Association of Questionnaire-Assessed Fall Risk With Uncontrolled Blood Pressure and Therapeutic Inertia Among Older Adults

Abstract

Therapeutic inertia (TI), or failure to escalate or initiate BP lowering medications when BP is uncontrolled, increases with advancing age and may in part be due to perceived fall risk. This study examined the association of a fall risk assessment, based on patient response to three questions administered by trained staff, with uncontrolled BP (≥140/90 mmHg) during a clinic visit and with TI during clinic visits with uncontrolled BP among 13 893 patients age ≥ 65 years corresponding to 41 122 primary care visits. Separate generalized linear mixed effects models were used to examine the association of fall risk (low, moderate, and high) with uncontrolled BP and with TI at a clinic visit after adjustment for demographics, comorbidities, and total number of visits. Baseline mean age was 73.0 years (standard deviation [SD] 5.6), 43.3% were men and questionnaire-assessed fall risk severity was low in 73.6%, moderate in 14.3%, and high in 12.2%. Compared to low fall risk, the adjusted odds of uncontrolled BP during a clinic visit were 0.97 (95% CI: 0.89, 1.06) and 0.90 (95% CI: 0.82, 0.98) with moderate and high fall risk, respectively. In contrast, adjusted odds of TI during a clinic visit with BP ≥ 140/90 mmHg was 1.16 (95% CI: 1.01, 1.34) and 1.30 (95% CI: 1.11, 1.52) with moderate and high fall risk, respectively, compared to low fall risk. These findings suggest that perceived fall risk severity may be one of several factors that influence hypertension management in older adults.

Keywords: aging; blood pressure; falls; hypertension; inertia.

FIGURE 1

Flowchart for selection of patient data.

FIGURE 2

Adjusted odds ratios for therapeutic inertia by questionnaire‐assessed fall risk categories. Adjusted model controls for age group (65–74 and ≥75 years), sex, race/ethnicity, clinic site, body mass index, current smoking status, cardiovascular disease, diabetes, number of visits, number of 30‐day intervals between visits, systolic blood pressure at the clinic visit and patient random effects. Models include fall assessments concurrent with or preceding the clinic visit by up to 12 months.