Community Hypertension Assessment Trial (CHAT)

Hypertension affects about 22% of Canadian adults and is a modifiable risk factor for cardiovascular disease. Most countries have increasing older adult populations and hypertension prevalence increases with age. However, the "rule of halves" still applies in many primary health care settings: half of the hypertensive patients are undiagnosed, half of the diagnosed patients are untreated, and half of the treated patients are uncontrolled.

Family physicians were randomly selected and approached to participate in CHAT. Paired cluster randomization was used, with practices as the unit of randomization. Data on blood pressure (BP) readings in the last year, diagnosis of hypertension, diagnosis of diabetes and/or target organ damage, and a current antihypertensive profile was extracted from the health records of 55 patients randomly selected in each practice. In practices allocated to the Intervention group, physicians sent a letter, with assistance from research staff, to invite all of their eligible patients 65 years and older to attend one or two BP sessions in nearby pharmacies. Inclusion criteria required that patients be regular patients in the practice, community-dwelling, and mobile/well enough to attend community sessions.

Public health authorities trained older adult volunteer peer health educators to assist patients at the sessions to accurately measure their BP using a validated automated device, and record BP readings and cardiovascular risk factors. The volunteers provided education about hypertension but did not perform a clinical function. A recommendation protocol ensured that the pharmacist and, if necessary, the physician was notified if patients had a very high BP or a very low BP accompanied by symptoms. A public health nurse was 'on-call' during the sessions in case clinical advice was needed. A volunteer coordinator facilitated the transfer of readings and patient-reported cardiovascular risk factors to physicians.

Accurate BP readings were obtained from the BPM-100 automated BP measuring device. Questionnaires were used to obtain self-report data on cardiovascular disease risk factors from patients, knowledge and attitudes about BP control from physicians, and attitudes and satisfaction with the program from peer health educators, pharmacists, and public health nurses.

The primary outcome measure followed an 'intention to treat' approach. The primary endpoint was the difference in the change from 12 months pre- to 12 months post-intervention in the proportion of eligible patients with a BP reading recorded in their health record with mean systolic <=140 mm Hg or mean diastolic <=90 mmHg; <=130 or <=80 if target organ damage or diabetes is present.

This trial is a collaboration by the Department of Family Medicine, McMaster University and the Elisabeth Bruyère Research Institute, a University of Ottawa and SCO Service Partnership.

28 family physicians were recruited; all eligible patients in practices allocated to the intervention arm were invited to the program. A randomly selected set of 50 charts per practice were audited, regardless of session attendance.

Physician Practices

Inclusion criteria

• Non-academic (none of its members are full-time faculty members)
• Full-time
• Regular family practice in terms of size and case-mix
• Able to provide an electronic roster that includes a mailing address of their patients 65 years and older

Exclusion criteria

• work in walk-in clinics or emergency departments
• about to retire
• work part-time
• fewer than 50 patients 65 years or older
• specialized practice profile

Patients

Inclusion criteria

• Community-dwelling
• 65 years or older
• Visited the practice at least once in the last year
• Considered regular patients by the physician

Exclusion criteria

• Terminally ill
• Scheduled to have surgery during the program period
• Not mobile to visit the pharmacy
• Non-English-speaking and unable to attend with an English-speaking companion
• Significant cognitive impairment