Government brief

ATOF: threshold selection for risk-stratified melanoma screening

Background

The National Targeted Skin Cancer Screening Roadmap sets out the case for a risk-stratified approach: invitations would go to people above a chosen 10-year risk threshold rather than to every age-eligible adult. Choosing that threshold is both a clinical decision (what is the model worth?) and an operational one (can the workforce deliver?).

Choosing a national threshold on clinical grounds alone risks recommending a policy that some Australian regions could not deliver in practice. The regions least able to absorb the implementation burden are often those with the highest disease and lowest workforce density. That is an equity consideration the threshold debate should make explicit.

88
mapped SA4 regions (states, ACT, NT)
ABS SA4; 20 of the 108 codes have no mapped population
22 of 88
SA4s with essentially no dermatologist time
0–3 clinical hours a week, NHWDS 2024
67–76%
of melanoma in situ may be overdiagnosis
Lindsay 2024
1.8×
disparity in minimum feasible threshold
median 1.6% (most disadvantaged SEIFA quintile) vs 0.9% (least)

What ATOF does

ATOF is an interactive decision-analytic tool. It brings three usually-separate evidence streams into one screen, so a policy maker can see the trade-offs before committing to a threshold.

  1. Clinical net benefit. For each validated Australian risk model (QSkin MP16/MP7, MIA/Vuong, NCI MBRAT, Cho), ATOF shows where the model is actually clinically useful, by Vickers and Elkin's decision-curve framework. Before overdiagnosis is counted, every model beats screening everyone and screening no one across 0.5–15% 10-year risk; counting overdiagnosis narrows that to about 0.5–3% (28% of screen-detected melanomas overdiagnosed), or to the lowest thresholds only, around 0.5% (54%).
  2. Overdiagnosis harm. The same curves can be deflated by an evidence-based overdiagnosis rate (Lindsay 2024, Glasziou 2020). The useful range shrinks visibly, and the overdiagnosis penalty becomes an adjustable input rather than a footnote.
  3. Regional workforce capacity. For each of Australia's 88 SA4 regions, ATOF compares the screening demand a threshold would create against the dermatologist and SCCA-accredited GP workforce that can actually deliver it. Red regions cannot deliver; green regions could deliver more.

Five workforce scenarios are available, including two modelled extensions clearly flagged as such: Total Body Photography (decouples image capture from reading) and a tiered GP-triage / specialist model.

How to use the tool

  1. Start with a question. The planner opens with three questions: can my region deliver skin checks, who should do the checks, and what would a national programme cost. Each one takes you to that answer.
  2. Set three controls. The risk threshold, who does the checks and how often. Four numbers and one sentence update at once: adults offered checks, regions that can deliver, net cost a year and the cost of each year of healthy life gained.
  3. Open the detail when it is needed. Where the money goes, benefit and harm, equity and every assumption sit in collapsed sections under the answer. The detailed view holds the full calculators: decision curves, specialist capacity maps, a regional advisor and per-person cost-effectiveness.
  4. Review equity effects. The Equity tab of the detailed view shows the regions stratified by SEIFA disadvantage or remoteness. The tool states the disparity in plain numbers: the median lowest threshold a region's specialist workforce can deliver is 1.6% in the most disadvantaged quintile and 0.9% in the least, and the tool counts the residents of regions that cannot deliver the chosen national threshold.
  5. Export outputs. In the detailed view every chart exports as a PNG with the current configuration recorded in the filename, and the page prints to a paginated PDF with a state-stamped header banner. Every view has a sharable URL.
  6. Embed the tool. The detailed view can be loaded inside an iframe with /explore/?embed=1, ?embed=figure, or ?embed=map for use on a Roadmap page or a supplementary slide.

Benefits to the programme

  1. Explicit trade-offs. The threshold debate is currently a clinical decision with implicit assumptions about workforce. ATOF surfaces those assumptions so they can be challenged, refined, and owned by all parties.
  2. One shared tool for multidisciplinary discussion. Clinicians, epidemiologists, health economists, workforce planners, and equity advocates can use the same tool and see how their concerns interact.
  3. Traceable sources. Every number traces to a citation; every modelled scenario is flagged as modelled; every empirical scenario cites its source. Users can check any value directly.
  4. Quantified equity effects. The tool quantifies, by SEIFA quintile and by remoteness, exactly which Australians a uniform threshold under-serves. Adaptive per-region thresholds become a concrete policy option rather than an abstract aspiration.
  5. Exportable outputs. PNG export, print-to-PDF, share-link, and a standalone methods page mean the same tool that drives the meeting also produces the supplementary material.

What ATOF is not

The scope of the tool is limited in four ways:

  1. Not a microsimulation. ATOF does not track individual patients over time. The microsimulation arm is McLoughlin et al. (PLoS One 2025); the two are complementary.
  2. Not a submission-grade cost-effectiveness analysis. ATOF gives illustrative costs per QALY (Value per person page and Workforce Lab cost panel); the definitive estimate is Policy1-Melanoma's microsimulation. For scale it cites MSAC 1699 (National Lung Cancer Screening Program), a screening precedent, not a melanoma assessment (ICER AUD $62,754/QALY).
  3. Not a clinical decision support tool. ATOF informs programme design at the population level. Individual patient decisions still belong with the clinician.
  4. Not a finished product. ATOF v0.1 is a draft for demonstration. Numbers are directional, not ready-to-cite. See the Limitations page for the complete list of assumptions and the data needed to take it from draft to v1.

Summary

ATOF makes the trade-off between clinical net benefit, overdiagnosis harm, and regional workforce capacity visible in one interactive view, so the national risk threshold can be chosen with its equity consequences in view, rather than on clinical grounds alone and then applied to regions that cannot deliver it.