Detection and Decision-Making for Secondary Caries: A Scoping Review Informing a Clinical Decision Framework

1. Introduction

Secondary caries, also known as Caries Around Restorations or Sealants (CARS), is frequently cited as a cause of restoration failure and replacement in clinical practice [1,2,3]. When it comes to CARS, replacement of restorations is usually done to control the lesion; however, observational and clinical studies have raised concerns that secondary caries may be overdiagnosed, which could lead to premature intervention and the unnecessary removal of healthy dental tissue [2,3,4,5]. Repeated replacement procedures increase treatment complexity and costs over time, contributing to a restorative cycle [6,7].

Secondary caries presents distinct diagnostic challenges compared with primary caries. The presence of restorations limits direct visual and tactile access to the tooth surface and introduces confounding factors that complicate clinical interpretation [3,8]. Findings such as marginal staining, ditching, or small gaps are frequently observed during restoration assessment, yet their relationship with biologically active secondary caries remains uncertain when considered in isolation [9,10]. Consequently, clinicians are often confronted with a diagnostic grey zone, in which distinguishing lesion progression from aging-related restoration changes relies heavily on individual interpretation [8,11]. This diagnostic uncertainty is closely related to the structural complexity of the tooth–restoration interface, where restoration margins, gaps, staining, residual carious tissue, and subsurface lesion progression may produce overlapping clinical and radiographic appearances (Figure 1).

Visual–tactile examination is widely regarded as the primary detection method for assessing restorations suspected of secondary caries and is consistently recommended in clinical studies and guidelines [12,13,14,15]. Bitewing radiography is commonly used as an adjunct, particularly in proximal or submarginal areas where direct inspection is limited [9,16,17]. Conventional diagnostic approaches generally demonstrate high specificity but limited sensitivity for lesions beneath apparently intact restoration margins, contributing to diagnostic uncertainty depending on clinical context [16,18]. The frequent discrepancies between clinical and radiographic findings emphasize that radiographic data should serve as a complementary tool within a diagnostic strategy, rather than a standalone criterion for operative intervention [19,20,21].

To address the limitations of conventional assessment, several adjunctive diagnostic methods have been proposed, including fluorescence-based devices, optical imaging techniques and intraoral photography, as well as artificial intelligence (AI)–based systems designed to support the interpretation of diagnostic images [22,23,24,25]. These approaches have mainly been investigated as complementary tools to enhance lesion detection. However, their reported performance varies according to threshold definition, lesion location, and restoration-related confounders; moreover, concerns regarding false-positive findings have been raised [26,27]. Furthermore, the extent to which identifying additional suspicious findings translates into improved clinical management decision-making remains insufficiently explored [7,8].

Beyond identifying the lesion, a major clinical challenge is translating diagnostic findings into treatment decisions. The concept of diagnostic strategies can be understood as an overarching framework integrating detection (acquisition of clinical findings), diagnostic interpretation (assessment of lesion presence and activity), and management decision-making (selection of management approaches) [8,12,13,15]. Most research focuses on accuracy metrics, such as sensitivity and specificity, but often fails to explain how these results guide the choice between monitoring, repairing, or replacing a restoration [1,2,28]. Consequently, the clinical thresholds for operative intervention remain poorly defined and inconsistent. This lack of standardization explains why different clinicians may choose widely different treatments for the same clinical finding, frequently leading to unpredictable management and potential overtreatment [5,28,29].

Given the heterogeneity of study designs, diagnostic methods, reference standards, and outcome measures in this field, a comprehensive mapping of the available clinical evidence is warranted. Therefore, the aim of this scoping review was to map how diagnostic strategies are applied in the assessment of secondary caries lesions, from the identification of clinical findings to their interpretation and translation into management decisions and, based on this synthesis, to inform an evidence-based clinical decision framework.

2. Methods

This scoping review was conducted and reported in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) [30] and followed the methodological guidance of the Joanna Briggs Institute (JBI) for scoping reviews [31]. The review protocol was prospectively registered in the Open Science Framework (OSF; https://osf.io/jhzwu/?view_only=78efa71806624a68a695fa9727f27f61).

The review addresses the following research questions:

• Which detection or diagnostic methods are reported for the assessment of secondary caries lesions around restorations or sealants?
• Which diagnostic thresholds or criteria are used to support clinical decision-making for secondary caries lesion management?
• What recommendations can be made for future clinical studies investigating the detection and management decision-making for secondary caries?

2.1. Eligibility Criteria

Primary clinical studies of any design (randomized clinical trials, non-randomized controlled trials, and observational studies including cohort and cross-sectional designs) were eligible if they investigated detection or diagnostic methods and/or management decision-making thresholds related to secondary caries in restored primary or permanent teeth. All restorative materials and techniques were considered eligible, including direct and indirect restorations made of composite resin, amalgam, glass ionomer cement (GIC), gold, ceramic, or metal.

Studies were excluded if they were narrative or systematic literature reviews, in vitro or in situ using extracted teeth or artificially induced caries models, non-human studies, and purely technical or laboratory validation studies.

Clinical guidelines, consensus statements, or recommendation documents were included as a distinct evidence category when they explicitly addressed secondary caries (or equivalent terms) within detection, diagnostic, management decision-making thresholds, or management frameworks. For these documents, a report of clinical data was not required.

2.2. Information Sources and Search Strategy

Search strategies were developed with the assistance of an experienced health sciences librarian. No restrictions were applied for language or publication year. Full search strategies are provided in Supplementary Material A.

Electronic searches were conducted in PubMed, Web of Science, Scopus, LIVIVO, Cochrane, Embase, and Google Scholar (for gray literature, first 100 results sorted by relevance). Additional searches were performed in guideline repositories, including the National Institute for Health and Care Excellence (NICE), Guidelines International Network (GIN), and Trip Database. These searches were complemented by targeted searches of professional dental organizations and manual screening of reference lists from included titles. The initial search was conducted on March 3, 2024, and updated on July 13, 2026.

2.3. PCC Acronym and Selection Process

The review was guided by the Population-Concept-Context (PCC) framework, recommended for scoping reviews [31]:

• Population: Children and adults with restored primary or permanent teeth;
• Concept: Clinical detection methods, diagnostic interpretation, and management decision-making thresholds for secondary caries lesions;
• Context: Clinical assessment and management of restorations in dental care settings.

After duplicate removal using Mendeley, titles and abstracts were screened independently by two reviewers (E.T.C. and G.S.L.) using Rayyan [32]. Full text of potentially eligible records was retrieved and assessed independently using the same eligibility criteria. Disagreements were resolved through online discussion rounds, with arbitration by senior researchers in cariology (F.M.M. and M.S.C.).

2.4. Data Collection Process

When full texts or key information were unavailable, corresponding authors were contacted by e-mail and, when applicable, via professional platforms (e.g., ResearchGate). Studies were excluded if no response was obtained after three contact attempts spaced over one week apart. Data extraction used a standardized Excel spreadsheet, pilot-tested on five studies. Discrepancies were resolved by consensus between the reviewers and arbitrators.

2.5. Data Items

For primary clinical studies, extracted data included: study identification (authors, year), country and setting (university, private, public, or mixed), study design, dentition type (primary/permanent), follow-up period (when applicable), sample characteristics (number of patients, teeth, restorations, or images), restoration type and material, index diagnostic test(s), reference standard (when applicable), reported diagnostic or treatment decision thresholds (predefined or author-defined), diagnostic performance outcomes (e.g., sensitivity, specificity, AUC, kappa, ICC), and secondary caries prevalence when calculable.

For clinical guidelines and consensus documents, extracted data included document scope, terminology used for secondary caries, diagnostic approaches, management decision-making thresholds, management recommendations, and the availability of practical decision-support tools. For research protocols, extracted data focused on study objectives, targeted diagnostic methods or thresholds, planned outcomes, and anticipated contributions to clinical decision-making. The complete data extraction dataset is available at the OSF repository.

2.6. Data Synthesis

The extracted data were collected and summarized descriptively to map the existing evidence on secondary caries detection and management decision-making. Results were grouped according to the type of evidence: (i) primary clinical studies; subdivided by diagnostic method (e.g., visual–tactile, radiographic, adjunctive technologies) and by the decision thresholds that guided management (e.g., visual criteria, radiographic depth, surface texture, presence of soft tissue, or restoration integrity); (ii) recommendations and concepts from guidelines were synthesized into flowcharts and the critical appraisal was performed using Joanna Briggs Institute (JBI) tools [33,34,35]. The results were presented in tables and used to discuss the quality of the evidence but this did not influence the inclusion of studies.

3. Results

3.1. Sample Selection

The combined search yielded 11,479 records. After duplicate removal and two-stage screening, 48 documents met the PCC and eligibility criteria, comprising 34 primary studies [7,9,19,20,22,23,24,25,27,28,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59] and 14 guidelines/consensus statements [8,10,12,13,14,15,60,61,62,63,64,65,66,67] (Figure 2; Supplementary Material B). Primary clinical studies constituted the main body of evidence and were included for data extraction and methodological appraisal. Guidelines and consensus documents were analyzed separately to map how secondary caries detection, management decision-making, and management are framed in authoritative clinical recommendations. In addition, three published research protocols were identified [68,69,70], these were linked alongside the primary study that generated the data included in this review, for methodological context in the data collection phase (Table 1).

3.2. Studies Characteristics

Among the 34 primary clinical studies, twenty-one (21/34) primarily evaluated diagnostic performance, although analytical strategies varied across investigations. Nine studies quantified diagnostic performance using accuracy-based metrics, including sensitivity, specificity, AUC, or related measures, typically supported by an explicit reference standard (e.g., clinical inspection or expert consensus) [24,25,27,36,45,46,49,51,57]. Five studies assessed diagnostic performance through examiner agreement, reporting inter- or intra-observer reliability metrics such as kappa or intraclass correlation coefficients, particularly when evaluated methods relied on human interpretation of visual–tactile or radiographic findings [22,28,39,47,48]. Another eight investigations compared diagnostic methods based on the proportion of detected versus non-detected lesions and/or their resulting treatment decisions. These studies reported descriptive comparisons across methods, criteria or time points [19,23,37,40,41,43,58,59].

In addition, a subset of observational studies (7/34) was primarily designed to compare alternative diagnostic threshold frameworks for secondary caries while maintaining similar underlying detection methods [9,20,50,53,56,58,59]. These investigations evaluated how different visual criteria or grading systems, such as FDI-based criteria, CARS, modified USPHS, or combined clinical–radiographic frameworks, translated into treatment decisions (monitoring, repair, or replacement). One study adopted a health-economic perspective (1/34), evaluating secondary caries management strategies through cost-effectiveness modelling [7]. One randomized clinical trial (1/34) compared different visual criteria for secondary caries detection and their impact on treatment decisions [52].

Five non-randomized prospective clinical studies (5/34) evaluated secondary caries as a longitudinal clinical outcome within comparisons of restorative materials or restorative strategies. In these investigations, visual–tactile and radiographic criteria were applied pragmatically during follow-up, while the analytical emphasis was placed on caries occurrence, restoration failure, or replacement over time, rather than on diagnostic performance or decision thresholds [38,42,44,54,55]. Figure 3 summarizes the distribution of study designs, settings and origins, sample characteristics, and reference standards across the included studies, highlighting the predominance of diagnostic-oriented investigations (21/34).

3.3. Methodological Quality of the Evidence and Studies Limitations

A critical appraisal of the included primary studies was conducted using the appropriate Joanna Briggs Institute (JBI) checklists according to study design. Diagnostic accuracy studies showed moderate methodological heterogeneity. Most studies presented low risk of bias for outcome measurement and timing, as index tests and reference standards were generally applied within comparable periods [7,9,17,19,20,22,23,24,25,27,36,37,39,40,41,43,45,46,47,48,49,50,51,53,56,57,58,59].

However, important limitations were identified. Sampling methods were frequently unclear or at high risk of bias, with several studies relying on convenience or enriched samples, limiting representativeness [17,24,25,36,39,40,41,45,48,51,57,58]. In addition, threshold pre-specification was rarely reported [22,23,24,25,27,36,41,45,46,47,59], and blinding between index tests and reference standards was often unclear, increasing the risk of biased interpretation of diagnostic findings [24,25,43,45,46,47,51,57].

Regarding the reference standard, restoration removal was commonly used but typically restricted to teeth with clear clinical indication, while early or uncertain lesions were assessed using alternative comparators [22,40,41,43,57,59]. This selective verification may have influenced both diagnostic accuracy estimates and their translation into clinical management decision-making.

Quasi-experimental studies showed acceptable methodological structure but were limited by unclear group comparability, incomplete follow-up, and insufficient reporting of outcome assessment [38,42,44,54,55]. Only one randomized controlled trial was identified [52], presenting low risk of bias across all domains, but the limited number of RCTs restricts the strength of the available evidence. Overall, these limitations contribute to variability in how diagnostic information is interpreted and applied in clinical management decision-making. The overall risk of bias across study designs is summarized in Supplementary Material C.

3.4. Clinical Guidelines and Recommendations for Secondary Caries Management

Fourteen clinical guidelines, consensus statements, or recommendation documents addressing secondary caries were identified (Table 2). These documents were published between 2010 and 2024 and issued by international scientific societies and professional dental organizations, including the European Organization for Caries Research (ORCA), the European Federation of Conservative Dentistry (EFCD), the German Society of Conservative Dentistry (DGZ), the FDI World Dental Federation, the International Caries Classification and Management System (ICCMS™) and CariesCare International groups, and the American Academy of Pediatric Dentistry (AAPD).

Two documents (2/14) were specifically focused on secondary caries detection and management decision-making. The EFCD–ORCA–DGZ Delphi consensus on proximal and secondary caries [15] and the ORCA/EFCD consensus on secondary caries detection, management decision-making, and management [8] explicitly addressed diagnostic approaches and clinical decision criteria for secondary caries. Restoration-evaluation frameworks were represented by the original FDI criteria [12] and its revised version [13] (2/14), which incorporated secondary caries within standardized visual–tactile scoring systems linked to management actions.

Several documents addressed secondary caries within broader caries management systems (4/14), including the ICCMS™ Guide for Practitioners and Educators [64], the CariesCare practice guide [65], and the transition from ICDAS to CariesCare International frameworks [14]. Terminology-focused guidance was provided by the ORCA/IADR consensus on caries terminology [66], which defined secondary caries conceptually without proposing diagnostic thresholds or management algorithms.

Pediatric-oriented guidance documents were issued by the American Academy of Pediatric Dentistry, including the Guideline on Restorative Dentistry [60], Pediatric Restorative Dentistry: Best Practices [61], Policy on Medically-Necessary Care [62], and the Fluoride Therapy: Best Practices [63]. Additional recommendations relevant to lesion assessment and progression included the ORCA/EFCD clinical recommendations for caries diagnosis and activity assessment [10] and guidance on the evidence base for professional and self-care prevention [67].

Across documents, visual–tactile examination, commonly supplemented by bitewing radiographs, was consistently described as the primary approach for detecting secondary caries. Explicit numeric diagnostic thresholds were uncommon; instead, most guidelines operationalized management decision-making through categorical or score-based criteria linked to lesion severity, cavitation, dentine involvement, or lesion activity. Extracted data from clinical guidelines and consensus documents were systematically charted, and reported detection/diagnostic criteria and management pathways were organized into decision frameworks to visually summarize the findings.

3.5. Which Detection or Diagnostic Methods are Reported for Assessment of Secondary Caries Lesions Around Restorations or Sealants?

Visual–tactile (VT) examination, frequently operationalized through structured criteria (e.g., USPHS/Ryge, FDI, CARS, ICDAS-based approaches), was the core detection method for secondary caries around restorations or sealants, either as the primary index method or as the clinical comparator [9,20,22,23,27,28,36,38,39,40,41,42,43,44,48,49,52,53,55,56,58,59].

This VT-centered approach is consistent with guidance documents, which consistently position standardized visual–tactile assessment (often under clean, dry, and illuminated conditions) as the first-line method for secondary caries evaluation [8,10,12,13,14,15,60,61,62,63,64,65,66,67]. Specificity is generally high (≈ 0.85), but sensitivity is lower (< 0.60) for lesions hidden beneath apparently intact margins [36,48].

Radiographic assessment, particularly bitewing radiographs (BW), was the most frequently reported adjunct method and was used to support detection in areas where VT examination is limited (e.g., proximal or submarginal regions) and to support diagnostic interpretation, particularly regarding lesion depth staging [7,9,19,20,25,28,36,37,39,42,43,45,48,49,50,51,52,56]. Alternative radiographic strategies, including panoramic and periapical imaging, were occasionally reported, particularly in AI-assisted workflows [46,47]. In guidance documents, BW radiographs are consistently framed as an adjunct to VT examination, particularly when clinical inspection is limited [8,10,15,57,58,64,65]. Bitewings performed better than panoramic or periapical images for detecting proximal and submarginal caries, offering improved visualization of the cemento-enamel junction and reduced anatomical overlap [19,43,57].

A smaller group of studies evaluated light-based or optical adjuncts intended to complement VT and radiographic assessment. Laser fluorescence devices (e.g., DIAGNOdent) and fluorescence-based cameras (e.g., VistaCam) were assessed in clinical cohorts, typically alongside VT and/or BW, with reported performance varying according to threshold definition and the presence of confounders such as marginal staining or defects [7,22,27,36,49]. Fibre-optic transillumination (FOTI) was explored in earlier clinical comparisons [37,49], and optical coherence tomography (OCT) was evaluated as an adjunct imaging modality [37]. While these approaches were generally positioned as sensitivity-enhancing tools, their clinical interpretation remained dependent on threshold definition and verification strategy.

Intraoral and digital photography was reported either to support clinical documentation or as an input for automated analysis. Studies using photographic assessment included comparisons against clinical validation frameworks and expert-based interpretation [23,24], with additional use in multimodal workflows [27]. These approaches were consistently framed as complementary and required careful interpretation to distinguish discoloration and marginal defects from true lesions adjacent to restorations.

Finally, artificial intelligence (AI) approaches represented the most recent diagnostic strand, using radiographs and/or photographs as inputs to convolutional neural network-based or object-detection models. AI systems were evaluated on bitewings for secondary caries identification [25,45,46,51,57] and within multi-source imaging pipelines integrating panoramic and bitewing data [47]. AI applied to photographic datasets was also reported [24]. Across these studies, AI was generally positioned as a second-reader or decision-support tool, with acknowledged limitations related to single-centre datasets and limited external validation [25,45,47,51].

3.6. What Thresholds Guide the Decision-Making for the Treatment of Secondary Caries?

Standalone decision thresholds were rarely predefined in the included evidence. These thresholds do not operate in isolation but emerge from the integration of detection and diagnostic interpretation processes, ultimately guiding management decision-making. Instead, treatment decisions for secondary caries were guided by the concurrent use of clinical, radiographic, and index-based criteria, which together defined implicit operative thresholds.

Indicators of biological activity represented the most consistent trigger for operative and non-operative management. Soft or leathery dentine on probing, plaque stagnation adjacent to restoration margins, chalky or opaque enamel, and signs of rapid lesion progression were repeatedly associated with decisions toward active monitoring, non-operative, or operative approaches [9,20,22,27,38,39,40,41,48,50,52,53,54,56,58,59]. These criteria align with guideline frameworks that position lesion activity and cavitation as primary decision nodes preceding operative intervention [8,12,13,15,64].

Radiographic assessment was used as an auxiliary tool for lesion staging, particularly when restorations limited direct visualization and probing [19,36,41,43,48,49]. Radiographs contributed mainly by estimating dentine involvement beneath restorations, supporting decisions regarding monitoring, non-operative or operative approaches [52,56]. Within structured evaluation systems, radiographic findings were interpreted alongside clinical findings rather than as independent diagnostic thresholds [12,13,14,64,65].

Other modifiers, including patient symptoms, lesion depth toward the inner dentine third, and overall caries risk, consistently shifted decisions toward earlier intervention in higher-risk scenarios [12,16,17,38,40,64]. Marginal gaps or isolated staining were frequently associated with restoration replacement; however, most studies and guidelines agreed that these findings, when considered in isolation, should not be interpreted as evidence of secondary caries [9,13,20,48,50,52,53,54,56,58,59,65]. A decision-analytic model indicated that intervening before radiographic evidence of dentine involvement increases treatment costs and accelerates the restorative cycle, supporting higher operative thresholds in low-risk patients [7,14]. An integrated synthesis of the evidence identified in this review is presented in Figure 4, combining detection methods, diagnostic thresholds, and their translation into treatment decisions, along with current guidance and identified gaps.

3.7. What Recommendations can be Made for Future Clinical Studies Investigating the Detection and Decision-Making for Secondary Caries?

A consistent trend toward conservative, risk-oriented management of secondary caries lesions was observed in the literature. Monitoring, refurbishment, or repair were commonly associated with acceptable outcomes when lesions were inactive, uncertain, or limited in extent, whereas full restoration replacement was more frequently reported in association with lesions showing clear activity or progression [7,9,12,20,53].

An important methodological gap concerns how management decisions are operationalized and reported. Many studies described treatment outcomes without clearly specifying how lesion activity, accessibility, radiographic staging, and patient-level factors were integrated into management decision-making. Structured frameworks, such as the FDI criteria, ICCMS™, and CariesCare, offer shared terminology and decision structures that may enhance transparency compared with unstructured clinical judgment, although their application ultimately depends on professional interpretation and clinical experience [13,14,64,65].

Heterogeneity was also observed in the interpretation of restoration-related changes, including marginal staining, ditching, or gap formation. Several clinical cohorts indicated that these findings, when considered in isolation, were weak indicators of biologically active secondary caries and should not be equated with disease presence [9,48,50,52,53,54]. Distinguishing restoration aging or defects from true caries outcomes is therefore essential, as treating these entities as equivalent may inflate secondary caries prevalence and obscure disease-related events [7,10,12].

Overall, future investigations would benefit from explicitly situating management decisions within defined interpretative frameworks while acknowledging the inherent subjectivity of secondary caries assessment and the role of clinical experience. Clear reporting of how restorative defects, lesion activity, and progression are distinguished may improve consistency across studies and support more meaningful comparisons of management strategies.

3.8. Is it Possible to Create an Evidence-Based Framework from Detection to Decision-Making for Secondary Caries?

The synthesis of the included clinical studies and guidelines supports an evidence-informed framework linking secondary caries detection, diagnostic interpretation, and management decision-making. Rather than defining a single operative threshold, the available evidence suggests a sequential approach in which visual-tactile findings, radiographic information, lesion activity, and patient-level factors are interpreted together.

Figure 5 summarizes the detection and diagnostic interpretation pathway. Visual-tactile assessment remains the first-line approach, particularly for accessible occlusal and free surfaces, whereas bitewings provide complementary information when direct inspection is limited, especially in proximal or submarginal areas. Within this pathway, radiographic findings mainly support lesion depth estimation and suspected dentine involvement. However, enamel-limited or uncertain radiographic findings should be interpreted cautiously and not automatically translated into operative treatment.

Diagnostic categories were translated into management decision-making and presented in Figure 6. Findings consistent with sound restorations, inactive lesions, or uncertain/non-cavitated lesions generally support monitoring, preventive care, refurbishment, sealing, or repair. In contrast, operative intervention is more consistently supported when there is evidence of lesion activity, cavitation, soft dentine, progression, or likely dentine involvement. Restoration-related findings such as marginal staining, ditching, gaps, or discoloration may contribute to diagnostic uncertainty but should not be considered sufficient evidence of secondary caries when assessed in isolation.

Although this framework is consistent with the available evidence, its application remains limited by heterogeneous diagnostic thresholds, variable reference standards, and the scarcity of controlled studies evaluating management outcomes. Future studies should explicitly report how diagnostic findings are translated into management decisions and should evaluate whether these decisions improve clinically relevant outcomes beyond lesion detection alone.

4. Discussion

This scoping review mapped the available clinical literature on detection methods and management decision-making thresholds for secondary caries. The findings reinforce that secondary caries detection, diagnostic interpretation and management decision-making differ from primary caries, as existing restorations inherently limit direct lesion access and introduce diagnostic uncertainty at both early and advanced stages of disease [8,66], highlighting the structural features of the tooth–restoration interface that may complicate lesion detection and interpretation.

Visual–tactile examination combined with bitewings is the most commonly used approach for the assessment of secondary caries; however, important limitations persist, particularly in situations where restorations obscure lesion access. While this strategy performs reliably in clearly sound or cavitated conditions, its ability to resolve ambiguous findings (e.g., hidden lesions, minor restorative margin defects or suspected subsurface lesions) is limited. In these scenarios, conventional methods tend to show high specificity but only moderate sensitivity beneath apparently intact margins, increasing diagnostic uncertainty and variability in interpretation [9,36,48]. These observations are consistent with previous systematic reviews addressing CARS detection [16,17,18,26]. As a result, the clinical challenge extends beyond lesion detection to include diagnostic interpretation under uncertainty, particularly in avoiding both over- and under-interpretation of findings that may lead to inappropriate operative intervention [7,8].

Beyond lesion detection, the evidence supports the relevance of lesion activity–based criteria in guiding conservative and targeted management strategies. Longitudinal clinical cohorts showed that management decision-making grounded in lesion activity, progression, and biological behavior, rather than restorative marginal defects alone, was associated with lower restoration replacement rates without compromising clinical outcomes over time [9,20,52]. These findings align with contemporary caries management concepts that prioritize biological processes and patient risk over purely morphological or aesthetic features.

Radiographic assessment played a complementary role in this process, rather than being used in isolation to support diagnostic interpretation and, in turn, guide management decisions in uncertain clinical scenarios. Bitewings were mainly used to identify lesions not directly accessible to clinical inspection and to estimate dentine involvement beneath existing restorations, thereby estimating the expected invasiveness of management [9,20,41,48]. This perspective is further supported by clinical investigations in primary teeth, suggesting that the benefits of adjunctive methods, such as bitewings, might be limited depending on the clinical context [21]. While combining detection strategies often enhances sensitivity, it may concurrently lead to a reduction in specificity without necessarily improving the overall clinical utility [21,71]. Therefore, in scenarios with a lower prevalence of advanced lesions, a more cautious integration of adjunctive information is recommended to avoid shifts toward more invasive treatment decisions that do not translate into clear outcomes for the patient [12,13].

Other adjunctive diagnostic technologies, including fluorescence-based devices, high-resolution intraoral photography, and artificial intelligence–based systems, have been explored as complementary tools to conventional assessment of caries around restorations [8,26,57]. While several studies reported context-dependent gains in sensitivity, these often occur at the expense of specificity, which is a critical concern in populations where the prevalence of active secondary lesions is low. In such scenarios, even a marginal decrease in specificity disproportionately increases the rate of false positives, potentially driving a cycle of unnecessary re-treatments. This phenomenon reinforces spectrum bias often found in diagnostic studies, where technologies are tested against advanced lesions that do not represent the subtle diagnostic ‘grey zone’ encountered in high-functioning clinical practices [26,71,72]. In this context, methodological heterogeneity, inconsistent reference standards, and limited external validation restrict the current applicability of these technologies in routine management of secondary caries [8,26].

Importantly, adjunctive methods did not consistently improve management decision-making in clinically unequivocal situations. Their potential value emerged mainly in scenarios where visual–tactile assessment was constrained by restorative overlap or subsurface lesion progression, representing the diagnostic grey zone of secondary caries [22,27,37,49]. In such contexts, the clinical relevance of adjunctive tools depends less on maximizing sensitivity and more on supporting calibrated interpretation under conditions of uncertainty [22,27,37,49].

Across clinical studies and guideline frameworks, management decision-making thresholds consistently prioritized lesion activity and dentine involvement over marginal appearance alone [38,40,48,53]. The application of lesion-centred criteria, such as CARS-based thresholds, has been associated with reduced restoration replacement without compromising outcomes in observational and clinical studies [9,20,50,52,56]. These findings are coherent with guideline-based frameworks that position cavitation and lesion activity as primary decision nodes preceding operative intervention [8,12,13,64].

From a broader perspective, the findings suggest that advances in secondary caries detection should be judged by their ability to inform decision-making in clinically ambiguous situations rather than by gains in sensitivity alone [7,8]. Methods capable of resolving uncertainty without disproportionately increasing false-positive findings are more likely to contribute meaningfully to patient-centred care [23,53].

A previous appraisal of clinical practice guidelines using the AGREE II instrument identified substantial variability in methodological rigor and limited guidance on implementation, particularly regarding economic considerations and stakeholder involvement [73]. These observations are consistent with the present review, in which recommendations for secondary caries detection and management were conceptually aligned but inconsistently operationalized across documents.

Despite growing evidence, secondary caries remains insufficiently defined as a standardized outcome, with heterogeneous reporting of diagnostic criteria, lesion staging, and decision thresholds across studies. This lack of standardization limits comparability and complicates translating diagnostic performance into consistent clinical management decision-making. Preventive and risk-based strategies aimed at maintaining patients at low caries risk are therefore likely to reduce secondary caries incidence and progression, particularly in the early stages. However, limited evidence supporting minimally invasive interventions for initial and moderate secondary caries lesions represents a critical gap that may inadvertently contribute to unnecessary restoration replacement.

This review is subject to limitations, including the possibility of missed studies, publication bias, and the inherent subjectivity of study selection and evidence synthesis. Critical appraisal was conducted using JBI tools focused on diagnostic performance principles, even though not all included studies were designed as diagnostic accuracy investigations. Substantial heterogeneity in study design, diagnostic methods, reference standards, and outcome definitions also limited deeper comparative analyses. Furthermore, since management decision-making thresholds were frequently reported as secondary observations within restorative survival studies rather than as primary outcomes, the identification of specific intervention criteria often relied on interpretative extraction of implicit clinical protocols. Finally, consistent with the scoping review methodology, a quantitative synthesis of diagnostic accuracy or treatment effects was not performed; thus, the comparative effectiveness of the identified strategies is presented descriptively rather than statistically.

5. Conclusion

Clinical evidence and guidance documents converge on visual–tactile assessment complemented by bitewings as first-line strategies for CARS evaluation. Adjunct detection tools showed context-dependent performance, particularly in proximal or visually restricted areas, but still lack robust external validation. Management decision-making is predominantly driven by activity-related signs and suspected dentine involvement, supporting conservative management when diagnostic uncertainty is high. Standardized reporting of thresholds and decision frameworks is a key priority for future clinical research, alongside studies evaluating patient-reported outcomes such as sensitivity, satisfaction, and oral health-related quality of life.