📋 Sources (primary literature):
ABCDE: Friedman RJ et al. Early detection of malignant melanoma: The role of physician examination and self-examination of the skin. CA Cancer J Clin 1985;35(3):130–51.
7-Point Checklist: MacKie RM. Clinical recognition of early invasive malignant melanoma. BMJ 1990;301:1005–6. Revised 1989: Clin Exp Dermatol 1991;16:151–3. PMID 1867692 ↗
3-Point Checklist: Soyer HP et al. Three-point checklist of dermoscopy. Dermatology 2004;208:27–31. PMID 14730233 ↗
Chaos & Clues: Rosendahl C et al. Dermatoscopy in routine practice — ‘chaos and clues’. Aust Fam Physician 2012;41(7):482–7. PMID 22762066 ↗
Chaos & Clues (seborrhoeic keratosis exclusion & refined accuracy): Rosendahl C, Tschandl P, Cameron A, Kittler H. Diagnostic accuracy of dermatoscopy for melanocytic and nonmelanocytic pigmented lesions. J Am Acad Dermatol 2011;64(6):1068–73. PMID 21440329 ↗; and Rosendahl C, Cameron A, Tschandl P, Bulinska A, Gourhant J-Y, Keir J, Kittler H. Chaos & Clues: A Dermatoscopic Decision Algorithm for Pigmented Skin Malignancy [poster]. The University of Queensland & Medical University of Vienna, 2015.
Menzies Method: Menzies SW et al. Frequency and morphologic characteristics of invasive melanomas lacking specific surface microscopic features. Arch Dermatol 1996;132:1178–82. PMID 8859028 ↗
CASH: Henning JS et al. The CASH algorithm for dermoscopy. J Am Acad Dermatol 2007;56:45–52. PMID 17190620 ↗
Validation across methods: Carrera C et al. Validity and reliability of dermoscopic criteria. JAMA Dermatol 2016;152:798–806. PMID 27074267 ↗ All algorithms have known false-negative rates. No algorithm replaces clinical judgement. When in doubt, excise or refer.
🚫 Clinicians only. These tools calculate published algorithm scores. Scores are a reference only — excision, biopsy, and referral decisions are yours as the treating clinician. No algorithm replaces clinical judgement. All algorithms have known false-negative rates. When in doubt, excise or refer.
Clinical / naked eye assessment. No dermoscope required. Tick each feature that is present.
Sensitivity: varies (E is highest value feature)Friedman RJ et al, J Am Acad Dermatol 1985
✓
A — Asymmetry
Lesion is asymmetrical when divided through any axis. One half does not mirror the other.
✓
B — Border irregularity
Border is uneven, ragged, notched, or blurred. Not a smooth, well-defined edge.
✓
C — Colour variation
More than one shade of brown or black, or areas of red, white, or blue within the lesion.
✓
D — Diameter >6mm
Larger than a pencil eraser at broadest point. Note: early melanomas may be <6mm.
✓
E — Evolution / change
Any change in size, shape, colour, or new symptom (itch, bleed) over weeks to months. This is the most clinically significant feature.
Clinical / naked eye. Major features score 2 points; minor features score 1 point. Score ≥3 = refer for specialist opinion per original criteria.
The lesion has grown or changed in size recently (patient-reported or documented).
✓
Change in shape +2 pts
Irregular or changing outline, notching, or satellite lesions appearing.
✓
Change in colour +2 pts
Variation or mottling of colour within the lesion, or overall colour change from prior appearance.
Minor criteria (1 point each)
✓
Inflammation +1 pt
A red inflammatory halo or flush around the edge of the lesion.
✓
Crusting or bleeding +1 pt
Oozing, crusting, or spontaneous bleeding not attributable to trauma.
✓
Sensory change — itch +1 pt
New itch or altered sensation in or around the lesion. Note: itch alone has low specificity.
✓
Diameter ≥7mm +1 pt
Largest diameter is 7mm or more at clinical examination.
Dermoscopy. Simplified entry-level algorithm. Score 2 or 3 = suspicious.
Sensitivity: 96.3%Specificity: 32.8%Score ≥2 = suspiciousSoyer HP et al, Dermatology 2004; PMID 14730233
✓
Asymmetry of colour and/or structure
Asymmetry in one or both perpendicular axes for colour AND/OR dermoscopic structures (network, globules, dots, vessels).
✓
Atypical (irregular) pigment network
Black or brown network with irregular meshwork, variable thickness, or abruptly ending at the periphery.
✓
Blue-white structures
Any blue and/or white colour including blue-white veil, regression structures (white scar-like areas, blue pepper-like granules).
Dermoscopy. Australian algorithm. Step 1: Is there chaos? If yes, look for clues. Works for all pigmented lesions — melanocytic and non-melanocytic.
Sensitivity: 90.6%Specificity: 62.7%Rosendahl C et al, Aust Fam Physician 2012; PMID 22762066
⚠️ Exceptions — still consider biopsy or referral even with no chaos or no clue:
A changing lesion in an adult, especially with increasing age, with either a history of change or dermoscopic evidence of change (e.g. peripheral clods, radial lines/pseudopods).
A nodular lesion, or a very small lesion, with any clue to malignancy.
Head/neck: a pigmented lesion, or one with dermatoscopic grey colour.
Palms or soles (acral): a lesion with a parallel ridge pattern.
Per Rosendahl et al (2012) — this algorithm is a diagnostic aid, not a rule; no method can guarantee detection of every malignancy.
Step 1 — Is there CHAOS?
Chaos = asymmetry of structure and/or colour, regardless of the outline shape.
✓
Yes — chaos is present
Asymmetry of colour and/or structures. Proceed to Step 2 (clues).
✓
No — no chaos / lesion is symmetrical
If no chaos, the lesion is likely benign. Note exceptions (nodular melanoma, acral lesions). Stop here.
✓
Unsure
Per Rosendahl: if you cannot decide, manage as chaotic. Proceed to clues.
Step 2 — Are any CLUES present?
Tick all clues visible. Chaos + any one clue = consider excision or referral.
✓
Eccentric structureless area
Any colour except skin colour. An area devoid of structures, not central, not symmetric.
✓
Grey or blue structures
Any grey or blue colour in any pattern (dots, globules, structureless areas).
✓
Peripheral black dots or clods
Black dots or globules at or near the periphery of the lesion (not scattered throughout).
✓
Segmental radial lines or pseudopods
Radial projections or pseudopods (lines with bulbous ends) at the periphery, in a segmental (not all-round) pattern.
✓
White lines (shiny white streaks)
Orthogonal or parallel bright white lines visible on polarised dermoscopy. Not the same as regression white areas.
✓
Thick reticular lines
Pigment network with markedly thickened lines (broadened network), not fine regular network.
✓
Polymorphous vessels
Two or more different vessel types within the same lesion (dotted, linear, irregular, hairpin, etc.).
✓
Angulated lines (polygons)
Lines meeting at angles forming polygon-like structures. Associated with lentigo maligna.
✓
Parallel lines on ridges (acral)
For palmar/plantar/acral sites only. Parallel pattern on the skin ridges (not furrows) = suspicious.
Before biopsy — could this be a seborrhoeic keratosis?
Chaos + a clue can still be a seborrhoeic keratosis (SK) or solar lichen-planus-like keratosis (LPLK). Tick any of these pattern-analysis clues to SK/LPLK that are also present — an optional second look before proceeding to biopsy, not a replacement for it.
✓
Multiple orange clods
Comedone-like openings — several orange/brown round structures.
✓
Multiple white clods
Milia-like cysts — several white/yellow round structures.
✓
Thick curved lines
Broad, curved (not straight/angulated) ridge-and-fissure lines — a "brain-like" or cerebriform surface.
✓
Sharply demarcated border
A sharp, well-defined edge over the entire periphery of the lesion — the "stuck-on" look.
✓
Multiple similar lesions grouped nearby
Other lesions of similar appearance clustered in the same skin area.
Dermoscopy. Two-step: first exclude melanoma with negative features, then look for positive features. No negative features + ≥1 positive feature = melanoma per original criteria.