Effect of Home and Remote Blood Pressure Monitoring Via Home BP Monitoring System (HMS) on Adults With Hypertension in Taiwan

Design and setting: This is a single-center, open-label, randomized, two-period, two-sequence crossover trial conducted at the cardiology outpatient clinics of National Taiwan University Hospital Hsin-Chu Branch, Hsinchu, Taiwan. Participants were randomized 1:1 to receive the case-manager intervention for months 0-6 followed by the control condition for months 6-12, or the reverse order. Crossover occurred at 6 months without a washout period. Home monitoring common to both conditions: At enrollment, every participant received an upper-arm Bluetooth blood pressure monitor (OMRON HEM-7157T) paired with the HMS smartphone application. The application transmitted every reading automatically, provided a log for medication intake and side effects, and delivered the same system-generated educational messages and reminders in both conditions. Participants were trained to follow the Taiwan 722 home blood pressure protocol in the first week of each month; measurement at other times was allowed. Intervention condition (case-manager surveillance with physician-guided remote titration): Cardiovascular case managers reviewed the transmitted readings at the beginning of each month against two targets: home blood pressure below 130/80 mmHg and self-reported medication adherence on more than 50% of days. A predefined algorithm specified: telephone counseling when participants measured on fewer than 4 days, reported adherence below 50%, or reported side effects; a clinic visit within 2 weeks for home blood pressure above 180/110 mmHg; doubling of the current dose or addition of a drug class when home blood pressure exceeded 135/85 or 145/90 mmHg, respectively, for 2 consecutive months; a change in dosing time or frequency for morning-only or evening-only elevation; and withdrawal of a drug class for home systolic blood pressure below 110 mmHg. All prescription changes were made by the cardiologist at a clinic visit, scheduled earlier when the algorithm required it, and recorded in the medical record. Control condition (passive data review): Readings were stored but not reviewed between visits. At each scheduled clinic visit, the treating cardiologist received a summary of the home readings from the preceding 2 weeks and adjusted treatment at his or her discretion. Assessments: Office blood pressure was measured at enrollment and at 3, 6, 9, and 12 months. 24-hour ambulatory blood pressure monitoring, brachial-ankle pulse wave velocity, and blood and urine tests were performed at enrollment and at approximately 6 and 12 months. Medication changes, side-effect reports, and health care use were recorded throughout. Analysis: Treatment effects are estimated as the 6-month (period 1) difference between randomized sequences, adjusted for baseline values, and as the within-participant crossover treatment effect using the 6- and 12-month values. The sequence effect is examined as an indicator of carryover. This trial was registered retrospectively, after enrollment was completed.