Implicit bias is an unconscious preference for a specific social group that can have adverse consequences for patient care. Acute care clinical vignettes were used to examine whether implicit race or class biases among registered nurses (RNs) impacted patient-management decisions.

In a prospective study conducted among surgical RNs at the Johns Hopkins Hospital, participants were presented 8 multi-stage clinical vignettes in which patients' race or social class were randomly altered. Registered nurses were administered implicit association tests (IATs) for social class and race. Ordered logistic regression was then used to examine associations among treatment differences, race, or social class, and RN's IAT scores. Spearman's rank coefficients comparing RN's implicit (IAT) and explicit (stated) preferences were also investigated.

Two hundred and forty-five RNs participated. The majority were female (n = 217 [88.5%]) and white (n = 203 [82.9%]). Most reported that they had no explicit race or class preferences (n = 174 [71.0%] and n = 108 [44.1%], respectively). However, only 36 nurses (14.7%) demonstrated no implicit race preference as measured by race IAT, and only 16 nurses (6.53%) displayed no implicit class preference on the class IAT. Implicit association tests scores did not statistically correlate with vignette-based clinical decision making. Spearman's rank coefficients comparing implicit (IAT) and explicit preferences also demonstrated no statistically significant correlation (r = −0.06; p = 0.340 and r = −0.06; p = 0.342, respectively).

The majority of RNs displayed implicit preferences toward white race and upper social class patients on IAT assessment. However, unlike published data on physicians, implicit biases among RNs did not correlate with clinical decision making.