Overview

The study is a community-based single-blinded clustered randomised controlled trial involving school children aged 6–8 years. The present clinical trial is nested into a larger study that aims to evaluate the impact of oral health in children’s development (quality of life, anthropometric and cognitive development). For this study, recruitment and treatment is provided at different public schools in Paranoá, a deprived suburban area of Brasilia, Brazil.

For our trial, recruitment has been finalised and patients are re-examined for the first recall visit (12 months) in 2018. We enrolled patients with one or more deeply carious, vital and non-symptomatic primary molar(s).

The unit of randomisation was the child, with all primary molars with a deep carious lesion in the same child receiving one of the two treatments. Non-eligible teeth in eligible patients were treated according to their need using standard protocols either on site or via referral to the paedodontic service of the University of Brasília. These protocols included atraumatic restorative treatment using a high-viscosity glass-ionomer cement or bulk-fill resin composite placed using a universal adhesive system.

Total follow-up time will be 3 years after completion of the initial treatment (ie, until 2020); longer follow-ups are not planned as children will change schools and teeth will exfoliate. Re-examinations will be performed after 12, 24  and 36 months. The study is registered at ClinicalTrials.gov (NCT02754466).

Setting and participants

Study recruitment took place in public primary schools of Paranoá, in Brasilia, the capital of Brazil. Participants were invited to attend routine examinations and recruited to the trial if eligible. Treatment was provided in a mobile dental unit, as described below. We included 115 children aged 6–8 years with minimum one vital, clinically and radiographically non-symptomatic, retainable, deeply carious primary molar. The lesion needed to radiographically extend into the inner half of the dentin and show signs of activity, for example, softness and roughness of the surface, plaque stagnation. Single and multiple surface lesions were included. The size of the cavity after preparation (maximum orovestibular extension and included surfaces) was recorded for later analysis of possible influences on outcomes. The lesion should not have been associated with non-carious hard tissue defects like fluorosis or molar incisor hypomineralisation. Patients required parental consent for participation. In addition, patients’ cooperation for treatment under no or only local anaesthesia was needed. Patients with severe systemic diseases or disabilities were not included. Patients with known allergies to dental materials used within the study as well as those with teeth expected to exfoliate within the next 18 months were also not included.

Sample size

The unit of analysis was the tooth. Hence, this a multilayered cluster randomised trial, with the patient and the school being the clusters. Clustering needs to be taken into account. This is done using the so-called Lee, Wei and Amato (LWA) model for clustered survival data.13 Sample size estimation for this model was based on the ideas of Xie and Waksman.14 The necessary calculations were performed with the R program ssizecl.surv.15 Required sample size was calculated based on the primary outcome parameter ‘time until complications’. We anticipated that at 36 months, 90% of the objectively excavated teeth (polymer bur carious tissue removal) and 80% of the subjectively excavated teeth (subjective carious tissue removal) would not experience any complications; this is in the range of what has been reported for similar treatments.¹⁷ As mentioned, the assumption of superiority was made, as polymer burs involve additional costs and efforts, which need to be clinically justified. If α=0.05 (two-sided) and 1−β=0.9, a sample size of 45 children per group was needed when conservatively assuming an intercluster correlation (ICC) of 0.8 and a mean of 1.5 teeth per child being treated. Note that such ICC is relatively high, but accounts for not knowing the ICC in the specific population. Assuming a smaller ICC of 0.3 required 37 patients per group. Assuming an overall drop-out of 20%, the required sample size was 57 per group. We eventually recruited 115 patients (177 teeth). Note that we had originally aimed to recruit an even higher number of children, but that this was not feasible within the financial scope of this trial.

Recruitment

Recruitment was performed within six primary schools in Brasília. A routine examination was offered to all potentially eligible pupils. Eligible patients and their parents received the study information and consent forms. Written consent was given by the parents of the children. There was a minimum time period of 48 hours between study information and consent.

Allocation and blinding

Random allocation was performed as follows. When the child was seated in the chair, one opaque envelope, containing the allocation, was drawn from a total of 120 envelopes (60 per group). Opening the envelope revealed the allocation. Note that eventually, only 115 children were recruited (as described), which is why groups are not perfectly balanced (five envelopes were never opened).

Blinding of the operator and the patient was not possible. However, clinical follow-up examinations will be performed blinded. This will be possible, as provided restorations will be identical in both groups, making it impossible to identify groups without further knowledge. We will also ensure that neither the child nor the parents will provide any information regarding the group allocation to the examiner. If patients require radiographic evaluation during follow-up visits, an independent third researcher will modify the radiographs in such way that identification of groups (eg, via assessing the thickness of remaining residual carious dentin) is not possible for the examiner. Note that not all patients may require follow-up radiographs and that further radiographs are usually only taken for patients requiring retreatments due to radiation protection reasons.

Outcomes

Outcomes will be assessed by an independent dentist who will be blinded for the original intervention. The primary outcome of the study will be the time until complications (endodontic and/or restorative) occur. Secondary outcomes will include the time until tooth removal is needed, patients’ satisfaction with treatment using a 5-point Likert scale and cost-effectiveness (see below).

Intervention and data collection

Examination, intervention and data collection were or will be provided in a mobile dental unit. In the first examination visit, a full assessment and intraoral examination was performed. For patients who were possibly eligible, caries risk was estimated using the decayed missing filled teeth index (DMFT, measured using WHO criteria), with DMFT ≥2 counting as high risk. If informed consent was given, treatment was provided in the second visit, with local anaesthetics being applied if needed (only two cases eventually required this; the decision if the discomfort was bearable was made by the child). Note that all patients had been conditioned for dental therapy a priori.

Removal of enamel and cavity preparation were performed using water-cooled diamond instruments; removal of peripheral carious dentin was performed using conventional rose head burs and/or hand excavators until only hard, dry dentin remains. Pulpo-axial carious dentin was removed either using a hand excavator until leathery, slightly moist and reasonably soft dentin remains (ie, dentin which cannot be removed using an excavator without force being needed; subjective removal) or via polymer burs (Polybur, Komet) until a new bur abrades and does not remove any dentin anymore (objective removal).

Operating dentists were calibrated prior to study commencement regarding these criteria and methods using extracted teeth. Moisture control was performed using cotton rolls and continuous aspiration. Restorations were performed using a high-viscosity glass ionomer cement (Equia Forte, GC, Tokyo, Japan).

Pulpal exposure did not occur. Follow-up treatment will be provided according to standardised local dental routine. All retained teeth will be followed regardless of the treatment arm.

Data collection will be performed via case reports forms. The following data have been or will be collected:

a) At the initial examination:

• Age, gender, medical and dental history;

• Caries risk;

• Decayed, missing, filled teeth according to WHO criteria;

• For teeth which are potentially eligible: Caries Assessment Spectrum and Treatment code,16 involved surfaces and cavity size, extension, lesion activity, radiographic lesion depth;

• Endodontic status: absence of spontaneous pain, no swelling or sinus formation.

b) After treatment:

• Operator;

• Surfaces treated;

• Used anaesthetic material;

• Subjective assessment (satisfaction) after the treatment (a 5-point Likert face scale was used), and the question ‘how satisfied were you with this treatment?’ was asked;

• Transportation time and costs;

• Time, staff and material required for treatment.

c) At follow-up:

• Tooth retreated alieno loco (yes/no);

• Teeth extracted or extraction needed (yes/no);

• Tooth exfoliated (yes/no);

• Complications: pulpal complications (pain, pulpitis, sensitivity to percussion or cold/hot, swelling, sinus formation) non-pulpal complications (restorative need to re-intervene, eg, caries at restoration margins, fracture or loss the restoration; fracture of the tooth).

d) After follow-up treatments:

• If follow-up treatments are required, costs of these will be assessed as well. Retained teeth will be followed-up regardless if they required retreatment or not to assess their survival. A summary of performed procedures and recorded data can be found in table 1.

The study flow is summarised in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flow chart of the study. 

Data analysis and statistical evaluation

As outlined the primary efficacy analysis will use a variant of the Cox proportional hazards model, that is, the LWA model that allows for multiple observations per subject to compare both treatment groups, at 36 months. This analysis will be performed with proc phreg (SAS V.9.4). Details may be found in the study by Guo.17 The primary analysis population will be based on the intention-to-treat principle (ITT). The ITT population is defined as the population who received at least one dental treatment (full analysis set); patients who have failed at time zero will be reported but will not provide data for the survival analysis. Additionally, a per-protocol analysis will be performed. A sensitivity analysis with regard to missing values will use multiple imputation methods.18 No interim analyses are planned. As this will be the single confirmatory analysis, no adjustment for multiple testing is required.

All secondary analyses will be exploratory. In this sense, we will analyse effects of possible confounders (school, age, gender, dental anxiety, caries risk, cavity size and number of involved surfaces, dental arch, first or second molar, surfaces). Mixed models will be used to compare subjective satisfaction and costs between groups. Incremental cost-effectiveness ratios and cost-effectiveness acceptability via the net-benefit approach will be used to describe the relative cost-effectiveness.

Missing data

We anticipate various reasons for missing data (drop-out, complications, withdrawal, adverse events). Sample size calculation was performed under the consideration of possible loss-to-follow-up, that is, data missing at random. To prevent data missing systematically, that is, not at random, case report forms were designed in a way enforcing complete reporting. We will account for data not missing at random according to the ITT principle in our primary analysis.

Ethics and dissemination

In case of adverse events (AEs) and/or severe adverse events (SAEs), the committee will be informed (see below). Both parents and patients will receive detailed verbal and written explanations regarding the study and the procedures therein involved. Informed written consent of the parents was required and was given not earlier than 48 hours after study information.

All amendments to the protocol shall be agreed to by the principal investigator (LAH) and be recorded with a justification for the amendments. Amendments will be reviewed to determine the need for ethically re-approving the amended protocol. Amendments will be recorded and reported appropriately.

Patients and their parents were informed that patients have the right to withdraw from the study at any time without giving reasons. Withdrawal will be documented and consent will be sought from participants to retain data collected up to the point of withdrawal.

We expect only few specific AEs or SAEs in our trial, like allergy or severe reaction to a study material used (local anaesthetics, restoration material) or rare, uncommon reactions (unexpected bleeding, severe pain, uncontrollable anxiety) to or during the performed interventions. Any AE will be investigated and reported to the ethics committee. In case of SAEs requiring unblinding, code-breaking is possible. If the investigation is terminated prematurely or suspended, the principal clinical investigator (LAH) will execute this termination and inform the ethics committee accordingly.

The results of this study will be published in international peer-reviewed journals. A summary of study results will also be saved at ClinicalTrials.gov to allow general access to obtained findings. Original data will be made available by authors on request.

Patient and public involvement

Neither patients nor the public were involved in the design of the study, including conceiving the research questions and deciding the outcomes. However, our study will assess the subjective burden on the patients as described.

Trial status

The trial was  registered  at ClinicalTrials.gov and the study recruitment has been finished.