Decision-making of general practitioners on interventions at restorations based on bitewing radiographs

Challenges in conducting clinical research in primary care dentistry

The integration of dentistry into primary health care is crucial for promoting patient well-being. However, clinical studies in dentistry face challenges, including issues with study design, transparency, and relevance to primary care. Clinical trials in dentistry often focus on specific issues with strict eligibility criteria, limiting the generalizability of findings. Randomized clinical trials (RCTs) face challenges in reflecting real-world conditions and using clinically relevant outcomes. The need for more pragmatic approaches and the inclusion of clinically relevant outcomes (CROs) is discussed, such as tooth loss or implant success. Solutions proposed include well-controlled observational studies, optimized data collection tools, and the integration of artificial intelligence (AI) for predictive modelling, computer-aided diagnostics and automated diagnosis. In this position paper advocates for more efficient trials with a focus on patient-centred outcomes, as well as the adoption of pragmatic study designs reflecting real-world conditions. Collaborative research networks, increased funding, enhanced data retrieval, and open science practices are also recommended. Technology, including intraoral scanners and AI, is highlighted for improving efficiency in dental research. AI is seen as a key tool for participant recruitment, predictive modelling, and outcome evaluation. However, ethical considerations and ongoing validation are emphasized to ensure the reliability and trustworthiness of AI-driven solutions in dental research. In conclusion, the efficient conduct of clinical research in primary care dentistry requires a comprehensive approach, including changes in study design, data collection, and analytical methods. The integration of AI is seen as pivotal in achieving these objectives in a meaningful and efficient way.

Detection of caries around restorations on bitewings using deep learning

OBJECTIVE

Secondary caries lesions adjacent to restorations, a leading cause of restoration failure, require accurate diagnostic methods to ensure an optimal treatment outcome. Traditional diagnostic strategies rely on visual inspection complemented by radiographs. Recent advancements in artificial intelligence (AI), particularly deep learning, provide potential improvements in caries detection. This study aimed to develop a convolutional neural network (CNN)-based algorithm for detecting primary caries and secondary caries around restorations using bitewings.

METHODS

Clinical data from 7 general dental practices in the Netherlands, comprising 425 bitewings of 383 patients, were utilized. The study used the Mask-RCNN architecture, for instance, segmentation, supported by the Swin Transformer backbone. After data augmentation, model training was performed through a ten-fold cross-validation. The diagnostic accuracy of the algorithm was evaluated by calculating the area under the Free-Response Receiver Operating Characteristics curve, sensitivity, precision, and F1 scores.

RESULTS

The model achieved areas under FROC curves of 0.806 and 0.804, and F1-scores of 0.689 and 0.719 for primary and secondary caries detection, respectively.

CONCLUSION

An accurate CNN-based automated system was developed to detect primary and secondary caries lesions on bitewings, highlighting a significant advancement in automated caries diagnostics.

CLINICAL SIGNIFICANCE

An accurate algorithm that integrates the detection of both primary and secondary caries will permit the development of automated systems to aid clinicians in their daily clinical practice.

Artificial Intelligence as a Decision-Making Tool in Forensic Dentistry: A Pilot Study with I3M

Expert determination of the third molar maturity index (I3M) constitutes one of the most common approaches for dental age estimation. This work aimed to investigate the technical feasibility of creating a decision-making tool based on I3M to support expert decision-making. Methods: The dataset consisted of 456 images from France and Uganda. Two deep learning approaches (Mask R-CNN, U-Net) were compared on mandibular radiographs, leading to a two-part instance segmentation (apical and coronal). Then, two topological data analysis approaches were compared on the inferred mask: one with a deep learning component (TDA-DL), one without (TDA). Regarding mask inference, U-Net had a better accuracy (mean intersection over union metric (mIoU)), 91.2% compared to 83.8% for Mask R-CNN. The combination of U-Net with TDA or TDA-DL to compute the I3M score revealed satisfying results in comparison with a dental forensic expert. The mean ± SD absolute error was 0.04 ± 0.03 for TDA, and 0.06 ± 0.04 for TDA-DL. The Pearson correlation coefficient of the I3M scores between the expert and a U-Net model was 0.93 when combined with TDA and 0.89 with TDA-DL. This pilot study illustrates the potential feasibility to automate an I3M solution combining a deep learning and a topological approach, with 95% accuracy in comparison with an expert.

Training in radiographic caries detection and staging using an interactive tool

INTRODUCTION

The objective of the study was to compare a dental student's practical ability to detect and stage radiographic caries per International Caries Detection and Assessment System (ICDAS), following a traditional lecture and a lecture containing an interactive session using an audience response system (ARS). Associations between the order of instructions and student performance were also evaluated.

MATERIALS AND METHODS

Eighty-three dental students were randomly assigned to groups A and B. On the first day, group A received a traditional lecture and group B received content using the ARS. All students then took an electronic quiz (T1) identifying and staging caries on radiographs per ICDAS. For the second day, group A received the content using the ARS system and group B received a traditional lecture. All students subsequently took a second electronic quiz (T2). Two survey questions about the learning experience were also included.

RESULTS

Wilcoxon rank-sum analysis of scores from consenting students (81) showed no difference between the quiz 1 scores of two groups (p=.61). Whilst not statistically significant (p = .07), the group that had the ARS initially scored marginally higher on quiz 2. Survey results showed that most participants preferred either the ARS alone (49.38%) or a combination of the ARS and a traditional lecture (40.74%). A majority of them (80%) found the ARS helpful.

CONCLUSION

When training students in practical skills of detection and staging radiographic presence of dental caries per ICDAS, hands-on learning tools, such as an ARS, complement traditional lectures.

Additional information of bitewings to first time clinical examination of caries and restoration status in permanent dentition

OBJECTIVES

Identifying additional information obtained by X-rays combined with clinical examination concerning primary caries, caries adjacent to restorations and quality of restorations.

MATERIAL AND METHODS

A total of 240 adult patients, equally distributed in gender and six age-groups, were randomly selected from an original study population of 4,402 subjects (DANHES). Clinical and radiographical registrations on occlusal and approximal surfaces in posterior teeth were categorized into unrestored surfaces (sound/primary caries) and restored surfaces (without/with caries adjacent to restorations). Material and quality of restorations were also recorded. Chi-square and Fisher-exact tests were used for statistical analyses.

RESULTS

Of potentially 11,520 surfaces, 3,015 occlusal and 5,112 approximal surfaces were analysed. Occlusal: Of 907 unrestored surfaces, 110 had primary caries and 53% were detected radiographically. A total of 183 of 2,108 restored surfaces had caries adjacent to restorations, and 99% were found radiographically. A total of 190 restorations were over/under-extended, and 89% were registered radiographically. Approximal: Of 2,649 unrestored surfaces 648 had primary caries, and 92% were registered radiographically. A total of 565 of 2,463 restored surfaces had caries adjacent to restorations, and 99% were found radiographically. A total of 638 restorations were over/under-extended, and 98% were found radiographically. For all restorations, material and depth influenced quality of restorations and incidence of caries adjacent to restorations. At least one independent variable (gender/age group/tooth-type/jaw/side/mesial-distal surface) influenced frequencies of primary caries, caries adjacent to restorations and quality of restorations.

CONCLUSIONS

First-time clinical examination must be supplemented with X-rays to obtain a complete impression of caries status in posterior regions regarding diagnostics of caries, assessment of lesion depth and quality of restorations.

Study protocol for a diagnostic randomized clinical trial to evaluate the effect of the use of two clinical criteria in the assessment of caries lesions around restorations in adults: the Caries Cognition and Identification in Adults (CaCIA) trial

Background

The assessment of restored teeth in dentistry remains a challenge, mainly related to the detection of caries around restorations. There is a diversity of clinical criteria available to assess the caries lesions, resulting in differences in the dentists’ diagnosis and treatment decisions. In addition, there is a lack of evidence regarding the best criteria to detect caries lesions around the restorations. Thus, the present protocol aims to evaluate the effect of using 2 visual criteria to assess restored teeth on the outcomes related to oral health in adults.

Methods

The design protocol of the Caries Cognition and Identification in Adults trial correspond to a triple-blind randomized, controlled clinical trial with parallel-groups. Two groups will be compared: patients who will receive the diagnosis and treatment decision according to FDI (World Dental Federation) criteria—FDI group; and patients who will receive diagnosis and treatment decision according to the “Caries Associated with Restorations or Sealants” criteria defined by the International Caries Classification and Management System (ICCMS group). The participants will be followed up after 6, 12, 18, 24, and 60 months, and the restoration failure will be the primary outcome. The analysis will be conducted through Cox regression with shared frailty. The impact of oral health on quality of life and the cost-effectiveness of the methods used will be the secondary outcomes. Two-tailed analyzes will be used, considering a level of significance of 5%.

Discussion

This is the first clinical trial to assess the effect of using two visual methods to detect caries lesions around restorations on the outcomes related to oral health in adults. The findings of this study will define what is the best diagnostic strategy for the assessment of caries around restorations in permanent teeth.

Trial registration NCT03108586 (registered 11 April 2017).

Practice based research in dentistry: an alternative to deal with clinical questions

Clinical interventions in dental practice should be determined based on the best scientific evidence available. Well-designed randomized clinical trials (RCTs) provide important evidence supporting the efficacy of interventions and are usually considered as the best primary evidence. However, the strict criteria adopted by most RCTs reduce their external validity since some findings from these studies might not work under usual conditions. On the other hand, practice-based research (PBR) studies have been designed to better define the effectiveness of clinical interventions under settings closer to "real-world" conditions. Therefore, this review aimed to describe different PBR designs discussing some advantages and limitations of such studies. The stimulus to organization PBR networks is discussed since the studies performed by these networks involve large number of clinicians and important conclusions can be drawn. Designs of observational studies including surveys and cohort studies based on practice are presented. Survey methods are important to know the behavior of practitioners regarding diagnostic and decision of treatment. Cohorts allow assessing different cofounders contributing to some outcome since large sample sizes and long follow-up periods can be observed in some of these studies. Pragmatic trials designed to take place in real-world clinical practice settings are also discussed as a useful design to assess the effectiveness of clinical interventions. In conclusion, this review sought to present PBR studies as alternative designs to answer clinical questions, but not replacing randomized clinical trials.

Influence of different clinical criteria on the decision to replace restorations in primary teeth

OBJECTIVES

This cross-sectional study is nested in a randomized clinical trial. It was designed to evaluate the influence of using two different clinical criteria to assess caries lesion around restorations on the decision to replace restorations in primary posterior teeth. Variables that might influence this decision were also considered.

METHODS

One trained and calibrated examiner assessed 550 restorations of 160 children (3-10 years old). Children were randomized to have their restorations evaluated and subsequently treated according to World Dental Federation (FDI) or Caries Associated with Restorations and Sealants (CARS) criteria. After reaching the treatment decision, the same examiner performed another evaluation using the other criteria. Spearman's correlation coefficients and 95 % confidence intervals (95 %CI) between the scores obtained with both criteria and respective treatment decisions were calculated. Poisson multilevel regression analysis were performed between the exploratory variables related to children, restored tooth and restoration assessment; the outcome variables were decisions related to restoration replacement, any operative intervention and presence of secondary caries.

RESULTS

The strongest correlation observed between the methods was for recurrence of caries. A total of 94 restorations (17.1 %) were indicated for replacement with FDI criteria and 30 (5.5 %) were indicated for replacement with CARS. Besides the diagnostic method used, number of decayed teeth and restorations with two and three restored surfaces were associated with the decision of replacement and presence of recurrent caries lesions.

CONCLUSIONS

The decision to replace posterior restorations in primary teeth is influenced by the criteria used for the restorations assessment and also by the children's caries experience and multisurface restorations. The restoration material did not influence the decision of restorations replacement.

CLINICAL SIGNIFICANCE

The decision to replace posterior restoration in primary teeth is strongly related to the evaluation method and not only by patients' risk factors.

Risk Factors for Dental Restoration Survival: A Practice-Based Study

To improve patient dental care, it is necessary to identify possible risk factors for the failing of restorations. This practice-based cohort study investigated the performance and influence of possible risk factors at the level of the practice, patient, tooth, and restoration on survival of direct class II restorations. Electronic patient files from 11 Dutch general practices were collected, and 31,472 restorations placed between January 2015 and October 2017 were analyzed. Kaplan-Meier statistics were performed; annual failure rates (AFRs) were calculated; and variables were assessed by multivariable Cox regression analysis. The observation time of restorations varied from 0 to 2.7 y, resulting in a mean AFR of 7.8% at 2 y. However, wide variation in AFRs existed among the operators, varying between 3.6% and 11.4%. A wide range of patient-related variables is related to a high risk for reintervention: patient age (elderly: hazard ratio [HR], 1.372), general health (medically compromised: HR, 1.478), periodontal status (periodontal problems: HR, 1.207), caries risk and risk for parafunctional habits (high: HR, 1.687), restorations in molar teeth (HR, 1.383), restorations placed in endodontically treated teeth (HR, 1.890), and multisurface restorations (≥4 surfaces: HR, 1.345). Restorations placed due to fracture were more prone to fail than restorations placed due to caries. When patient-related risk factors were excluded, remaining risk factors considerably changed in their effect and significance: the effect of operator, age of the patient, and endodontic treatment increased; the effect of the diagnosis decreased; and the socioeconomic status became significant (high: HR, 0.873). This study demonstrated that a wide variation of risk factors on the practice, patient, and tooth levels influences the survival of class II restorations. To provide personalized dental care, it is important to identify and record potential risk factors. Therefore, we recommend further clinical studies to include these patient risk factors in data collection and analysis.