9.3.1 Focus Groups

In a series of nine nominal groups, 63 providers were asked to discuss and rank barriers and facilitators to conducting HBPM and ABPM [14]. Providers suggested that there were several barriers that prevented the use of HBPM in their clinic. These were grouped into themes according to the Theoretical Domains Framework. The most commonly reported barrier that prevented the conduct of HBPM related to beliefs about capability and consequences (e.g., ability of patients to correctly perform HBPM, test results being inaccurate due to the use of an invalid device or patients not following the BP measurement protocol). Additionally, the cost of the HBPM devices, low reimbursement to physicians, and lack of time to train patients were noted as barriers to performing HBPM. A second study that used focus group discussions with physicians in the Netherlands had similar findings. Specifically, physicians reported that HBPM was inferior to ABPM and that not all of their patients would be capable of conducting HBPM [7]. In this study, it was found that physicians discouraged the use of HBPM as they believed ABPM was a superior approach to measuring BP outside of the clinic setting.

9.3.2 Results from Structured Provider Surveys

In a study from a single health system in the US, more than 75% of providers completing a structured questionnaire on HBPM considered the procedure part of standard care in their practice [15]. Over 90% of providers who reported using HBPM said it was to guide treatment and two-thirds used it to improve adherence. Barriers to conducting HBPM included the lack of knowledge regarding validated devices and lack of data on the scientific evidence that HBPM will result in better BP control. Additionally, providers reported that patients’ poor eyesight and lack of confidence would result in not obtaining valid BP measurements on HBPM. Also, a high proportion of providers (~40% to 50%) reported that they preferred measuring BP in the office setting. Nearly 33% thought that patients would be anxious if their BP was not controlled when measured on HBPM and over 40% of providers reported no one was available in their practice to teach patients how to properly conduct HBPM.

Among a random sample of all primary care providers in Hungary (n = 405; 58% response rate), 98.5% agreed that HBPM was part of standard of care and 94.4% often and almost always encouraged their patients to perform HBPM [16]. HBPM was considered to be equal or more important that office-based BP measurements by over 95% of providers. Despite the high proportion of providers who reported using HBPM, only 67% stated their service taught patients how to conduct HBPM. Barriers to conducting HBPM included concerns about the availability of validated devices (79%), that patients would become preoccupied with their BP (54%), and that most patients were not properly trained (40%). Also, over 25% of providers were concerned that the HBPM results would make their patients anxious and would result in frequent phone calls to the office. Facilitators to increase the use of HBPM included the availability of training facilities, inclusion of diagnostic and treatment protocols based on HBPM, programs that tabulate/display the HBPM results and evidence that it improved BP control.

In a third study, a random sample of primary care providers in Ontario, Canada, was mailed a survey on HBPM [11]. Among 478 providers (response rate 55%) who reported treating patients with hypertension, 52% considered HBPM part of standard of care and 63% often or almost always encouraged HBPM. Overall, 98% of providers reported using HBPM to detect white coat hypertension, 93% to guide antihypertensive treatment, 69% to improve medication adherence, and 56% to confirm the presence of resistant hypertension. Similar to the study in Hungary, a high percentage of providers reported barriers to HBPM with 70% being concerned that patients would become preoccupied with their BP and 65% uncertain of the accuracy of home devices. Also, 63% stated they would use HBPM more often if they had a list of validated devices, 49% if devices were more affordable and 45% if more evidence were available showing that HBPM improves BP. Only 5% of providers reported having someone in the office available to train patients.

Finally, a structured telephone survey was conducted among primary care providers in Greece (n = 366; 87.4% participation rate) to investigate the implementation of HBPM guidelines [17]. Overall, 94% of providers reported using HBPM for their patients with hypertension. The most common indications were white coat hypertension, treatment titration, and detection of hypertension, while only 1% reported using HBPM to detect masked hypertension. Only 30% of providers based treatment decisions on the results of the HBPM. The main limitations noted for HBPM included 80% of providers who expressed concerns that patients were not reliable in reporting their BP and 41% who questioned the accuracy of the HBPM devices. Additionally, 86% of the providers who reported not using HBPM stated that they did not trust BP readings recorded by patients. While many patients do not accurately report their BP values on HBPM, this problem can be minimized by having patients use a device that stores readings and having them bring their device into the clinic [18]. Also, some HBPM devices have the capability of transmitting BP readings to the clinic which eliminates the need to rely on the accuracy of patients reporting their own BP [19].

9.4.1 Data from Randomized Controlled Trials

A randomized trial conducted in the Kaiser Permanente Medical System evaluated the cost of HBPM versus usual care among 430 patients who were randomized at a 1:1 ratio to a HBPM intervention or usual care [23]. The intervention included receipt of a HBPM device with the request to measure BP twice weekly and mail a record of the recordings along with changes in medications/side effects every 4 weeks. After a 1-year intervention period, the decline in SBP and DBP were 3.3 mm Hg and 1.6 mm Hg larger among participants randomized to the HBPM intervention compared with usual care. Participants randomized to the HBPM intervention had 1.2 fewer office visits and 0.8 more telephone calls with medical staff compared to their counterparts randomized to usual care. In 1986 US dollars, the cost of hypertension care was lower in the HBPM versus usual care randomization arm ($88.28 versus $125.37). Even considering the cost of the HBPM device and patient training, the authors of this study concluded HBPM to be cost saving.

The cost-effectiveness of HBPM was evaluated in a randomized trial that showed HBPM in conjunction with clinical pharmacist specialist meetings reduced SBP by 21 mm Hg versus 8 mm Hg for those randomized to usual care [24]. Over 6 months of follow-up, the HBPM intervention was associated with hypertension-related costs of $455 per-patient versus $179 per-patient for those randomized to usual care. The higher costs with HBPM resulted from increased contact with healthcare providers, laboratory monitoring, medication use, and the HBPM device. HBPM did not reduce the need for outpatient, hospital or emergency department visits. Total healthcare costs were also higher among participants randomized to the intervention versus usual care ($1530 versus $1283). Extrapolating the BP-lowering of the HBPM intervention across the lifespan, it was associated with a favorable incremental cost-effectiveness ratio (ICER); $20.50 for each 1 mm Hg lowering of SBP and $3330 per additional life-year gained.

9.4.2 Modeling Studies

Using a Markov model based on healthcare expenditures in Japan and the prevalence of white coat hypertension from the Ohasama study, HBPM was reported to be associated with cost savings for the Japanese population (medical costs: $9.33 million US dollars [1.09 billion Yen] per 1000 patients over a five-year period with implementation of HBPM versus $10.89 US dollars [1.27 billion Yen] without implementation of HBPM) [25]. The authors reported that the cost-effectiveness of HBPM would be more favorable when conducted in populations with a higher prevalence of white coat hypertension and a lower annual transition rate from white coat hypertension to hypertension based on HBPM.

In a US-based study of patients with health insurance, a decision-analytic model was used to evaluate the short- and long-term cost-benefit and return on investment comparing HBPM versus BP recorded in the clinic setting for the diagnosis and treatment of hypertension [26]. From the health insurer’s perspective, HBPM was associated with net savings that were higher at older age and increased over progressively longer follow-up. For example, over a 10-year time horizon, HBPM was estimated to result in cost-savings of $414.81, $439.14, and $1364.27 among adults 20–44, 45–64, and ≥ 65 years of age, respectively. When the cost-effectiveness of HBPM was divided into its use for the diagnosis versus treatment of hypertension, it was estimated that HBPM provided a better return on investment for diagnosis at younger ages and for guiding treatment at older ages.

Lovibond and colleagues conducted a Markov model-based analysis to compare the cost-effectiveness and quality adjusted life years gained when diagnosing hypertension based on BP recorded monthly in the clinic setting over 3 months and by HBPM over 1 week [27]. At each age evaluated (40, 50, 60, 70, and 75 years), diagnosing hypertension by HBPM and clinic-measured BP were equivalent in terms of costs and quality-adjusted life years gained for both men and women. However, HBPM was deemed to be cost-effective in sensitivity analyses wherein it was presumed to have the same sensitivity and specificity for identifying hypertension as ABPM or in younger age groups when the frequency of repeat monitoring following a normal results was reduced from 5 years to 1 year. A subsequent modeling analysis using data from the US reported HBPM to be associated with higher costs and lower quality-adjusted life years than using clinic-measured BP for the initial diagnosis of hypertension [28]. It should be noted that these studies were strongly influenced by the sensitivity and specificity of HBPM for diagnosing hypertension and there are few data available for generating these estimates [29]. Additionally, data are needed on the cost-effectiveness of screening for masked hypertension.