Examining the Effects of Contextually-Imposed Cognitive Load on Providers' Chronic Pain Treatment Decisions for Racially and Socioeconomically Diverse Patients

Abstract

Compared to people who are White and have high socioeconomic status (SES), those who are Black and have low SES are more likely to receive suboptimal pain care. One potential contributor to these disparities is biased provider decision-making—there is compelling evidence that providers are influenced by patient race and SES when making pain treatment decisions. According to the dual process model, people are more likely to be influenced by demographic stereotypes, including implicit beliefs, when they are under high cognitive load (i.e., mental workload). One stereotype belief relevant to pain care is that Black and low SES people are more pain tolerant. Aligned with the dual process model, providers who are under high cognitive load and have strong implicit beliefs that Black and low SES people are more pain tolerant may be particularly likely to recommend fewer pain treatments to them. To test this hypothesis, I recruited physician residents and fellows (n=120) to make pain treatment decisions for 12 computer-simulated patients with back pain that varied by race (Black/White) and SES (low/high). Half of the providers were randomized to the high cognitive load group in which they were interrupted during the decision task to make conversions involving hypertension medications for another patient. Remaining providers completed the task without being interrupted. Providers’ implicit beliefs about race and SES differences in pain tolerance were measured with two separate Implicit Association Tests (IATs). Multilevel modeling indicated that providers recommended stronger medications to low than high SES patients (OR=.68, p=.03). There was also a significant interaction between patient SES and cognitive load (OR=-.56, p=.05) and a trending interaction between patient race and cognitive load (OR=1.7, p=.07). Under low cognitive load, providers recommended more pain treatments to high SES (vs. low SES) and Black (vs. White) patients, but under high cognitive load, providers recommended more pain treatments to low SES (vs. high SES) patients and equivalent treatment to Black and White patients. There were no three-way interactions between patient demographics (race or SES), cognitive load, and providers’ implicit beliefs (race-pain or SES-pain IAT scores). However, there was a trending interaction between patient race and race-pain IAT scores (OR=2.56, p=.09). Providers with stronger implicit beliefs that White people are pain sensitive and Black people are pain tolerant recommended more pain treatments to White patients and fewer pain treatments to Black patients. Lastly, there was a trending effect that providers with stronger implicit beliefs that high SES people are pain sensitive and low SES people are pain tolerant recommended stronger medications in general (OR=13.03, p=.07). Results support that provider cognitive load is clinically relevant and impacts clinical decision-making for chronic pain for racially and socioeconomically diverse patients. Future studies are needed to further understand the impact of cognitive load on providers’ pain care decisions, which may inform evidence-based interventions to improve pain care and reduce disparities.