Study design
This trial was approved by the French ethics committee (N°19.02.18.67521) and was registered prior to patient enrollment in agreement with French legislation (ID-RCB: 2019-A00459-48). Written informed consent was obtained from the parents before inclusion of the each child. This manuscript adheres to the applicable CONSORT guidelines.
Between March 2019 and March 2020, we conducted a single-center, randomized, controlled, single-blind, parallel-group, trial with 1:1 allocation at the Marie Lannelongue Hospital. Consecutive patients between 3 and 10 years of age admitted for cardiac catheterization to assess congenital heart disease were eligible, based on the manufacturer’s recommendations on the use of the car. Exclusion criteria were body weight greater than 30 kg (upper limit for which using the car was approved), cognitive or psychiatric disorder, neurosensory pathology making the child unable to drive the car, lack of understanding of French, and parental refusal. For eligible patients meeting none of these criteria, inclusion in the study was proposed on the day before catheterization, after a full explanation of the procedure and study protocol had been given to the parents and children.
Interventions
The study intervention consisted in having the child drive a toy electric ride-on car from the ward to the catheterization laboratory (Figure 1). On day before the intervention, the physician in charge of each patient randomized to the intervention group provided the child with accurate and playful explanations about the car and how it would be used. In the control group, the children were taken to the catheterization as usual lying supine on a gurney. In both groups, the parents and an orderly accompanied the child. In the intervention group, the orderly had received training about the car and have a remote control that could override the child’s actions, to ensure safety.
All children in both groups received standardized preoperative information from a nurse, who used a play kit with figurines to represent the operating room and the course of the operation. No premedication was given in either group. Children in either group who exhibited extreme symptoms of anxiety deemed incompatible with the study protocol were excluded from the study and given vigilant sedation adapted to the level of anxiety.
Once at the catheterization laboratory, the child was separated from the car and parents and given anesthesia promptly. The mode of anesthesia was at the discretion of the anesthesiologist in both groups, a standardized anesthesia protocol was not necessary since we not measure post-procedural status (post-emergence delirium for example).
Anxiety assessment
Anxiety in the child and parents was assessed on the day before the procedure, in the child's room (T0) and on the day of the procedure just before anesthesia induction (T1). The T1 assessment took place after the child was separated from the parents. We used two scales, the short form of the modified Yale Preoperative Anxiety Scale (mYPAS-SF) [16, 17] and the Visual Analog Scale for Anxiety (VAS-A). Both scales are validated for assessing anxiety in pediatric patients and adults [18–20].
The mYPAS-SF is validated for children 2 years of age and older [21]. The level of anxiety is proportional to the score. The minimum score is 22/100 and scores above 30/100 reflect anxiety [16]. The scale has four categories illustrating the different forms of expression of anxiety (activity, verbal behavior, expression, and arousal) and each containing four to six items.
The VAS-A can be used for children aged 6 years or older. For younger children, the scale is marked by the parents. The level of anxiety is indicated by placing a mark on a line from 0 to 100. Scores taken to indicate clinically significant anxiety have ranged across studies from 30 to 50/100 [18, 20]. We used a cutoff of 30/100. Thus in this trial, we considered a child to be anxious if the mYPAS-SF or the VAS-A-child score were above 30.The parents also completed the VAS-A to reflect their own level of anxiety. We defined anxious parents with a VAS-A-parent score above 30.
Thus, anxiety was assessed by three scores (mYPAS-SF, VAS-A-child and VAS-A-parent), by the same blinded investigator, at T0 and T1. The difference between the T1 and T0 scores was taken to reflect anxiety induced by the proximity of the procedure.
Outcomes
The primary outcome was the median mYPAS-SF score at T1. Secondary outcomes were the median VAS-A-child score at T1, the median VAS-A-parent score at T1, and the median differences between T1 and T0 (∆T1-T0) in the mYPAS-SF, VAS-A-child and VAS-A-parent scores.
Randomization
The randomization sequence was generated by a central computer (STATA® Software) in blocks of 4 by a statistician. On the day before catheterization, each patient was enrolled in the study and allocated to one of the two groups by a clinical research associate, with a sealed envelope technique to one of the two groups.
Blinding of the healthcare professionals was not feasible. A physician who had no other role in the study and was unaware of the allocation group conducted the anxiety evaluations and recorded the anonymized results.
Sample size calculation
The mYPAS-SF, VAS-A-child, and VAS-A-parent scores were considered to indicate anxiety if greater than 30/100. Based on a previous comparable study, we hypothesized that the median T1 mYPAS-SF score would be 60/100 in the control group and 45/100 in the intervention group[15]. With a standard deviation of 15, an alpha risk of 0.05, and a beta risk of 0.10, 22 patients were required in each group to detect a statistically significant difference in the median mYPAS-SF score at T1 (primary outcome) between the two groups.
Statistical Analysis
The statistical analysis was carried out using the per-protocol approach. Non-normal and ordinal values were described as median [interquartile range]. Quantitative non-normal values, such as the mYPAS-SF score, were compared using the non-parametric Mann-Whitney test. Qualitative variables were described as percentages and compared using Fisher’s exact test. Quantitative normally distributed variables were compared using the Wilcoxon rank test. P values <0.05 were considered significant.