Chuchu 2018 — Cochrane review: teledermatology for diagnosing skin cancer in adults
Citation
Chuchu N, Dinnes J, Takwoingi Y, Matin RN, Bayliss SE, Davenport C, et al. Teledermatology for diagnosing skin cancer in adults. Cochrane Database Syst Rev. 2018 Dec 4;12(12):CD013193. DOI: 10.1002/14651858.CD013193.
Study design and population
Cochrane-grade systematic review of diagnostic-accuracy studies; 22 studies included; QUADAS-2 risk-of-bias assessment. Adults with skin lesions assessed for skin cancer via teledermatology vs. in-person dermatologist or histopathology reference.
Reported metrics
- Pooled summary sensitivity for correct malignant classification via photographic images: 94.9 % (95 % CI 90.1–97.4); specificity 84.3 % (95 % CI 48.5–96.8)
- Referral-accuracy analyses: teledermatology correctly triaged > 90 % of lesions that face-to-face specialists considered to require action
- 7 studies reported diagnostic accuracy for any skin cancer; 9 for invasive melanoma; 4 for BCC; 6 reported only referral accuracy
Surrogate-to-outcome linkage
Cochrane-grade evidence linking teledermatology triage to the patient-relevant outcome of appropriate referral for suspected skin cancer. Formally confirms referral accuracy as a clinically accepted endpoint for teledermatology and, by analogy, for AI-triage CDS devices operating in the same workflow.
CRIT1–7 appraisal
| Criterion | Score | Justification |
|---|---|---|
| CRIT1 Relevance | 3 | Direct — teledermatology skin-cancer triage. |
| CRIT2 Methodology | 3 | Cochrane methodology; QUADAS-2; formal diagnostic-test-accuracy review. |
| CRIT3 Reporting | 3 | Pooled sensitivity / specificity with 95 % CIs reported. |
| CRIT4 Applicability | 3 | Directly applicable; Cochrane-grade authority for HTA / regulator. |
| CRIT5 Evidence weight | 3 | Cochrane systematic review — highest tier. |
| CRIT6 Risk of bias | 2 | Heterogeneity precluded full pooling; many constituent studies observational. |
| CRIT7 Contribution | 3 | Core aggregate anchor — highest evidence-quality endorsement of referral-accuracy metrics. |
Aggregate: very strong.
Limitations and notes
Heterogeneity constrained meta-analytic pooling; specificity CI very wide (48.5–96.8) reflecting that heterogeneity.
Strength as anchor
Very strong — Cochrane-grade anchor for the accepted-surrogate claim. Regulator-facing weight maximal. Pairs with Finnane 2017 and Bourkas 2023 as the systematic-review core of Domain 3.