Ethical approval, protocol registration, and open science

The present research protocol was approved by the Local Human Research Ethics Committee (Approval No. 80918624.9.0000.0121; Opinion No. 7.015.281). The study protocol was registered on the Brazilian Registry of Clinical Trials (ReBEC) platform (Registration No. XXXX) on 12 March 2025 and is publicly available at XXXX. The study was conducted in accordance with the principles of the Declaration of Helsinki (Carlson et al. 2004) and national legislation governing research involving human participants. Written informed consent was obtained from parents or legal guardians, and written informed assent was obtained from the participating children through the Informed Consent Form and the Free and Informed Assent Form, respectively.

The manuscript was reported in accordance with the 2025 recommendations of the Consolidated Standards of Reporting Trials (CONSORT) (Hopewell et al. 2025) and its extension for non-inferiority and equivalence trials (Piaggio et al. 2012). De-identified data generated in this study are available from the corresponding author upon reasonable and ethically justified request.

Testing settings, data collection locations, and recruitment

The present study was designed as a randomised, controlled, open-label clinical trial with two parallel groups (1:1) and a non-inferiority hypothesis. The test group received topical cryoanaesthesia using an ice stick, whereas the control group received topical anaesthesia with a 20% benzocaine-based gel. The study was conducted at the clinical facilities of the Postgraduate Programme in Dentistry at the XXX Brazil. Eligible participants were recruited between 17 March and 8 April 2025, during which the operator provided study information and obtained written informed consent and assent (ICF and TALE). Data collection was conducted between 09 April and 22 May 2025.

Eligibility criteria

Eligibility was assessed through clinical examination using a dental mirror, an exploratory probe, and a periodontal probe. A total of 40 children aged 6 to 9 years, of both sexes, classified as ASA I or II, and with previous dental experience—defined as having previously undergone dental procedures under local anaesthesia, were included. All participants required treatment of a mandibular primary molar under local anaesthesia. Exclusion criteria included the presence of odontogenic infection associated with systemic signs and symptoms, acute pain, a reported history of allergy to any drugs or materials used in the study, a history of systemic or haemorrhagic disorders, or the use of analgesic or anti-inflammatory medication within 6 h before the procedure.

Sample size

The sample size calculation was performed using the Sealed Envelope® online tool (https://www.sealedenvelope.com/power/continuous-noninferior) based on parameters described in a previous clinical trial that evaluated self-reported pain following topical anaesthesia with benzocaine 20% compared to cryanaesthesia (Lakshmanan and Ravindran 2021). Considering a significance level (α) of 5%, 90% statistical power, a standard deviation of outcome of 9.7 (Lakshmanan and Ravindran 2021) and a non-inferiority margin of 10% (10 mm in a 100 mm VAS scale), a minimum required sample size of 34 participants was estimated. Adding 20% expecting possible discontinued interventions, the final sample size was 40 participants.

The non-inferiority margin was defined based on approximately half of the difference in mean pain scores (20.7) reported in the previous study (Lakshmanan and Ravindran 2021), according to recommendations from Honório et al. (2020). Furthermore, from a clinical perspective, the authors considered that differences inferior to 10 mm on a 100 mm VAS were unlikely to represent relevant changes in pain perception.

Randomisation

Randomisation was performed using the online tool Sealed Envelope®, employing permuted blocks of 4 or 6 to generate a sequential allocation list, thereby ensuring balanced distribution of participants between groups. Both sequence generation and allocation were undertaken by an independent third party who was not involved in the study procedures. Participants were randomly assigned to either the cryoanaesthesia group or the 20% benzocaine-based gel group according to the generated sequence. Allocation concealment was ensured using sealed, opaque, sequentially numbered envelopes prepared by an independent individual. Each envelope was opened only at the beginning of the intervention appointment, after confirmation of eligibility and receipt of parental consent. One tooth per child was included in the study.

Blinding

Owing to the inherent differences in texture and temperature between the topical anaesthetic methods evaluated, blinding of the operator and participants was not feasible.

Although the study was open-label, the main outcome was recorded at a standardised time point, during the puncture of the first 2 mm of the needle, which occurred after 2 min of topical anaesthesia application. Thus, the pain measurement reflected the effect of topical anaesthesia evaluated during anaesthetic puncture. In addition, all participants received standardised instructions and were treated under similar clinical conditions.

Intervention

Before the intervention, all parents or legal guardians completed the Portuguese version of the Dental Anxiety Scale (DAS) in a neutral setting (waiting room). This instrument comprises four items, each with five response options, yielding a total score ranging from 0 to 20, and assesses the respondent’s level of anxiety regarding the forthcoming dental treatment (Hu et al. 2007). Additionally, the child’s anxiety was assessed using the Facial Image Scale (FIS) upon entering the surgery and once seated in the dental chair, to record pre-treatment anxiety levels (Buchanan and Niven 2002). This scale consists of five facial drawings ranging from very happy to very unhappy. The child was provided with the following explanation: “These pictures show how a child feels, from very happy (pointing to the happiest face) to very unhappy (pointing to the unhappiest face”. The child was then asked to indicate the face that best represented how they felt at that moment by answering the question: “Which picture shows how you feel right now?” Subsequently, the VAS was presented to the participant, and the procedure for recording pain during the appointment was explained.

Subsequently, the envelope was opened and the allocation carried out according to the designated sequence. All clinical procedures were performed by a single operator, previously trained and calibrated for the study protocol, and with clinical experience in paediatric dentistry, including the routine management of paediatric patients and the administration of local anaesthesia. In the experimental group, during cryoanaesthesia, the tip of the standardised-sized ice stick was placed in an empty anaesthetic tube, with a tip of diameter 6 mm, which was placed on the previously dry mucosa at the base of the buccal vestibule, precisely at the site designated for the local insertion of the anaesthetic needle. During the two-minute application period, the suction tip was positioned next to the ice stick to prevent excessive moisture in the area caused by melting. Until the immediate moment of application, the ice sticks were stored in a thermal container with frozen cooling packs to maintain a low temperature for a prolonged period, thus preventing the tubes from starting the melting process (Fig. 1).

Fig. 1

Depiction of ice stick storage for cryoanaesthesia. A Thermal cooler; B Frozen gel plates placed inside the cooler box; C Ice sticks for use in cryoanaesthesia

In the positive control group, a 20% benzocaine-based topical anaesthetic gel (DFL®, Rio de Janeiro, Brazil), stored at room temperature, was applied with a sterile cotton pellet to the previously dried mucosa at the base of the buccal vestibule. The gel remained in situ for two minutes, in accordance with the manufacturer’s instructions.

In both groups, following the topical anaesthetic procedure, local anaesthesia was administered using a buccal infiltrative technique with one cartridge of 2% lidocaine (Alphacaine 2%, DFL®). An extra-short (12 mm), 30 gauge needle (Septodont®, Saint-Maur-des-Fossés, France) was fitted with a sterile silicone stopper positioned 2 mm from the tip to limit the depth of needle penetration, given that topical anaesthetics are effective only within the superficial 2 to 3 mm of the mucosa (Agarwal et al. 2017; Malamed 2005). During needle insertion, a trained and calibrated researcher assessed and recorded the child’s behaviour using the Frankl Behaviour Rating Scale (Frankl et al. 1962). The operator performed the initial needle puncture whilst holding only the anaesthetic cartridge, avoiding contact with the plunger to prevent inadvertent deposition of the anaesthetic solution. At the end of the initial needle penetration, the needle was maintained in position—up to the limit established by the silicone stopper—without initiating injection. At this point, a trained researcher presented the VAS to the child and asked them to rate the pain experienced up to that moment. Immediately after the child’s response was recorded, needle insertion was completed, and the anaesthetic solution was administered at a rate of 1 mL per minute. Subsequently, the planned dental procedure—extraction, endodontic treatment, or restoration under rubber dam isolation—was performed. The clinical environment was standardised for all participants to ensure a welcoming atmosphere. The same non-pharmacological behaviour management techniques were employed throughout, including tell–show–do, distraction, and positive reinforcement.

Outcome assessment

The clinical procedures were performed by the same operator and assistant, who, together with the outcome assessors, underwent comprehensive theoretical training covering all stages of the study, from the clinical procedures to the administration of the assessment scales.

The primary outcome was self-reported pain following anaesthetic needle puncture (penetration of the initial 2 mm of the needle), assessed using VAS. The scale consists of a 100 mm horizontal line displayed on a laminated sheet, on which the participant indicates the intensity of pain perceived at a pre-specified moment. The line includes small vertical markings at 10-mm intervals, with only the extreme values explicitly labelled: 0, representing “no pain”, and 100, representing “the worst pain imaginable” (Castarlenas et al. 2017; Santos et al. 2020). The hypothesis for this outcome was that self-reported pain during infiltrative anaesthetic needle puncture in children receiving topical cryoanaesthesia would not exceed the 10 point non-inferiority margin on the VAS (Lakshmanan and Ravindran 2021) when compared with children who received 20% benzocaine gel as topical anaesthesia.

As a potential predictor of pain reporting, an examiner assessed the child’s behaviour using the Frankl Behaviour Rating Scale, which categorises behaviour into four levels according to the child’s response to the procedure: definitely negative, negative, positive, and definitely positive (Frankl et al. 1962). Other potential predictors included the children’s anxiety, assessed using the FIS, which comprises five facial drawings ranging from very happy to very unhappy, scored from 1 to 5. In addition, the anxiety of parents or legal guardians regarding dental care was evaluated using the DAS.

Training, calibration, and pilot study

Training of the operator, evaluators, and assistants consisted of theoretical sessions during which the clinical care protocol was presented and discussed. Subsequently, the evaluators were calibrated for the Frankl Behaviour Rating Scale using videos of paediatric clinical procedures, and inter- and intra-examiner reliability were assessed using the weighted kappa coefficient (κ > 0.70). Inter-rater agreement was 0.71 (moderate agreement) for both evaluators, whilst intra-rater agreement was 0.70 (moderate agreement) and 0.85 (strong agreement), respectively (McHugh 2012). To assess the feasibility and applicability of the protocol, a pilot study was conducted with 10 participants. As the minor adjustments made during the pilot phase did not affect the reliability of the results, these participants were subsequently included in the main study.

Harms

Harm was defined as any local or systemic adverse event related to the application of cryoanaesthesia, including discomfort, changes in mucosal colour or integrity, or any other unexpected reaction. A follow-up assessment was conducted within 24 h of the appointment via telephone contact with the child’s legal guardian.

Statistical analysis

Data were entered into an electronic spreadsheet (Excel®, Microsoft Corp., WA USA). Statistical analyses were performed using NCSS Statistical Software (version 24.3, 2021; NCSS, LLC, UT, USA) and the Statistical Package for the Social Sciences (SPSS for Windows, version 21.0; IBM Corp., NY, USA). A significance level of 5% was adopted for all analyses.

A per protocol analysis was conducted. Initially, a descriptive analysis of the main characteristics of the sample was performed, including age, gender, and baseline anxiety of the participants and their guardians. To evaluate baseline balance between groups, the child’s anxiety was dichotomised as absence of anxiety (score 1) or presence of anxiety (scores 2–5) (Buchanan and Niven 2002). Caregiver anxiety was categorised according to DAS scores as follows: DAS ≥ 15 indicated extreme anxiety, DAS 12–14 indicated moderate anxiety, and DAS ≤ 11 indicated low levels of anxiety (Hu et al. 2007). Subsequently, participants’ sex and anxiety levels were compared between groups using Pearson’s chi-squared test, whilst caregiver anxiety was compared using Fisher’s exact test. The mean age of participants was compared between groups using the Mann–Whitney U test.

A one-sided Student’s t-test for non-inferiority was used to compare mean differences between groups, considering a non-inferiority margin of 10% (corresponding to 10 points on the VAS), as a unilateral test with a pre-specified margin. To evaluate potential factors associated with pain scores at anaesthetic needle puncture (dependent variable), multiple linear regression analysis was performed. For the adjusted model, automatic stepwise procedures were not applied; instead, variables were selected using a purposeful selection approach. Variables with a p-value ≤ 0.20 in the unadjusted analysis were manually entered into the adjusted model (Mickey and Greenland 1989). The assumptions of the linear regression model were assessed through residual analysis, including evaluation of normality and homoscedasticity.

Changes to the registered protocol

During research, changes were made to the initial registered protocol. After trial registration, it was decided not to record the Facial Image Scale (FIS) at the initial consultation, as the analysis did not include comparisons of anxiety levels across consultations. Similarly, the FLACC scale was not used, given that all participants were verbal and self-reported pain should be prioritised whenever feasible (Raja et al. 2020). To ensure greater sample homogeneity, only participants with previous dental experience were included. In addition, children who had used analgesic or anti-inflammatory medication within six hours before the consultation were excluded. Furthermore, due to the period required for protocol amendments on the trial registration platform, adjustments to the study timeline were necessary, resulting in a delay in the initiation of the recruitment phase and, consequently, subsequent stages of the study. The protocol was registered on March 12, 2025, and the first participant was recruited on March 17, 2025. No changes were based on collected data or preliminary study results. The modifications were due to considerations identified during the preparatory phase, before the inclusion of the first participant.