Study participants, all age 20 years or older, were recruited at Föllinge Health Centre (population 1500), located in a rural area in Region Jämtland Härjedalen, northern part of Sweden. Eligible study participants were primary care patients treated for essential hypertension (International Classification of Diseases 10; I10) or examined for other conditions for which blood pressure control is important, e.g. cardiovascular disease, previous ischemic stroke or transitory ischemic attack, diabetes mellitus type 2, chronic kidney disease, obesity, hypercholesterolemia, and previously untreated high blood pressure. Exclusion criteria were need for emergency care, advanced stages of disease or dementia, inability to perform self-directed BP measurement, pregnancy, atrial fibrillation, arm circumference too small or wide for the device, lack of consent, or re-entry into the same study. Enrollment was conducted from February 1, 2021, through December 30, 2022.

All participants underwent a 12-lead electrocardiogram, biochemical panel, physical examination, and physician-measured office BP. The time span from enrollment to ABPM and the last BP measurement in the office or at the kiosk was normally within 2 days, up to a maximum of 7 days, depending on logistics (Fig. 1).

Flowchart of study procedure, Abbreviation BP; blood pressure

From the enrollment visit through the last study-related BP measurement, study participants stayed on the same medication. The ABPM protocol was available to staff and participants only after the in-office and kiosk BP measurements were completed. Otherwise, there were no blinding procedures.

To randomize participants to kiosk and office BP measurement before ABPM was set up, sealed envelopes were used at the enrollment visit. After participants completed ABPM, the kiosk and office BP measurements were repeated in reverse order (Fig. 1).

BP measurements were performed with the following meters: in-office, Omron M7 Inteli IT (CE. 01997), certified for arm circumferences 22–42 cm; in kiosk, Omron i-Q142 (CE. 0197), certified for arm circumferences 22–42 cm; and for ABPM, Meditech ABPM-05 Blue BP-05 (CE 0120), cuff widths 120–250 mm for 23–33 cm arm circumference and 150–330 mm for 31–40 cm arm circumference. All study equipment was validated by the department of medical technology, Östersund Hospital, before and during the trial on a yearly check-up scheduled.

Kiosk and office BP values were registered after 5 minutes of quiet rest, with the patient sitting in a chair with a backrest, arm supports, and feet on the floor, with the BP cuff on the left upper arm at the level of the heart. Three BP values were recorded at an interval of 1–2 minutes, with further measurements only if the first two values differed by > 10 mmHg. Mean BP was taken using the last two measurements, and the total mean BP was taken as the average of four measurements, two before ABPM in the BP kiosk and two after ABPM, by the nurse in-office. BP at the enrollment visit was recorded as the mean of the last two measurements in a series of three (all physician-measured BP). Trained study nurses conducted the other office BP measures. Kiosk BP was assessed in a separate kiosk for self-directed BP measurement, adjacent to the clinic. Study participants were instructed about using the device before BP measurement in the kiosk, with only one study participant and no staff present in the kiosk. The study participants were instructed to register their BP values by hand on their study forms, and a mean value was subsequently calculated.

ABPM was registered every 30 minutes from 06 A.M. until 22 P.M. and every 60 minutes at night, for a total of 24 hours. The mean value calculated from the measures taken during 06 A.M. to 22 P.M was recorded as the daytime AMBP.

Measurement of in-office BP and ABPM was performed on the left upper arm for comparison with the corresponding measurements in the kiosk, where the device was adapted for BP measurement on the left upper arm, with a button control for each measurement on the right-hand side.

The primary outcome was comparative measurements of systolic and diastolic BP among in-office (including nurse- versus physician-measured BP), kiosk, and daytime ABPM values. The secondary outcome was the diagnostic performance of in-kiosk and in-office (including nurse- versus physician-measured) BP, with daytime ABPM as reference. For daytime ABPM and kiosk BP, we defined the diagnostic cut-off as ≥135 mmHg systolic and/or ≥ 85 mmHg diastolic BP. For office BP, we defined the diagnostic cut-off as ≥140 mmHg systolic and/or ≥ 90 mmHg diastolic BP.

In exploratory analyses, we evaluated the performance of ≥140 mmHg systolic and/or ≥ 90 mmHg diastolic BP in kiosk and ≥ 160 systolic and/or ≥ 100 diastolic BP in office and in kiosk, with daytime ABPM as reference. We also evaluated the diagnostic performance of office and kiosk BP measures with the total 24-hour ABPM as reference (diagnostic threshold ≥130 systolic and/or ≥ 80 diastolic for 24-hour ABPM). In addition, we conducted this analysis separately for participants with drug treatment for hypertension. The diagnostic cut-off levels for in-office and kiosk BP and ABPM were selected according to recommended reference standards [4].

For sample size calculations, the difference among BP settings was set at ≥4 mmHg systolic BP with a standard deviation (SD) of 15, nominal power 0.8, alpha 0.05, and an assumed correlation of 0.5. The calculation resulted in a necessary minimum of 113 participants. The estimated SD for systolic BP was based on previous BP registrations in the same population. To account for potential late exclusions and drop-outs, we aimed to recruit up to 150 study participants.

Standard descriptive measures, means, and proportions were presented for baseline variables, and the paired t-test (two tailed) was used for comparisons of means. The statistical significance level was set at p < 0.05. SPSS (version 25, IBM Corp) and a diagnostic test evaluation calculator (MedCalc Software Ltd.) were used for the analyses [20].