Background: Pain is ubiquitous; the 2019 National Health Survey indicated that 20.4% of adults had chronic pain and that for 7.4% of adults, pain limited work or life activities. Earlier research found racial disparities in the management of pain and opioid prescriptions in emergency department and surgical settings, but research has been more limited in the inpatient setting. In addition, there are various tools used to assess and document a patient’s level of pain in the electronic health record (EHR). This quantitative study explores whether there are disparities in the assessment of pain based on race/ethnicity, gender, and limited English proficiency (LEP) status.
Methods: This retrospective cohort study identified all adults discharged from the general medicine service at an 800-bed urban academic teaching hospital from January 2013 through September 2021. Using data collected from the EHR, we extracted nursing pain assessments to explore differences in pain assessment type and the average number of pain assessments across race/ethnicity, gender, and LEP status. Nurses have the option of using various pain assessment tools, including numeric pain score, Wong-Baker FACES Pain Rating Scale, verbal descriptors, Critical Care Pain Observation Tool (CPOT), Checklist of Nonverbal Pain Indicators (CNPI), and Face Legs Activity Cry Consolabilty (FLACC) pain scale. Our institutional gold standard is the numeric pain score (0-10 scale). Additionally, we calculated the average number of documented pain assessments per day for patients. Patients’ self-reported race/ethnicity and gender were used. LEP status was determined by primary language and whether the patient required interpretative services.
Results: We identified 60,900 patients who received pain assessments while on the hospital medicine service from January 2013 through September 2021. There were 3,035,677 individual pain assessments completed on these patients. We found statistically significant differences in pain assessment type for patients with LEP status (p-value < 0.001). Patients with LEP status were less likely to receive our institution’s gold standard of numeric pain score than patients proficient in English (34.16% vs. 58.83%). LEP patients were also more likely to receive nonverbal pain assessments than their English proficient counterparts (14.10% vs. 5.73%). Differences also existed in the average number of pain assessments performed. For instance, Black patients and patients identified as “Other”, on average, had the lowest number of pain assessments of any race/ethnicity.
Conclusions: There were significant differences in pain assessment type for patients with LEP status as compared to English-speaking patients. While Black patients, on average, had among the fewest pain assessments, a higher percentage of their pain assessment types were numeric pain score compared to other racial/ethnic groups. Given that numeric pain score is our gold standard for pain assessment, its relative underuse for patients with limited English proficiency merits future investigation. Further research needs to investigate the factors that contributed to these disparities and if frequency and type of pain assessments performed impacted patient care, including pain medication and administration, and other clinical outcomes.
