Abstract
The increasing trend of exposing patients seeking health advice to numerical information has the potential to adversely impact patient–provider relationships especially among individuals with low literacy and numeracy skills. We used the HINTS 2007 to provide the first large scale study linking statistical confidence (as a marker of subjective numeracy) to demographic variables and a health-related outcome (in this case the quality of patient–provider interactions). A cohort of 7,674 individuals answered sociodemographic questions, a question on how confident they were in understanding medical statistics, a question on preferences for words or numbers in risk communication, and a measure of patient–provider interaction quality. Over thirty-seven percent (37.4%) of individuals lacked confidence in their ability to understand medical statistics. This was particularly prevalent among the elderly, low income, low education, and non-White ethnic minority groups. Individuals who lacked statistical confidence demonstrated clear preferences for having risk-based information presented with words rather than numbers and were 67% more likely to experience a poor patient–provider interaction, after controlling for gender, ethnicity, insurance status, the presence of a regular health care professional, and the language of the telephone interview. We will discuss the implications of our findings for health care professionals.
Notes
1N = 3806; 2N = 2218.