The authors conducted a retrospective chart review of Hispanic and non-Hispanic white patients treated for isolate long-bone fracture at the UCLA Emergency Medical Center from 1990 to 1991. The authors hypothesized that non-Hispanic whites would be more likely to receive pain medication than Hispanics.

Non-Hispanic whites were more than twice as likely as Hispanics to receive an analgesic (55% versus 23%). Controlling for various factors, including physician characteristics, did not reduce this association. In analyses which simultaneously controlled for sex, language, insurance status, severity, hospital characteristics, and potential for association with ethanol intoxication, Hispanic ethnicity was the strongest predictor of receiving no analgesic (OR = 7.46, 95% confidence interval = 2.22 to 25.04). Hispanic patients were equally likely to receive a high dose of analgesic as non-Hispanic whites but less likely to receive a low dose (19.4% of Hispanics versus 45.4% of non-Hispanic whites received a low dose). Hispanics were slightly less likely to receive a parenteral analgesic but much less likely to receive an oral analgesic (6.5% of Hispanics versus 30.6% of non-Hispanic whites).

While this study is suggestive, it did not control for all relevant factors, such as clinical factors (e.g., medication/alcohol use), severity of pain as rated by the patient, and a specific request for analgesic by the patient. (However, the authors clearly state their doubt that these findings are due to Hispanic patients feeling less pain or refusing analgesics.) Nor does it test for potential mediating factors, such as patient advocates (e.g., the presence of family and friends). However, it is important that large differences between ethnic groups were observed in a small sample (108 non-Hispanic whites and only 31 Hispanics), and after controlling for a range of potentially confounding factors. The authors conclude these findings, if replicated, are "of concern and require further elucidation."