Generalizability of Blood Pressure Lowering Trials to Older Patients: Cross-Sectional Analysis

Background/objectives: Randomized controlled trials are used to inform clinical guidelines on the management of hypertension in older adults, but it is unclear to what extent these trials represent the general population attending routine clinical practice. This study aimed to define the proportion and characteristics of patients eligible for hypertension trials conducted in older people.

Design: Cross-sectional study.

Setting: A total of 24 general practices in England.

Participants: Anonymized electronic health record data from all individuals aged 80 and older.

Measurements: Descriptive statistics were used to define the proportion and characteristics of patients eligible for two previous medication intensification trials (HYVET, SPRINT) and one medication reduction trial (OPTiMISE). A logistic regression model was constructed to estimate predictors of eligibility for each trial.

Results: Of 15,376 patients identified, 268 (1.7%; 95% confidence interval [CI] = 1.5-2.0%), 5,290 (34.4%; 95%CI = 33.7-35.2%), and 3,940 (25.6%; 95%CI = 24.9-26.3%) were eligible for the HYVET, SPRINT, and OPTiMISE trials, respectively. Between 5.6% and 30.7% of exclusions from each trial were due to eligibility criteria excluding those with high or uncontrolled blood pressure. Frailty (odds ratio [OR] = .44; 95%CI = .36-.54 [OPTiMISE]), cardiovascular polypharmacy (OR = .61; 95%CI = .55-.68 [SPRINT]) and multimorbidity (OR = .72; 95%CI = .64-.82 [SPRINT]) were associated with a lower likelihood of being eligible for one or more of the trials.

Conclusion: A possible unintended consequence of blood pressure criteria used by trials attempting to answer different primary questions is that for many older patients, no trial evidence exists to inform treatment decisions in routine practice. Caution should be exercised when applying results from existing trials to patients with frailty or multimorbidity.