How Common Disease Is Missed or Misattributed in Exposure-Defined Populations: A Framework-Based Synthesis

Technical Summary: How Common Disease Is Missed or Misattributed in Exposure-Defined Populations

Problem Statement
Common diseases such as chronic kidney disease (CKD), asthma, chronic obstructive pulmonary disease (COPD), and pulmonary fibrosis are frequently missed or assigned incorrect or incomplete causes in populations defined by significant work or environmental exposures (e.g., chemicals, heat, dust). While these conditions are familiar in general clinical practice, the systems used to detect, classify, and count them are often organized around risk profiles typical of lower-exposure populations. Consequently, individuals in high-exposure settings may develop these common diseases through different drivers (e.g., heat stress and agrochemicals rather than diabetes and hypertension) and fall outside standard screening pathways. This leads to two primary failures: undercounting (fewer cases entering record systems than actually exist) and misattribution (records containing incorrect or incomplete causes). Current literature often treats these issues separately or focuses on rare, exposure-specific diseases, leaving a gap in understanding how common multifactorial diseases are lost or misclassified across the entire spectrum of clinical and administrative systems.

Methodology
This study employed a framework synthesis approach, adapted from framework analysis and informed by care-cascade principles. The review was guided by the Population, Concept, and Context (PCC) approach:

• Population: Individuals identified through work or environmental exposures.
• Concept: Disease undercounting and misattribution.
• Context: Clinical care, surveillance, reporting, compensation, death registration, and burden estimation.

The authors developed a six-stage staged framework to map where disease loss or misrepresentation occurs:

1. Clinical Detection: Failure to detect disease in the population.
2. Disease Classification: Incorrect diagnosis among detected patients.
3. Etiologic Attribution: Absent, incomplete, or incorrect cause assignment among diagnosed patients.
4. Reporting or Compensation: Failure to report diagnosed or eligible cases.
5. Death Registration: Failure to register a death.
6. Cause-of-Death Coding: Assignment of an unusable or incorrect cause to a registered death.

Data collection involved searching Europe PMC, Google Scholar, Semantic Scholar, African Index Medicus, and major burden-estimation reports (e.g., Global Burden of Disease) for literature from January 2000 onward. Sources were categorized as direct evidence (measuring undercounting/attribution in exposure-defined populations) or contextual/boundary evidence (informing patterns without directly measuring the failure). Screening and extraction were performed by a single reviewer; formal risk-of-bias assessment was not conducted due to the heterogeneity of study designs.

Key Results
From approximately 6,020 identified records, 61 sources were charted, contributing 64 findings. The synthesis revealed a significant scarcity of direct patient-level evidence across the framework:

• Direct Patient-Level Evidence: Only one study provided direct evidence of diagnostic reclassification. Morell et al. found that 20 of 46 patients previously diagnosed with idiopathic pulmonary fibrosis were reclassified as having chronic hypersensitivity pneumonitis after systematic exposure investigation (often due to occult avian antigens).
• Population-Level Attribution: Two studies provided population-level estimates suggesting exposure associations within "idiopathic" or "unknown cause" categories (e.g., occupational exposures accounting for 21% of idiopathic pulmonary fibrosis), but these did not verify errors in individual patient records.
• Evidence Gaps: No direct common-disease evidence was identified for clinical detection, reporting/compensation, death registration, or cause-of-death coding specifically within exposure-defined populations.
• Contextual Findings:
  • Undercounting: General population studies show high rates of undiagnosed hypertension and diabetes, but do not isolate exposure-defined subgroups. Active screening for exposure-specific diseases (e.g., silicosis) detects cases missed by passive reporting.
  • Reporting Loss: Occupational disease reporting is known to be incomplete (50–95% underreporting in some studies), and exposure histories are frequently not elicited in routine clinical care.
  • Phenotypic Differences: Exposed populations often present with different risk patterns (e.g., CKD in agricultural workers without diabetes/hypertension; respiratory symptoms in biomass-exposed rural populations with low spirometry-defined COPD).
  • Burden Estimation Limits: Methods like garbage-code redistribution and comparative risk assessment can recover aggregate population burden but cannot reconstruct missing individual records or establish individual causation.

Key Contributions

1. A Staged Framework: The paper introduces a structured, six-stage framework distinguishing analytically distinct failure points in the disease lifecycle (detection, classification, attribution, reporting, registration, coding). This clarifies that different stages require different denominators and have different implications for correction (e.g., a registered death with a wrong cause can be statistically reassigned, whereas an unregistered death leaves no record).
2. Evidence Mapping: The synthesis explicitly maps the availability of evidence, highlighting that while "misattribution" in idiopathic categories is supported by population-level data, direct evidence of individual diagnostic failure in common diseases is extremely limited.
3. Distinction of Problems: The work differentiates between undercounting (missing cases) and misattribution (wrong causes), and between exposure-specific diseases and common diseases that present differently in high-exposure settings.

Significance and Claims
The paper claims that the current evidence base is insufficient to quantify the frequency of disease loss across all six stages. The primary significance of the work is diagnostic and structural rather than prescriptive:

• It identifies where disease may be lost or misattributed and what evidence is needed at each stage to measure these failures.
• It supports the rationale for exposure-informed clinical care, suggesting that incorporating exposure history into assessment (rather than waiting for familiar explanations to fail) may improve detection and attribution. However, the paper explicitly states it does not evaluate the effectiveness, prevalence, or cost of such interventions.
• It calls for future research to test whether structured exposure assessment improves diagnostic accuracy, time to diagnosis, and management outcomes, particularly in diseases currently labeled as idiopathic, sporadic, or of unknown cause.

The authors conclude that while the framework clarifies the mechanisms of failure, direct patient-level evidence remains limited, necessitating further evaluation of exposure-informed approaches to disease detection, classification, and management.