Objective: To determine whether Hispanic patients with isolated long-bone fractures are less likely to receive emergency department (ED) analgesics than similar non-Hispanic white patients.
Design: Retrospective cohort study.
Setting: The UCLA Emergency Medicine Center, a level I trauma center.
Participants: All Hispanic and non-Hispanic white ED patients aged 15 to 55 years, seen between January 1, 1990, and December 31, 1991, with isolated long-bone fractures, identified by ICD-9 codes 812, 813, 821, and 823, were eligible for inclusion. Exclusion criteria included injury more than 6 hours prior to presentation, "possible" or chip fractures only, altered mentation, or ethanol intoxication.
Main outcome measures: Emergency department administration of analgesic or no analgesic.
Results: The study group consisted of 139 patients meeting inclusion criteria, of whom 31 were Hispanic and 108 non-Hispanic white. Non-Hispanic whites were significantly more likely to speak English, be insured, and suffer nonoccupational injuries. Hispanics were twice as likely as non-Hispanic whites to receive no ED pain medication (crude relative risk [RR], 2.12; 95% confidence interval [CI], 1.35 to 3.32; P = .003). The RR for ethnicity was similar and significant (P < .05) after controlling by stratification for covariates related to patient, injury, or physician characteristics. After controlling for several covariates simultaneously through multiple logistic regression, ethnicity remained the strongest predictor of ED analgesic administration (odds ratio [OR], 7.46; 95% CI, 2.22 to 25.04; P < .01).
Conclusions: Hispanics with isolated long-bone fractures are twice as likely as non-Hispanic whites to receive no pain medication in the UCLA Emergency Medicine Center. No covariate measured in this study could account for this effect. An ethnic basis for variability in analgesic practice needs to be further characterized.