Prior research suggests that stereotypes and bias may affect mental health diagnoses (Abreu, 1999; Dovidio, Evans, & Tyler, 1986; Loring & Powell, 1988; Whaley, 1997). Across mental health care settings including Psychiatric Emergency Settings (PES), African-American patients are more likely to be diagnosed with schizophrenia and less likely to be diagnosed with affective disorders as compared to White patients (Starkowski, Lonczak, Sax, West, Mehta, and Theinhaus, 1995). Additionally, women receive higher rates of affective disorders (Starkowski et al., 1995), while men are more commonly diagnosed with schizophrenia. These clinical findings contrast the epidemiological literature which has not found evidence of racial or gender differences in prevalence rates of schizophrenia (Robins & Regier, 1991; Starkowski et al., 1995) nor bipolar disorder in community samples (Bland, 1997; Lohr & Cook, 2003; NIMH, 2000; Soreff, 2004; Weissman et al., 1991). Contrasts between clinical and epidemiological findings raise the possibility that patients are being misdiagnosed because of racial and gender stereotypes, which may have devastating clinical and social consequences. According to social-psychological and cognitive theories, patient characteristics may influence clinicians' tendencies to stereotype, especially under conditions of limited time and high patient pressures. Cognitive load refers to the magnitude of demands imposed within specific time constraints and limitations that restrict the amount of available cognitive resources that can be dedicated to a particular task (Gerson & Bassuk, 1980). This study investigated the effects of patient race and gender on PES diagnostic decisions of bipolar disorder made under high and low levels of cognitive load.

Patterns of decisionmaking were examined through retrospective record reviews of 1236 psychiatric patients, treated by 75 clinicians in a PES in a Mid-western city, and interviews with a sub-sample of clinicians. Patient records were randomly sampled according to the clinician's level of cognitive load, controlling for the average number of patients typically seen and the actual volume of patients seen by the particular clinician during that shift.

Multinomial logistic-regression analyses replicated previous clinical diagnostic trends with Black (RR=6.44,p=.000) and male (RR=2.130,p=.026) patients receiving higher rates of Psychotic Disorder compared to Bipolar Disorder. White (RR=3.479,p=.002) and female patients receiving higher rates of Bipolar Disorder. Results suggest that cognitive load may affect PES decisionmaking, especially the assignment of bipolar disorder for female patients specifically (p<.02). The number of Black patients diagnosed with Bipolar Disorder was sparse in both cognitive load conditions (N=10, 4.7%; N(low)=6; N(high)=4).

Although epidemiological studies do not reveal race and gender differences in rates of bipolar disorder in the general population, divergent rates do exist in clinical samples. Results highlight the potential difficulty of differentiating bipolar disorder from psychotic disorders (as well as other affective disorders) especially when evaluating patients with particular sociodemographic characteristics and/or under high load. Contextual factors, like cognitive load, may help explain gender and racial disparities in mental health. Findings may also have important practice implications for the use of psychotropic medications. Further examination of these factors is critical to the delivery of quality mental health care across all subgroups of patient populations.