Abstract / Summary
Importance Diagnostic accuracy is a key driver of emergency department (ED) quality, yet little consensus exists on which conditions to target. Objective To develop a list of emergency diagnosis–sensitive conditions (EDSCs) for measuring diagnostic error in the ED. Design, Setting, and Participants This retrospective cross-sectional study examined a national sample of claims for fee-for-service Medicare beneficiaries 65 years or older with nonelective hospitalizations from January 1, 2022, to December 31, 2024. Exposure Potential diagnostic errors in the ED. Main Outcomes and Measures A panel of 5 clinician investigators reviewed principal discharge International Statistical Classification of Diseases and Related Health Problems, Tenth Revision diagnosis codes and determined whether each met criteria for an EDSC—a disease that is emergency care sensitive, at risk of diagnostic error, and should be diagnosed in the ED or on admission. The percentage of each EDSC’s admissions preceded within 9 days by a potential diagnostic error (ED discharge) was characterized. To account for background ED utilization, an adjusted potential diagnostic error percentage was calculated by subtracting the percentage of a comparison cohort with at least 1 ED visit in the 9 days preceding a randomly selected date. The incidence of potential diagnostic errors per 10 000 ED discharges was calculated. Thirty-day morbidity (number of healthy days at home [HDAH]) and mortality were characterized. High-priority EDSCs were those in the top half of diagnostic errors per 10 000 ED discharges and in the top quartile of either mortality or HDAH lost. For high-priority conditions, the 20 most frequent diagnoses on preceding ED discharges were tabulated. Results Of 19 973 diagnosis codes identified, 1802 were aggregated into 76 EDSC categories. Of 14 671 540 admissions, 2 852 901 (19.4%) were for EDSCs. Of these EDSC admissions, 713 601 (25.0%) occurred among adults 85 years or older, and 1 493 822 (52.4%) were among women. The EDSCs with the highest and lowest adjusted percentages of admissions preceded by a potential ED diagnostic error were Guillain-Barré syndrome (16.6%; 95% CI, 15.4%-17.9%) and acute appendicitis (1.4%; 95% CI, 1.3%-1.5%), respectively. Pneumonia had the highest adjusted incidence of potential diagnostic errors per 10 000 ED discharges (9.30; 95% CI, 9.18-9.42) despite a moderate adjusted potential diagnostic error percentage per 10 000 ED discharges (4.4%; 95% CI, 4.3%-4.5%). Symptom-based diagnoses were frequent on ED discharges preceding EDSC admissions. Conclusions and Relevance In this retrospective observational study of Medicare beneficiaries, EDSCs were used to characterize the population burden of potential diagnostic errors in emergency care and associated clinical presentations.