This study examines the use of patient initiated discussions of blood pressure medication in the primary care setting. Additionally, this study examines different levels of patient incentive(s) to initiate discussions with providers including information only, information plus a financial incentive and reimbursement for 6 months of copay, and information/monetary incentive plus a reminder call prior to the index visit. Patient opinions about doctor/patient relationships will be assessed. Provider attitudes and prescribing behaviors will be analyzed as well.

Patients will be randomized to Pure control, intervention control, intervention group a, intervention group b or intervention group c. Pure control patients will not receive any study information about diuretics and their providers will not see any patients who receive our study intervention letter. Patients in the intervention control group will not receive the intervention but will see providers who have other patients in the intervention groups. Patients in intervention group a will receive a customized letter prior to their next primary care appointment with information regarding the patient's specific blood pressure, current blood pressure medication and recommendation for switching or adding a thiazide. Patients in group b will receive the customized intervention letter as well as a financial incentive for discussing the information with their provider and a copay reimbursement for 6 months if a prescription is written for a thiazide. Patients in group c will receive the intervention letter, financial discussion incentive, copay reimbursement as well as a phone call reminder 1-2 days prior to their primary care visit.

Inclusion Criteria:

The eligible population includes patients less than 80 years of age enrolled in primary care clinics at the study sites that meet the following additional inclusion criteria:

• No active prescription for a thiazide diuretic (or a medication that combines a thiazide and another agent (e.g., hydrochlorothiazide and triamterene)
• Prior diagnosis of hypertension, as documented in VA outpatient administrative files
• Two or more visits to a primary care clinic in the prior 12 months;
• Blood pressure above treatment goal at the two most recent VA outpatient clinic visits Above goal is defined as a systolic pressure >140 mmHg or diastolic pressure >90 mmHg. For diabetics above goal is defined as systolic pressure >130 mmHg or diastolic pressure >80 mmHg
• Blood pressure at goal (as defined above) during one of the last two outpatient clinic visits, but the patient is receiving a prescription for a calcium channel blocker (CCB)

Exclusion Criteria:

Patients meeting the above inclusion criteria will be excluded for the following reasons:

• Documented allergy to thiazides or to sulfa agents
• Previously documented intolerance or adverse drug reaction to thiazide diuretics
• Active prescription for a loop diuretic agent (e.g. furosemide)
• Renal insufficiency, defined by a glomerular filtration rate less than 30 ml/min
• No serum creatinine in the past year (to permit calculation of a creatinine clearance)
• Prior history of hypokalemia or serum potassium less than 3.5 meq/l in the prior year
• Diagnosis of gout or active prescription for allopurinol
• Congestive heart failure (CHF) due to systolic dysfunction with a documented left ventricular ejection fraction < 35% by echocardiography, nuclear medicine study, or ventriculography
• Residence in a long-term care facility
• No telephone for follow-up calls
• Life-expectancy < 6 months
• Inability to give informed consent or impaired cognitive function (defined as > 4 errors on the 10-item Pfeiffer Portable Mental Status Questionnaire82, administered during study intake)

Baseline data were only collected as combined control and intervention groups and it is not possible to separate by the 3 arms of the intervention for separate analysis. Because of randomization, the difference in baseline characteristics should not be meaningfully different.

There were not any reported Adverse Events, so there is not separate reporting of events, because there were no events. The numerator for all is zero.

Because there were no events, the denominator was combined into all the Control Patients and all the Intervention Patients. It is thus not possible to separate them since there were no events.