PASS II) using McNemar's test with α (Type I error) = 0.05, β (Type II error) = 0.2, and proportion discordance of 0.6, which resulted in a minimum of 18 patients. Participants were recruited out of 59 children referred to Mofid Children Hospital, Tehran, Iran, during November 2015–June 2016, by simple sampling method Figure 1. Patients with American Society Anesthesiologists I, with bilateral and identical restorative treatment needs, requiring 2 or more dental visits and no previous dental sessions, were selected. The exclusion criteria included a history of gastrointestinal tract disorder, renal and hepatic impairment, known drug hypersensitivity, and upper respiratory tract infection or obstruction, which made breathing through nose difficult. Figure 1

Participant flow chart.

As demonstrated in Figure 1, of 59 children assessed for inclusion in this study, 35 children did not meet the inclusion criteria or met any of the mentioned exclusion criteria, six parents were unwilling to take part leaving 18 children. The mean age of the participants (9 [50%] boys and 9 [50%] girls) was 49.16 ± 14.9 (range: 35–73) months, and the mean weight was 15.57 kg (range; 11.85–29).

Intra- and inter-rater reliabilities were established as κ = 0.84 and κ = 0.91, indicating high and excellent agreement, respectively.

Physiologic parameters

As demonstrated in Table 2, there was no statically significant difference in OS and HR among the two groups at the times (t) = 0, 10, 20, 30, and discharge (P > 0.05). The lowest and highest OS was 91% and 100%, respectively, which were seen in both groups. The least HR (79 beats/min) was observed in one patient at t = 0 in midazolam group, whereas the highest HR was reported in one of the children in promethazine group at t = 20 (171 beats/min).{Table 2}

Behavior evaluation

The behavior of each patient was evaluated according to the four categories of Houpt behavioral scale; alertness, movement, crying, and the overall behavior of participants at the first and last phases of each treatment.

Sleep (alertness)

During the first 15–20 min, the majority of participants (94.4%) sedated with midazolam were drowsy (Code 2); in contrast, only 44.4% of children sedated with promethazine were classified as drowsy during the same period Table 3.{Table 3}

In the last 20 min of the experiment, 77.7% of children sedated with midazolam fell in code 2 of sleep rating, which was significantly different compared to children treated with promethazine (44.4%) (P < 0.05). None of the children were categorized as asleep (Code 3) following treatment with either midazolam or promethazine Table 3.

Movement

As shown in Table 4, Houpt movement scale in the first and last 15–20 min showed no significant difference between the two drug regimens (P > 0.05).{Table 4}

Crying

Within the first phase, most of the children in midazolam group showed no crying or mild crying (88.9%). On the contrary children sedated with promethazine showed more intensive crying throughout the first 20 min, resulting in a significant difference between the two groups during this time period (P < 0.05). On the contrary, during the last 20 min, we found no significant difference between the two treatment regimens regarding crying scores (P > 0.05) Table 5.{Table 5}

Overall behavior

As demonstrated in Table 6, children in midazolam group showed significantly superior overall behavior compared to those sedated with promethazine, in the first phase (P < 0.05). However, no significant difference was seen during the last phase between the two groups (P > 0.05).{Table 6}

Acceptability of the methods

Parents' and patients' satisfaction is illustrated in Figure 2and Figure 3. Children sedated by midazolam expressed lower anxiety during the treatment in comparison to children in promethazine group (P < 0.05). However, the majority of parents were satisfied with both regimens (P > 0.05). Figure 2

Feelings experienced by patients during the treatment. Significant difference was found between the two groups ( P= 0.042).

Our investigation compared safety and efficacy of two combination sedative regimens. In the present study, we added N ₂O either to promethazine or midazolam in two following sessions. Utilizing the combo sedation in this study was aimed to benefit from the synergetic effects of the drugs in obtaining optimum sedation and cooperation.

In the current study, an onset time of 30 min for midazolam and 45 min for promethazine was designated as optimal before initiating dental treatment. This was in accordance with other researches with the same drugs. ¹ , ² , ¹⁴ , ²¹ This time interval was based on the required latent period necessary to reach the maximum peak plasma level of the drug. ¹ , ²²

In the present study, 0.5 mg/kg midazolam, 1 mg/kg promethazine, and 50% N ₂O/50% oxygen were administrated, which was in line with previous investigations. ¹ , ² , ⁸ , ¹⁴ Adding N ₂O is permitted to all sedation methods, leading to an increased sedation and improved oxygen delivery to the patient. ² A concentration of 50% N ₂O–50% oxygen provides safe anxiolytic/analgesic effect by the activation of opioid and gamma-aminobutyric acid receptor. ² , ⁸ , ¹¹ Preceding researches advocated 0.5 mg/kg midazolam, as no difference was seen in sedating efficacy of 0.5, 0.75, and 1 mg/kg midazolam and some side effects were reported with 1 mg/kg midazolam. ¹ , ² , ²³

Our results confirmed the overall safety of both regimens based on OS and HR of participants. The OS levels were comparable in two groups and did not surpass the safe range. ¹ The lowest OS in our research was 91%, which is defined as mild hypoxemia. Mild hypoxemia (POS: 91%–95%) does not alter other physiologic parameters such as respiratory, HR, and blood pressure, thus, needs no intervention or treatment. ²⁴ The episodes of OS below 95% were transient and corrected immediately by head repositioning. Lindh Stromberg similarly showed a higher than 92% OS in 60 children sedated by 0.3 mg/kg rectal midazolam. ²⁵ Chowdhury and Vargas attributed the transient oxygen desaturation (lower than 95%) to the violent movements and crying of the uncooperative children rather than the pharmacologic action of sedative agents. ²² In an investigation conducted by Needleman et al., children who were moderately to heavily sedated, demonstrated OS <95%. They concluded the level of sedation is an essential factor in oxygen desaturation. ²⁶ In our study, all the participants were fully awake or drowsy.

Statistically, we found no significant differences in the HR of the two groups during the sedation procedures and the average HR in both groups were within the normal limits. The highest pulse rate observed in the present investigation was 171, which exceeded the physiologic rate of 130 beats/min (bpm). However, this HR is not considered as life threatening; as in physiologic conditions, the HR may pass 170 bpm during struggling or crying. ¹ , ²⁷ Similarly, Vahid et al. reported normal HR and OS in sedation with a combination of promethazine and midazolam. ³

In the present study, both of the medications proved efficient sedative results. As 89.2% of children in midazolam group and 72.3% of children in promethazine group in initial phase and 77.8% of children sedated with midazolam and 66.7% of children medicated with promethazine in final phase were good-to-excellent behaving according to Houpt scale. ²⁰

The overall behavior in the first and second phase of the treatment was different between the two groups in the current investigation. During the initial phase of the treatment, children in midazolam group demonstrated significantly superior behavior compared to promethazine. In a comparison, among three sedation drugs; midazolam, promethazine, and triclofos, conducted by Singh et al., they also reported significantly a better sedation effect by midazolam compared to two other sedative medications during the treatment procedure. ¹⁸

In contrast to the first phase, there was no significant difference between the two groups in terms of overall behavior in the present study. This was predictable, as more children in midazolam group demonstrated behaviors classified as mild crying and continuous crying during the last time period compared to the first part of the study, while children in promethazine group showed similar overall and crying behavior during the first and second treatment phase. This may be attributed to the short half-life of midazolam compared to promethazine, which results in a shorter duration of action for midazolam (30 min) versus promethazine (4–6 h). ² , ⁴ , ¹³ Wilson et al. stated that midazolam has a short half-life, which causes minimum crying and struggling during the first 10–15 min following local anesthesia; thus, they suggested midazolam as an ideal choice in short procedures. ²⁸

Regarding the sedation level, we observed a deeper sedation during the dental treatment with midazolam compared to promethazine. Almost 94.4% of children in midazolam group were drowsy (Code 2) during the first 20 min. As a result of short time of action for midazolam, the proportion of children with drowsiness decreased (77.7%) during the last phase. In contrast to midazolam, most of the children in promethazine group remained fully awake in both time periods. This may also explain the superior overall behavior in sedation with midazolam compared to promethazine in the first time period and their insignificant difference in the last phase of treatment. We used VAS to investigate the level of anxiety of children following each treatment session. VAS is a simple and reliable means in evaluating dental anxiety with a 5-point Lickert scale and scores ranging from “relaxed/not anxious” to “very anxious” including five cartoon faces. ²⁹ Our results showed a lower anxiety level among children sedated with midazolam compared to promethazine. This may be associated with deeper sedation levels of children in midazolam group, which may affect the overall assessment of the participants positively. ¹² However, the parents did not prefer one drug combination over the other and were totally satisfied with both drug regimens. Although self-report assessments are widely approved in evaluating the acceptability of sedation procedures, the opinion of parents about their children's treatment is very subjective, and thus a more comprehensive questionnaire may be helpful in achieving more accurate and detailed information to overcome this shortcoming. ²³

To decrease the impact of confounding variables and preventing bias, this study was designed in a crossover manner and the operator, patients, parents, and the observer were blinded throughout the study.

The serum concentration of the drug may vary during the treatment session, leading to probable various behavioral reactions throughout the treatment. On the other hand, comprehensive assessment of each phase of treatment may be compromised or even impossible because of clinical demands of the sedation. Thus, video recording the whole appointment may have resulted in a more detailed behavior evaluation in this study. In our research, though parents were asked to report any postsedation complications after discharge, the main purpose of this study was to evaluate the efficacy and safety of the two drug combinations. As postoperative complications may affect the patients' and parents' overall satisfaction regarding the sedative regimens, thus, we suggest a similar investigation on the potential postoperative complications and recovery characteristics of the two medication regimens.

Both of the drug combinations resulted in acceptable, efficient sedation outcomes in uncooperative children and were safe regarding pulse rate and OS. We observed a significantly deeper sedation and improved overall behavior by midazolam in the initial phase of treatment compared to promethazine; however, the overall behavior did not differ significantly between the two medications during the final phase of the procedures.

Financial support and sponsorship

Nil.

Conflicts of interest

The authors of this manuscript declare that they have no conflicts of interest, real or perceived, financial or nonfinancial in this article.