Asbestos Burden and Evidence Atlas

What Is Official vs. Inferred

The intended official inputs for this atlas are IHME burden estimates, IHME source metadata, and IHME uncertainty outputs. By contrast, the coverage score, blind-spot score, and uncovered population share shown in this prototype are inferred constructs created for this website.

• Official or intended official inputs: GBD burden estimates, source metadata, uncertainty intervals.
• Inferred prototype inputs: coverage score, blind-spot score, uncovered population share, and weighting scheme.
• These inferred metrics are visual aids for exploration, not formal IHME indicators.

Why These Weights Were Used

The weights 0.35 / 0.25 / 0.20 / 0.10 / 0.10 were judgment-based design choices. They were chosen to emphasize disease burden first, then population potentially outside observed coverage, then uncertainty, with smaller penalties for older and sparser source environments.

These weights were not copied from IHME and do not represent an endorsed methodology from IHME, WHO, OECD, or any other institution. They were informed by general guidance on composite indicators and health data quality, especially the principle that weighting should follow an explicit conceptual framework and that weights often reflect analytic judgment rather than objective truth.

Assumptions

• Higher estimated burden should increase priority even when evidence quality varies.
• Larger uncovered population share may indicate greater risk of under-observation.
• Greater uncertainty can act as a caution signal for weaker or less stable evidence environments.
• Older source inputs reduce confidence in how current the evidence picture is.
• Fewer available sources may indicate thinner evidence support, while not measuring quality directly.
• Combining these dimensions into one score is useful for communication, even though it compresses complexity.

Possible Limitations

• The blind-spot score is not an official IHME indicator.
• The weighting scheme is heuristic and judgment-based.
• Different weighting choices could materially change the rankings.
• Uncertainty is an imperfect proxy for evidence weakness.
• Source count does not equal source quality.
• Population outside reported coverage is an inferred approximation, not a universal observed statistic.
• Country-level averages may hide important subnational asbestos hotspots.
• This visualization should identify questions and priorities, not make definitive claims about surveillance adequacy.

Why These Simplifications Are Used

These simplifications are acceptable for an exploratory dashboard because the aim is to help users detect patterns, compare places, and identify where deeper evidence review may be warranted. They are not a substitute for formal methodological reporting, and this site should always label the evidence-gap layer as experimental, inferred, or prototype.

References

• OECD/JRC Handbook on Constructing Composite Indicators: supports the view that weights in composite indicators follow a conceptual framework and are ultimately value judgments.
• WHO Data Quality Review Toolkit, Module 1: supports data-quality dimensions such as completeness, timeliness, consistency, and denominator comparison.
• IHME GBD 2021 Data Tools Guide: explains how to interpret GBD outputs and why larger uncertainty can reflect thinner or less consistent evidence.
• IHME GBD 2021 disease and injury analysis summary: supports uncertainty-aware interpretation of burden estimates.
• IHME GBD 2021 causes of death analysis summary: supports the role of propagated uncertainty in cause-specific estimates.

The choice of weights in this prototype is an inference informed by these general methodological references. It is not a published rule from any of the sources above.