evidencegaps
Findings

What do systematic reviews say?

Refine by population, intervention, comparison or outcome
More filters
Evidence found and GRADE certainty

398 questions · 757 findings · 66 reviews · Oral Health

Sort:Best matchNewestHighest certaintyDownload CSV
No studies found 20 · 2.6%GRADE not reported 574 · 75.8%Very low 39 · 5.2%Low 49 · 6.5%Moderate 53 · 7%High 22 · 2.9%

Laser fluorescence (DIAGNOdent 2190/pen) vs Histological validation (serial sectioning, grinding, hemisection, or microradiography)

Diagnostic accuracy for occlusal caries detection at the caries detection level (D0 versus D1-D4) · Extracted human permanent posterior teeth (premolars and molars), occlusal (pit-and-fissure) surfaces

In in vitro, laser fluorescence pen (2190) pooled sensitivity was 0.78 (95% CI 0.44 to 0.94) for occlusal caries at the caries detection level; ungraded.

4 findings from 1 review
Show 1 more finding

Digital bitewing radiography, sensor-based (CCD/CMOS) vs Direct visual examination after tooth separation or bioptical cavity preparation

Diagnostic accuracy for proximal caries detection (dentin detection level) · Patients with permanent posterior teeth, proximal surfaces assessed clinically (in vivo)

Under in vivo conditions, sensor-based digital bitewing radiography had a pooled sensitivity of 0.96 (95% CI 0.91, 0.98) for proximal caries detection at the dentin detection level; ungraded.

4 findings from 1 review
Show 1 more finding

Quantitative light-induced fluorescence (QLF) vs Histological validation (in vitro) or cavity preparation/biopsy (in vivo)

Occlusal caries detection accuracy · Human permanent posterior teeth (premolars and molars) in patients, occlusal (pit-and-fissure) surfaces

No eligible QLF studies met the meta-analysis inclusion criteria (candidate QLF studies were identified but none had acceptable index/reference tests and reporting quality).

1 finding from 1 review

Fibre-optic transillumination (FOTI) vs Histological validation (serial sectioning, grinding, hemisection, or microradiography)

Diagnostic accuracy for occlusal caries detection at the caries detection level (D0 versus D1-D4) · Extracted human permanent posterior teeth (premolars and molars), occlusal (pit-and-fissure) surfaces

In in vitro, fibre-optic transillumination pooled sensitivity was 0.97 (95% CI -0.92 to 0.99) for occlusal caries at the caries detection level; ungraded.

4 findings from 1 review
Show 1 more finding

Visual examination (VE) vs Direct visual examination after tooth separation or bioptical cavity preparation

Diagnostic accuracy for proximal caries detection (dentin detection level) · Patients with permanent posterior teeth, proximal surfaces assessed clinically (in vivo)

Under in vivo conditions, visual examination had a pooled specificity of 0.76 (95% CI 0.11, 0.99) for proximal caries detection at the dentin detection level; ungraded.

4 findings from 1 review
Show 1 more finding

Visual examination (VE) vs Histological validation of dental tissues (gold standard)

Diagnostic accuracy for proximal caries detection (dentin detection level) · Extracted permanent posterior teeth with proximal surfaces (in vitro)

Under in vitro conditions, visual examination had a summary ROC AUC of 0.85 for proximal caries detection at the dentin detection level; ungraded.

4 findings from 1 review
Show 1 more finding

Fibre-optic transillumination (FOTI) vs Histological validation (in vitro) or direct visual examination after tooth separation / bioptical cavity preparation (in vivo)

detection of proximal caries · Permanent posterior teeth, proximal surfaces (in vitro extracted teeth and in vivo patients)

For fibre-optic transillumination (FOTI), studies were identified (in vitro N=4, in vivo N=3) but none had sufficiently low risk of bias and adequate reporting to be pooled; no meta-analytic estimate could be derived.

1 finding from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (bitewing radiography, ADA classification, ICDAS II)

Occlusal caries detection: specificity · Patients (in vivo); permanent and primary teeth, occlusal surfaces

In vivo, QLF pooled specificity for occlusal caries in permanent teeth was 0.80 (3 studies; ungraded).

2 findings from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (histology; some ICDAS/radiography)

Occlusal caries detection: sensitivity · Extracted human teeth (in vitro); permanent and primary teeth, occlusal surfaces

In vitro, QLF pooled sensitivity for occlusal caries detection was 0.83 (5 studies; ungraded).

3 findings from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (histology; some ICDAS/radiography)

Proximal caries detection: sensitivity · Extracted human teeth (in vitro); permanent teeth, proximal (approximal) surfaces

In vitro, QLF pooled sensitivity for proximal caries detection was 0.83 (3 permanent-teeth studies; no primary-dentition data; ungraded).

1 finding from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (bitewing radiography, ADA classification, ICDAS II)

Proximal caries detection: specificity · Patients (in vivo); permanent and primary teeth, proximal (approximal) surfaces

In vivo, QLF pooled specificity for proximal caries in permanent teeth was 0.81 (3 studies; ungraded).

2 findings from 1 review

AI algorithms (predominantly CNN/deep learning) as index test vs Dental expert / visual-tactile clinical assessment (reference standard)

Sensitivity · Patients undergoing dental caries detection on dental images (teeth/surfaces across imaging modalities)

AI algorithms have high pooled sensitivity (0.85) for detecting dental caries; ungraded, with very high heterogeneity (I2>95%) and evidence of publication bias.

1 finding from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (histology; some ICDAS/radiography)

Occlusal caries detection: specificity · Extracted human teeth (in vitro); permanent and primary teeth, occlusal surfaces

In vitro, QLF pooled specificity for occlusal caries in permanent teeth was 0.66 (3 studies; ungraded).

3 findings from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (bitewing radiography, ADA classification, ICDAS II)

Occlusal caries detection: sensitivity · Patients (in vivo); permanent and primary teeth, occlusal surfaces

In vivo, QLF pooled sensitivity for occlusal caries detection was 0.86 (4 studies; ungraded).

2 findings from 1 review

AI algorithms (predominantly CNN/deep learning) as index test vs Dental expert / visual-tactile clinical assessment (reference standard)

Specificity · Patients undergoing dental caries detection on dental images (teeth/surfaces across imaging modalities)

AI algorithms have high pooled specificity (0.90) for detecting dental caries; ungraded, with very high heterogeneity (I2>95%).

1 finding from 1 review

AI algorithms (predominantly CNN/deep learning) as index test vs Dental expert / visual-tactile clinical assessment (reference standard)

Area under the summary ROC curve · Patients undergoing dental caries detection on dental images (teeth/surfaces across imaging modalities)

The area under the summary ROC curve for AI caries detection is 0.86, indicating high overall diagnostic accuracy; ungraded.

1 finding from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (histology; some ICDAS/radiography)

Proximal caries detection: specificity · Extracted human teeth (in vitro); permanent teeth, proximal (approximal) surfaces

In vitro, QLF pooled specificity for proximal caries detection was 0.74 (3 permanent-teeth studies; no primary-dentition data; ungraded).

1 finding from 1 review

AI algorithms (predominantly CNN/deep learning) as index test vs Dental expert / visual-tactile clinical assessment (reference standard)

Positive likelihood ratio · Patients undergoing dental caries detection on dental images (teeth/surfaces across imaging modalities)

A positive AI test raises caries probability substantially (LR+ 10.44; positive post-test probability 79%); ungraded.

1 finding from 1 review

AI algorithms (predominantly CNN/deep learning) as index test vs Dental expert / visual-tactile clinical assessment (reference standard)

Log diagnostic odds ratio · Patients undergoing dental caries detection on dental images (teeth/surfaces across imaging modalities)

AI algorithms show strong pooled discriminative accuracy for dental caries (log DOR 4.37; DOR 70.9); ungraded.

1 finding from 1 review

Quantitative light-induced fluorescence (QLF) vs Reference standard (bitewing radiography, ADA classification, ICDAS II)

Proximal caries detection: sensitivity · Patients (in vivo); permanent and primary teeth, proximal (approximal) surfaces

In vivo, QLF pooled sensitivity for proximal caries detection was 0.74 (4 studies; ungraded).

2 findings from 1 review