Objective: To estimate the prevalence and demographic disparities in limited numeracy among emergency department (ED) patients.
Methods: We performed two cross-sectional studies of ED patients with sub-critical illness in 2000-2001 and 2006. We enrolled 959 adult patients from 28 EDs in 17 US states and measured numeracy based on four validated questions.
Results: Rates of correct responses for individual numeracy questions ranged from 15% to 68%; only 11% of participants answered all questions correctly. Several demographic characteristics were independently associated with frequency of correct answers, including age (OR 0.92 [95% confidence interval (CI), 0.87-0.97] per (upward arrow) 5 years), race/ethnicity (compared to whites: OR 0.35 for blacks [95%CI, 0.20-0.63]; and OR 0.36 for Hispanics [95%CI, 0.19-0.69]), education (OR 4.74 [95%CI, 2.01-11.14] for high school graduates vs. not), health insurance (OR 1.70 [95%CI, 1.06-2.71] for those with private insurance vs. not), and income (OR 1.13 [95%CI, 1.05-1.22] per (upward arrow) $10,000).
Conclusion: We found a higher prevalence of limited numeracy among ED patients compared to the general population. Significant demographic disparities are consistent with previous observations for general health literacy.
Practice implications: Greater understanding of the high prevalence of limited numeracy may guide healthcare providers to simplify messages and communicate health information more effectively.