Orientation Before Reading
Critical Axes and Publication Boundaries
- Article Type: Diagnosis. The topic is organized across five clinical decision axes, not through a global judgment.
- Evidence Basis: Grade A evidence, moderate conclusion strength. 9 evaluated studies, of which 4 have full-text extraction and 5 are metadata-based. Quality Mix: 7 supporting, 2 with reservations, 0 critical.
- Bias Risk: low to moderate. Source Integrity: clean (1 PMID correction documented).
- Core Message: Caries diagnosis is not a method competition, but a concept problem. Detection and need for treatment must be evaluated separately.
Clinical Question
Which caries detection method achieves what—and where does the line begin between useful early detection and clinical overdiagnosis?
Executive Summary
The current body of literature includes 9 systematic reviews and meta-analyses on caries detection diagnosis, including studies on visual inspection, bitewing X-rays, and AI-assisted approaches. DDJ reads the evidence base as overall supportive (Grade A), but with moderate conclusion strength—because strong pooling data coexist with high heterogeneity and a partially immature methods landscape (especially for AI). [1-8]
The main direction of the corpus is diagnostic utility, but this utility is method-specific and stage-dependent. Visual inspection remains indispensable but systematically achieves low specificity for initial lesions. Bitewings complement precisely where visual inspection falls short. AI shows promising pooling data but is not yet a clinical standard due to bias risks and lack of external validation. [1,3,5]
The article organizes these findings along five clinical decision axes, linking strong statements to concrete evidence clusters rather than a global overall judgment.
How DDJ Reads This Topic
DDJ treats this topic as a diagnostic article. This means: The question is not "Does caries diagnosis work?", but rather "Which method works for which stage and which surface—and what follows therapeutically?"
The central zone of contradiction lies not between the methods themselves, but between detection performance and therapeutic consequence. High sensitivity without adequate specificity systematically leads to overdiagnosis and overtreatment in caries diagnosis. [1,3]
Clinical questions, conflict zones, and consequences become visible.
Claim Clusters and Decision Axes
Claim Cluster 1 — ddj0017_c1
Visual Inspection as Baseline Diagnostics
Clinical Axis: Visual Inspection vs. Radiographic Supplementation
Why this axis matters: Visual inspection is the universally available baseline diagnostic tool, but its limitations for initial and proximal lesions are systematically documented in large meta-analyses.
Evidence status: In a meta-analysis of 37 studies, visual inspection showed an in vivo sensitivity of 0.70 for the caries detection threshold with a specificity of only 0.47. At the dentin caries threshold, the AUC increases to 0.89. [1] For proximal caries, the AUC is 0.84, although VE alone is considered insufficient for proximal lesions. [3]
Where the signal is stable: For carious occlusal lesions and at the dentin caries threshold, visual inspection is clinically sufficiently sensitive.
Where uncertainty begins: For initial, non-carious lesions, specificity drops to 0.47 — meaning more than half of healthy areas are falsely classified as carious. For proximal lesions, sensitivity is lacking without radiographs.
Clinical implication: Visual inspection is indispensable but not a complete diagnostic concept on its own. Supplementation with bitewing radiographs must be risk-based.
Claim Cluster 2 — ddj0017_c2
Bitewings as Diagnostic Supplementation
Clinical Axis: Bitewing Indication and Added Value Compared to VE
Why this axis matters: Bitewings supplement visual inspection exactly where it systematically fails: for proximal caries and the detection of deeper dentin lesions.
Evidence status: The meta-analysis by Janjic Rankovic et al. shows that for digital bitewings, the AUC values for proximal dentin caries detection range from 0.81–0.92 with high specificity up to 0.97 in vitro. [3] The complementary diagnostic strength compared to VE for proximal lesions is documented in both large meta-analyses. [1,3]
Where the signal is stable: For proximal dentin caries, bitewings consistently provide higher detection rates than visual inspection alone.
Where uncertainty begins: Routine bitewings without risk stratification are not evidence-based. The indication should be guided by individual caries risk, not a fixed time interval.
Clinical implication: Bitewings are indicated for increased caries risk or clinical suspicion of proximal caries. Blanket interval recommendations are not supported by the corpus.