Classifying Race in Out-of-Hospital Cardiac Arrest and Potential Disparities: A Retrospective Cohort Study.

Abstract

Background: Although racial disparities have been described in resuscitation, little is known about potential bias in race classification of out-of-hospital cardiac arrest (OHCA).

Methods: We conducted a retrospective cohort study of adults treated by emergency medical services (EMS) for nontraumatic OHCA in King County, WA between January 1, 2018, and December 31, 2021. We assessed agreement using κ and evaluated patterns of missingness between EMS-assessed race versus comprehensive race classification from hospital and death records. Using multivariable logistic regression adjusting for Utstein data elements, we analyzed the association between race and OHCA survival across different sources.

Results: Among 5909 eligible OHCA patients, the average age was 64.0 years, 35.4% were female, and 16.1% survived to hospital discharge. Based on comprehensive race classification, 68.7% were White, 12.8% Black, 12.1% Asian, 2.5% multiracial, 2.3% Native Hawaiian/other Pacific Islander, and 1.6% American Indian/Alaska Native. EMS did not classify race in 43.7%. The κ coefficient between EMS and comprehensive race classification was 0.88 (95% CI, 0.86-0.90), though agreement varied substantially by specific race and was lowest among American Indian/Alaska Native (39.5%). Missingness in EMS records varied according to race and was greater among those classified as American Indian/Alaska Native (60.8%), Native Hawaiian/other Pacific Islander (58.8%), Asian (57.8%), or multiracial (54.1%) compared with White (40.6%) or Black (40.4%). In multivariable models using EMS-classified race, the odds ratio (OR) of survival was not significantly different for any race group compared with the White race, that is, OR. However, when using comprehensive race classification, OR of survival was significantly lower among Native Hawaiian/other Pacific Islander (OR, 0.57 [95% CI, 0.33-0.97]) and among multiracial (OR, 0.40 [95% CI, 0.20-0.75]) compared with White race.

Conclusions: In adult OHCA, race misclassification and missingness influenced its association with survival. Efforts should continue to evaluate best practices to classify race correctly and comprehensively.