Defining the "Critical Casualty" through Reference-Standard Constructs, Reporting Completeness, and Reproducibility in Mass-Casualty Incident Triage Research: Methodology Scoping Review.

INTRODUCTION/STUDY

OBJECTIVE

Mass-casualty incident (MCI) triage algorithms are evaluated against reference standards that operationally define the "critical casualty," yet the constructs underlying these standards and the completeness of their reporting have not been systematically characterized.

METHODS

A PRISMA-ScR methodology scoping review was conducted using a pre-registered Open Science Framework (OSF) protocol; MEDLINE, Scopus, Web of Science, and CINAHL were searched on May 5, 2026 for English-language full-text sources that described, derived, validated, or applied a reference standard for distinguishing critical from non-critical casualties in MCI triage. Reference standards were classified using a pre-specified taxonomy, and reporting completeness was assessed with the eight-item Reference Standard Reporting Index for MCI Triage (RSRI-MCI), developed a priori for this review and not externally validated. Two reviewers independently classified and coded all sources.

RESULTS

Sixty sources published from 2001 through 2025 were included. Operationalized reference standards clustered around four foundational methodological sources, while only five sources (8%) analyzed retrospective data from actual MCIs. Five reference-standard families were identified: Life-Saving Intervention (LSI)-Based (38%), Composite/Hybrid (30%), Expert Panel (15%), Outcome/Resource (12%), and Anatomic Severity (5%); LSI-Based and Composite constructs together accounted for approximately two-thirds of the field. Reporting was asymmetric: definition (98%), components (97%), and rationale (88%) were reported almost universally, whereas time window (40%), dead/expectant handling (57%), binary mapping (65%), missing-data handling (58%), and reference-standard limitations (58%) were reported less consistently. Inter-rater agreement was substantial overall (83%) and for family classification (93%; Gwet's Chance-Corrected Agreement Coefficient [AC1] = 0.919; 95% CI, 0.837-0.980).

CONCLUSION

Mass-casualty incident triage studies routinely report what their reference standard is but inconsistently report how it is operationalized, particularly its time window. Five reporting recommendations are proposed for future MCI triage validation research.