ATOF
Adaptive Threshold Optimisation Framework
🧮 Calculator📋 Brief📖 Methods⚠ LimitationsInteractive · v0.1 AEST
National Targeted Skin Cancer Screening Roadmap · Methodological framework

At what risk scoredo you draw the line between screen and don't screen?

ATOF is a decision-analytic framework that integrates clinical net benefit, overdiagnosis harms, and regional workforce capacity into a single answer to that question - for every region of Australia, under every plausible workforce scenario.

Validated models
6
45 and Up Study
AUC range
0.66–0.76
external validation
In situ overdiagnosis
67–76%
Lindsay 2024
QLD SA4s with zero derms
53%
Lindsay 2026
3.0%
today
Framework overview · Figure 3

The three-component story, in one frame

One master threshold drives all three panels. Drag below - or press â–¶ Play to walk the cascade from 1% to 7% and watch the curves deflate as the map turns red.

Master threshold3.0%
1

Net benefit

Where each model beats both 'screen everyone' and 'screen no one'. Useful range = green band.

2

Overdiagnosis penalty

Same curves, deflated by the 54% OD rate (Glasziou/Lindsay). The useful range shrinks.

3

Capacity by SA4

Workforce-feasibility at this threshold under Combined (Derm + GP). Red = over capacity.

Workforce capacity vs screening demand

Each SA4 coloured by capacity utilisation at the selected threshold. Drag the slider above and watch Australia respond.

88 SA4 regions
Component 1 · Decision Curve Analysis

Net benefit at every threshold

Decision curve analysis (Vickers & Elkin 2006) plots the population-level net benefit of using a risk model at each threshold, against screen everyone and screen no one. Curves are modelled from published AUCs using a binormal approximation.

Useful range 0.5–15.0%
2.0%
0.5%5%
QSkin MP16AUC 0.76QSkin MP7AUC 0.73MIA/VuongAUC 0.70·Screen all (treat everyone)Screen none (NB = 0 baseline)Selected threshold (live)Policy reference thresholds (3% Roadmap · 5% · 7% Ireland)Clinically useful range (0.5-4%)
Primary model
QSkin MP16
At threshold
3.0%
Sensitivity
55%
False positive rate
14.9%
Number needed to screen
92
Decision curves modelled from published AUCs using a binormal approximation (Vickers & Elkin 2006). Final ATOF analysis will reconstruct curves from individual-level 45 and Up Study data.
Component 3 · Threshold Optimiser

Where can screening start tomorrow?

For every SA4, the minimum feasible thresholdis the lowest risk cut-point at which the region's annual screening flow stays within available workforce capacity. Regions below the current national slider can run screening today; regions above it cannot, regardless of willingness.

≤2%24%
21 / 88
High capacity
2–5%59%
52 / 88
Moderate
5–7%11%
10 / 88
Constrained
>7%6%
5 / 88
Tight workforce
Unfeasible0%
0 / 88
Workforce ≈ 0
What this means for the Roadmap
Under the combined (derm + gp) scenario, a uniform national threshold of 3.0% 10-year risk is feasible in 55 of 88 SA4 regions. The remaining 33 would either need to be excluded from the program or accept demand that exceeds their workforce. 15 regions require thresholds above 5%, primarily where dermatologist density is zero or skin-cancer GP supply is thin.

Minimum feasible threshold by SA4 - strip plot

n = 88 SA4s · sorted within each state
1%2%3%5%7%10%15%NSWSydney - City and Inner South (NSW) · min 1.6% · MM1Sydney - Parramatta (NSW) · min 1.6% · MM1Sydney - Eastern Suburbs (NSW) · min 1.8% · MM1Sydney - Blacktown (NSW) · min 1.8% · MM1Sydney - Ryde (NSW) · min 1.8% · MM1Sydney - South West (NSW) · min 1.9% · MM1Sydney - Inner West (NSW) · min 2.0% · MM1Sydney - Inner South West (NSW) · min 2.0% · MM1Sydney - North Sydney and Hornsby (NSW) · min 2.0% · MM1Sydney - Outer South West (NSW) · min 2.2% · MM1Sydney - Northern Beaches (NSW) · min 2.3% · MM1Sydney - Baulkham Hills and Hawkesbury (NSW) · min 2.4% · MM1Sydney - Sutherland (NSW) · min 2.6% · MM1Sydney - Outer West and Blue Mountains (NSW) · min 2.7% · MM1Far West and Orana (NSW) · min 2.9% · MM5Illawarra (NSW) · min 2.9% · MM1Newcastle and Lake Macquarie (NSW) · min 2.9% · MM1Riverina (NSW) · min 3.1% · MM4Capital Region (NSW) · min 3.1% · MM2Central Coast (NSW) · min 3.2% · MM1Murray (NSW) · min 3.2% · MM3Central West (NSW) · min 3.4% · MM3Hunter Valley exc Newcastle (NSW) · min 3.5% · MM2Coffs Harbour - Grafton (NSW) · min 3.5% · MM3Southern Highlands and Shoalhaven (NSW) · min 3.5% · MM2New England and North West (NSW) · min 3.9% · MM3Mid North Coast (NSW) · min 4.0% · MM3Richmond - Tweed (NSW) · min 4.0% · MM3VICMelbourne - Inner (VIC) · min 1.4% · MM1Melbourne - Inner South (VIC) · min 1.7% · MM1Melbourne - Inner East (VIC) · min 1.8% · MM1Melbourne - North West (VIC) · min 1.9% · MM1Melbourne - West (VIC) · min 2.0% · MM1Melbourne - North East (VIC) · min 2.0% · MM1Melbourne - South East (VIC) · min 2.1% · MM1Melbourne - Outer East (VIC) · min 2.2% · MM1Geelong (VIC) · min 2.4% · MM1Ballarat (VIC) · min 2.5% · MM2Hume (VIC) · min 2.6% · MM3Shepparton (VIC) · min 2.6% · MM3Mornington Peninsula (VIC) · min 2.7% · MM1Bendigo (VIC) · min 2.8% · MM3Latrobe - Gippsland (VIC) · min 2.9% · MM3North West (VIC) · min 2.9% · MM4Warrnambool and South West (VIC) · min 3.0% · MM3QLDQueensland - Outback (QLD) · min 2.6% · MM6Brisbane Inner City (QLD) · min 2.9% · MM1Cairns (QLD) · min 4.5% · MM3Brisbane - South (QLD) · min 4.6% · MM1Ipswich (QLD) · min 5.3% · MM1Townsville (QLD) · min 5.9% · MM3Logan - Beaudesert (QLD) · min 6.0% · MM1Gold Coast (QLD) · min 6.3% · MM1Moreton Bay - South (QLD) · min 6.5% · MM1Brisbane - West (QLD) · min 6.5% · MM1Brisbane - North (QLD) · min 6.6% · MM1Central Queensland (QLD) · min 6.7% · MM4Sunshine Coast (QLD) · min 6.7% · MM2Mackay - Isaac - Whitsunday (QLD) · min 6.8% · MM4Toowoomba (QLD) · min 7.3% · MM3Darling Downs - Maranoa (QLD) · min 7.9% · MM4Wide Bay (QLD) · min 9.8% · MM3Brisbane - East (QLD) · min 10.2% · MM1Moreton Bay - North (QLD) · min 10.7% · MM1WAPerth - Inner (WA) · min 0.8% · MM1Western Australia - Outback (North) (WA) · min 1.8% · MM7Western Australia - Outback (South) (WA) · min 2.0% · MM6Perth - North West (WA) · min 2.5% · MM1Perth - South East (WA) · min 2.5% · MM1Perth - North East (WA) · min 2.5% · MM1Perth - South West (WA) · min 2.7% · MM1Bunbury (WA) · min 2.9% · MM2Mandurah (WA) · min 2.9% · MM2Western Australia - Wheat Belt (WA) · min 3.3% · MM4SAAdelaide - Central and Hills (SA) · min 1.6% · MM1South Australia - Outback (SA) · min 1.8% · MM6Adelaide - North (SA) · min 2.2% · MM1Adelaide - West (SA) · min 2.3% · MM1Adelaide - South (SA) · min 2.3% · MM1South Australia - South East (SA) · min 3.0% · MM3Barossa - Yorke - Mid North (SA) · min 3.1% · MM3TASHobart (TAS) · min 2.6% · MM1Launceston and North East (TAS) · min 3.0% · MM2West and North West (TAS) · min 3.0% · MM3South East (TAS) · min 3.7% · MM3ACTAustralian Capital Territory (ACT) · min 2.2% · MM1NTNorthern Territory - Outback (NT) · min 1.5% · MM7Darwin (NT) · min 1.7% · MM2national 3.0%Minimum feasible threshold (10-yr melanoma risk)
feasible at ≤2% 2–5% 5–7% >7% current threshold (3.0%)

Hardest 5 regions to screen

RegionStateRemotenessMin thresholdSource
Moreton Bay - NorthQLDMM110.7%Lindsay 2026 (real)
Brisbane - EastQLDMM110.2%Lindsay 2026 (real)
Wide BayQLDMM39.8%Lindsay 2026 (real)
Darling Downs - MaranoaQLDMM47.9%Lindsay 2026 (real)
ToowoombaQLDMM37.3%Lindsay 2026 (real)
Capacity model uses adjusted FTE × annual consultation rates ÷ 2-year screening cadence. Lindsay et al. 2026 (PMID:41674191) values for QLD; Blake et al. 2023 (PMID:37353974) anchors for NSW; AIHW NHWDS + ACD workforce snapshot for other states.
Equity geography

Where the workforce is vs. where the disease is

Four lenses on the same 88 SA4s. The same region appears in every panel - hover any panel to inspect it across all four at once. The maldistribution between workforce (cluster in MM1) and incidence (rural sun-exposed) is the equity problem ATOF's adaptive threshold is designed to reconcile.

Capacity at threshold
3.0% · combined
≤60%60–80%80–100%100–120%>120%
Workforce density
Derm + GP-skin per 100k pop
0<44–88–12≥12
Incidence
Melanoma cases per 100k / yr
<3030–6060–100100–150150–200≥200
Remoteness (MM1–MM7)
Modified Monash Model
MM1MM2MM3MM4MM5MM6MM7
Component 4 · Equity reveal

Who bears the burden of a uniform threshold?

A single national threshold is mathematically simple but pretends that workforce is evenly distributed. ATOF makes the trade-off explicit: every region's min feasible threshold, stratified by SEIFA disadvantage or ARIA+ remoteness.

What this means for equity
Under the current scenario, the median minimum feasible threshold for SEIFA Q1 (most disadvantaged) regions is 2.8%, compared with 1.9% for Q5 (most advantaged) - a 1.5× disparity. At the current uniform threshold of 3.0%, 847k Q1 residents live in regions whose workforce cannot deliver screening, versus 199k in Q5. A uniform threshold systematically transfers the under-supply burden to disadvantaged communities.
SEIFA quintileSA4sPopulationMedian minMean minPop in over-capacity SA4s
Q1 (most disadvantaged)101.38M2.8%3.3%847k(61%)
Q2286.12M3.1%3.5%2967k(49%)
Q3268.85M2.7%3.5%2581k(29%)
Q4103.91M2.6%4.2%1306k(33%)
Q5 (most advantaged)144.22M1.9%2.2%199k(5%)

Distribution of minimum thresholds within each stratum

one dot per SA4
1%2%3%5%7%10%15%Q1 (most disadvantaged)Far West and Orana (NSW) · 2.9%Mid North Coast (NSW) · 4.0%New England and North West (NSW) · 3.9%Queensland - Outback (QLD) · 2.6%Wide Bay (QLD) · 9.8%South Australia - Outback (SA) · 1.8%Western Australia - Outback (North) (WA) · 1.8%Western Australia - Outback (South) (WA) · 2.0%West and North West (TAS) · 3.0%Northern Territory - Outback (NT) · 1.5%Q2Central West (NSW) · 3.4%Coffs Harbour - Grafton (NSW) · 3.5%Murray (NSW) · 3.2%Richmond - Tweed (NSW) · 4.0%Riverina (NSW) · 3.1%Sydney - Blacktown (NSW) · 1.8%Sydney - Outer South West (NSW) · 2.2%Sydney - South West (NSW) · 1.9%Bendigo (VIC) · 2.8%Hume (VIC) · 2.6%Latrobe - Gippsland (VIC) · 2.9%Melbourne - North West (VIC) · 1.9%North West (VIC) · 2.9%Shepparton (VIC) · 2.6%Warrnambool and South West (VIC) · 3.0%Cairns (QLD) · 4.5%Darling Downs - Maranoa (QLD) · 7.9%Central Queensland (QLD) · 6.7%Ipswich (QLD) · 5.3%Logan - Beaudesert (QLD) · 6.0%Townsville (QLD) · 5.9%Adelaide - North (SA) · 2.2%Barossa - Yorke - Mid North (SA) · 3.1%South Australia - South East (SA) · 3.0%Mandurah (WA) · 2.9%Western Australia - Wheat Belt (WA) · 3.3%Launceston and North East (TAS) · 3.0%South East (TAS) · 3.7%Q3Capital Region (NSW) · 3.1%Central Coast (NSW) · 3.2%Hunter Valley exc Newcastle (NSW) · 3.5%Illawarra (NSW) · 2.9%Newcastle and Lake Macquarie (NSW) · 2.9%Southern Highlands and Shoalhaven (NSW) · 3.5%Sydney - Inner South West (NSW) · 2.0%Sydney - Outer West and Blue Mountains (NSW) · 2.7%Sydney - Parramatta (NSW) · 1.6%Ballarat (VIC) · 2.5%Geelong (VIC) · 2.4%Melbourne - North East (VIC) · 2.0%Melbourne - South East (VIC) · 2.1%Melbourne - West (VIC) · 2.0%Gold Coast (QLD) · 6.3%Mackay - Isaac - Whitsunday (QLD) · 6.8%Moreton Bay - North (QLD) · 10.7%Moreton Bay - South (QLD) · 6.5%Toowoomba (QLD) · 7.3%Adelaide - West (SA) · 2.3%Bunbury (WA) · 2.9%Perth - North East (WA) · 2.5%Perth - South East (WA) · 2.5%Perth - South West (WA) · 2.7%Hobart (TAS) · 2.6%Darwin (NT) · 1.7%Q4Sydney - Sutherland (NSW) · 2.6%Melbourne - Inner South (VIC) · 1.7%Melbourne - Outer East (VIC) · 2.2%Mornington Peninsula (VIC) · 2.7%Brisbane - East (QLD) · 10.2%Brisbane - North (QLD) · 6.6%Brisbane - South (QLD) · 4.6%Sunshine Coast (QLD) · 6.7%Adelaide - South (SA) · 2.3%Perth - North West (WA) · 2.5%Q5 (most advantaged)Sydney - Baulkham Hills and Hawkesbury (NSW) · 2.4%Sydney - City and Inner South (NSW) · 1.6%Sydney - Eastern Suburbs (NSW) · 1.8%Sydney - Inner West (NSW) · 2.0%Sydney - North Sydney and Hornsby (NSW) · 2.0%Sydney - Northern Beaches (NSW) · 2.3%Sydney - Ryde (NSW) · 1.8%Melbourne - Inner (VIC) · 1.4%Melbourne - Inner East (VIC) · 1.8%Brisbane - West (QLD) · 6.5%Brisbane Inner City (QLD) · 2.9%Adelaide - Central and Hills (SA) · 1.6%Perth - Inner (WA) · 0.8%Australian Capital Territory (ACT) · 2.2%Median: 2.8%Median: 3.1%Median: 2.7%Median: 2.6%Median: 1.9%national 3.0%Minimum feasible threshold (10-yr melanoma risk)
SEIFA quintile is the ABS Index of Relative Socio-Economic Disadvantage applied to each SA4 (1 = most disadvantaged, 5 = most advantaged). ARIA+ remoteness uses the Modified Monash Model (MM1 metropolitan → MM7 very remote). Min feasible threshold reflects the current workforce scenario setting at the top of the page.