Effect of a Humanoid Robot on Dental Anxiety in Children

Selection and description of participants: 115 healthy children aged between 4 and 10 years, having their first dental visit and requiring a pulpotomy for a primary molar were selected from Istanbul University, Department of Pediatric Dentistry. The children accepting to response questionnaires and informed consent, were accompanied by at least one parent was obtained. Children having any history of systemic diseases or medically compromised were not included to the study. The study protocol was approved by the Ethical Committee of Istanbul University (2014/461).

Inclusion criteria were (a) healthy children, (b) no previous dental experience, (c) need pulpotomy in at least one mandibular primary molar. Exclusion criteria were (a) children showing definitely negative behaviour (Frankl 1) during dental prophylaxis at the first visit.

Randomization: A total of 115 patients who met the inclusion criteria were examined. Six children didn't want to participate for the treatment and five children couldn't spit enough for saliva collection. From the eligible 104 children were randomly selected and equally allocated into two groups.

Treatment procedure: The study protocol was explained to parents and their and written consent was obtained. Pre-treatment questionnaires were asked to the participants. All children were treated under local anaesthesia. The treatments were performed by the same dentist (YK). The robot group (RG) were treated with the robot accompaniment (n=50). The control group (CG) were treated with positive reinforcement and tell-show-do (n=52). Treatments were completed in the same session. One calibrated pediatric dentist [intra-class correlation coefficient (ICC) score = 0.87] performed pulpotomy treatment with inferior alveolar block. In the robot group, distraction and tell-show-do techniques were used by the humanoid robot. The robot used in experiments is NAO, manufactured by Softbank Robotics. NAO is a 58-cm tall robot allowing it to perform the targeted motor tasks. The robot was programmed to perform the same series of instructions for every patient to guarantee that all children had the same experience with the robot. The movements of the robot can be controlled wirelessly from a computer by using Choregraphe software. The investigators prepared the transactions to let the operator run the corresponding commands for each period of the treatment in real time to create sequences of behaviours. After each task is defined individually, it is assigned to a keyboard event. Some combinations of robot movements are achieved by pressing a single key. With this keyboard interface, the learning time of the operator has been minimized. At the same time, the effort to control the robot has been simplified and mistakes have been reduced.

Anxiety status: A series of questionnaires measuring anxiety were administered to each child, before treatment and after treatment. Facial Image Scale (FIS) is a commonly used scale to determine the dental anxiety levels of children. Children were asked to point at FIS before treatment. It consists five faces numbered from 1 to 5 and ranging between "very happy" and "very unhappy". Each child is asked to point to the face that they feel most closely, according to their feelings at that moment. The questionnaire was repeated after treatment. The Children's Fear Survey Schedule-Dental Subscale (CFSS-DS) is used in the assessment of dental anxiety in children. It consists of 15 questions. Children are asked to give scores from 1 (not afraid) to 5 (very afraid) for each question. The total score is between 15 and 75 points. Children with CFSS-DS ≥38 were defined as dentally anxious. After the dental intervention, the children were asked would participants like to see the NAO again in their next appointment, who participated in robot group.

Behaviour assessment: Frankl's Behaviour Rating Scale (FBRS) were used by the dentist to assess each children's behaviour. The child's behaviour is classified by one of the following: definitely negative, negative, positive and definitely positive. Child's behaviour was evaluated at his/her first visit. Children with FBRS score 1 were excluded from the study. FBRS was repeated at the end of the treatment session.

Physiologic monitoring:

Pulse oximeter were used to measure physiological pulse rate (bpm) before, during and after treatment. Saliva was collected using the 'spitting method'. Each subject rinsed their mouth with water to reduce contamination of saliva with food debris and waited 5 minutes prior to sampling. Whole mouth saliva from the oral cavity was collected by asking the subjects to sit comfortably in an upright position and drop down their heads, let the saliva run naturally to the front of mouth without stimulating flow by oro-facial movements. Saliva that accumulates in the floor of mouth was expectorated into a graduated a polypropylene test tube every 30 s, for a total of 2 min. The amount of collected saliva in mL divided by the duration of the collection period, yielding ml/min was recorded as the mean salivary flow rate. Saliva samples were collected at three time points as follows: First saliva samples of children were collected 5 min before the dental intervention (pre-5, measuring the stress of being at the clinic). The patients then underwent dental procedures which lasted around 30 minutes. Right after dental care and after 10 minutes of resting (post-10), two new saliva samples were collected. Collection of at least 1 ml was required. After collection, the unprocessed samples were stored at -20°C until they were analyzed.