A 19-year-old male patient came to the department of pediatric dentistry of our hospital in order to get prosthetic treatment for his right mandibular posterior teeth. The patient was given a temporary bridge after preparation of the right mandibular posterior teeth under general anesthesia at our hospital about 2 months ago. On the day of presentation, final preparation and impression for the final bridge were scheduled. The patient had autism, with no other relevant medical history. His weight was approximately 90 kg; height was approximately 180 cm. He was overweight for his age. The degree of autism was so severe that he could not stand still and stubbornly resisted from entering the clinic. In addition, the muscular strength of the patient made it tougher to control his behavior. We decided to administer 10 mg of midazolam via the intramuscular route. However, the patient's resistance to intramuscular injection made it difficult to do so. As the patient was overweight, sedation via the oral route could not be maintained, and he resisted taking oral pills. Moreover, even intravenous injection was difficult, and there were no professionals and monitoring equipment to perform sedation via the intravenous route. Hence, as an alternative, we decided to try sedation via the submucosal route. First, protective stabilization for the patient who had strong involuntary movements was arranged using physical restraint. Then, local infiltration anesthesia was administered with about 0.6 ml of 2% lidocaine (1:100,000 epinephrine) to the left maxillary vestibule, and midazolam was injected into the submucosa at the same spot. The total dose of midazolam was 9 mg (5 mg/ml; 0.1 mg/kg). After about 5 minutes, the patient was under sedation. The level attained was that of conscious sedation, and the airway remained under the patient's voluntary control. During the treatment, he turned his head or moved his body, both of which did not interfere with the treatment. SpO₂ was continuously monitored during the treatment, and O₂ was continuously supplied through the nasal mask. The duration of the dental procedure was about 40 minutes, and the treatment was safely completed without any side effects of midazolam under good coordination. Approximately 30 minutes after the completion of the treatment, the patient recovered from calm state. After confirming that there was no physical problem, we explained the remarks, and allowed him to go home.

This study reports the case of a patient with autism whose behavior was difficult to manage. The patient displayed resistance to sedation via the oral, intramuscular, and intravenous routes. Thus, the patient was quickly and effectively stabilized by sedation via the submucosal route (Fig 1).

The effect of sedation via the submucosal route can be elucidated from comparative studies on other sedation routes. According to Lee et al. [7], sedation via the submucosal route was more effective than sedation via the oral, intramuscular, and intravenous routes. That is, on comparing submucosal midazolam (0.1–0.2 mg/kg) mixed with oral chloral hydrate (50–70 mg/kg) and oral chloral hydrate (50–70 mg/kg) alone, the combination had better behavior response rating and less vomiting response than oral chloral hydrate used alone. According to Alfonzo-Echeverri et al. [6], after comparing the differential absorption rates of midazolam (0.08 mg/kg to a maximum of 5 mg) in sedation via the submucosal route and the intramuscular route, there was no big difference in average absorption; however, sedation via the submucosal route reached the maximum absorption rate at 10 minutes on an average, and sedation via the intramuscular route reached the maximum absorption rate at 20 minutes on an average. These reports demonstrate that sedation via the submucosal route has better effectiveness and shorter onset time than sedation via the oral and intramuscular routes.

Sedation via the submucosal route shows a more profound effect and shorter working time compared to sedation via the oral route because it is not affected by the first pass effect, that is, the sedative does not pass through the liver [8]. In the case of sedation via the oral route, the sedative loses its effect as it passes through the liver and the gastrointestinal tract by the first pass effect and thus exhibits a longer onset time [9]. In contrast, in the case of sedation via the submucosal route, it directly goes into the cardiovascular system through intraoral veins, without passing through the liver [10].

The shortcoming of sedation via the submucosal route is the discomfort at the injection spot and the vomiting reflex. Moreover, there can be undesirable effects such as respiratory depression, despite the short working time and increased effectiveness. A case of serious side effects along with respiratory depression after sedation via the submucosal route has been reported. It was managed using Alphaprodine, an opium material [11]. However, it is not used anymore, and midazolam, which was used in this case, is safe and effective [12]. Furthermore, flumazenil, a benzodiazepine antagonist, may be used to reverse the effect of midazolam [13]. Flumazenil can reverse the effect of benzodiazepine drugs including midazolam by competitively binding to the GABA receptor, on which benzodiazepines act [14]. It is mainly injected intravenously, but can also be injected into muscles and submucosa; the recommended dose for intravenous injection is 0.3–0.6 mg. According to Schmitt et al. [15], sedation via the submucosal route was associated a greater local irritable reaction and pain compared to oral sedation; however, the differences were not considerable. In this case, even if the patient had local discomfort, it was much better to use sedation via the submucosal route, as no alternatives could be used.

Unlike this case, not every patient with autism has difficulties in being kept under sedation [16]. However, in this case, the patient was overweight for his age and had high physical strength and resistance. He also displayed autistic symptoms that resist relationships with others and changes in the surrounding environment; this in turn, made it difficult to perform sedative procedures other than sedation via the submucosal route.

In general, submucosal injection is not as technically feasible as intramuscular injection. The reason why submucosal injection could be administered, while intramuscular and intravenous injections could not, was that submucosal injection can be administered using physical restraints such as Pedi-Wrap. Use of physical restraints can hinder the access for intramuscular and intravenous injections. Inpatients in hospitals can be restrained using straps to administer intramuscular and intravenous injection; however, strings or beds are not available in dental clinics. Besides, in the case of intramuscular and intravenous injection, there is difficulty in exposing the injection spot when performing protective stabilization using physical restraints such as Pedi-Wrap. Especially in cases like this, it becomes more difficult to expose the site for intramuscular and intravenous injection under protective stabilization for patients with autism, who have involuntary movements and exhibit resistance with physical strength.

When sedation via submucosal route is performed, it is preferable to use local anesthetics without vasoconstrictors, as the vasoconstrictor may delay the effect of the sedative [17]. In this case, lidocaine containing epinephrine, which is mainly used in dental clinics, was used because sedation via submucosal route was not planned from the beginning, and it was necessary to perform local anesthesia quickly after restraining the patient. He was considered not to be affected by small amounts of vasoconstrictor because he was overexcited to cause vasodilatation, and overweight.

This study showed that sedation via the submucosal route is an effective way for managing the behavior of patients with autism in situations where other types of sedation are not available. This study highlights the importance of understanding the individual clinical situation and the potential of sedation via the submucosal route. Thus, sedation via the submucosal route is expected to be more frequently used for patients with autism who cannot be sedated by other routes, patients with other disabilities, and children. Follow-up studies about sedation via submucosal route in various situations are expected to be presented.