Cross-sectional study of healthcare professionals' estimates of the health numeracy of inpatients and outpatients in a UK secondary care setting.

Introduction: A cross-sectional study to identify whether healthcare professionals (HCPs) can accurately estimate the health numeracy of patients.

Methods: A convenient sample of inpatients and outpatients, in an urban UK general hospital, undertook a validated health numeracy assessment and associated demographics questionnaire. HCPs who had a care interaction with the patient were shown the health numeracy assessment, informed of the mean score and SD in the assessment's validation study, and were then asked to estimate their patient's score. Outcome measures were the proportion of underestimations, correct estimations and overestimations by HCPs and a comparison of HCP estimates to the patient's score on the assessment as assessed through the intraclass correlation coefficient (ICC).

Results: Health numeracy assessments were completed by 142 patients with a mean score of 38.9% and an SD of 33.4%. There were 220 estimations of patients' health numeracy obtained from HCPs. All HCP groups overestimated patient health numeracy with overestimates accounting for 66.8% of all estimates. ICC was below 0.4 for all HCP groups (ICC 0.054; 95% CI -0.078 to 0.185) indicating poor agreement between the HCPs' estimations and the patient's health numeracy as measured by the health numeracy assessment. Senior doctors (consultants and registrars) were most likely to correctly estimate patient health numeracy (20.8% and 20.0% of estimates, respectively).

Conclusions: Good health numeracy is vital to effective understanding of risk, shared decision-making and the consenting process. However, HCPs of varying professional backgrounds struggle to correctly estimate their patient's health numeracy and tend to overestimate it. Given that health numeracy is poor for a large proportion of patients, there is a risk that HCPs may fail to identify scenarios in which their patient's poor health numeracy could undermine shared decision-making and/or lead to poor outcomes.