Contribution of telemedicine and information technology to hypertension control

Many open issues stand in the way of satisfactory blood pressure (BP) control in hypertensive patients. The experience gathered in the last decades in the field of hypertension showed that the traditional BP measurement taken during in-office visits is not sufficient for optimal management of this condition. New phenotypes of hypertensive patients have emerged based on the combination of office and out-of-office measurements, such as white coat hypertension and masked hypertension. Despite their acknowledged clinical importance, the routine implementation of ambulatory BP monitoring and home BP monitoring (HBPM) in clinical practice is still limited by their costs, risks of bias in the case of HBPM, and low adherence to management. Another unresolved issue in hypertension management is how to deal with resistant hypertension. Resistant hypertension has been acknowledged to represent quite a hard challenge for physicians and to be one of the leading factors responsible for the hypertension burden worldwide, especially in Western countries [1, 2]. According to the 2018 guidelines for the management of arterial hypertension issued by the European Society of Cardiology (ESC) and the European Society of Hypertension (ESH) [3], resistant hypertension accounts for 30% of the total cases of arterial hypertension.

However, 50% of these patients might indeed be affected by “pseudoresistant hypertension,” and its failure to be controlled largely depends on their low adherence to the prescribed treatments [4]. In fact, low adherence to treatment is one of the two main recognized causes of unsatisfactory BP control [3], along with physician inertia, i.e., the inability of physicians to react to changes in patients’ BP values and clinical conditions over time. Low adherence is as simple to define as it is hard to manage. The measurement of drug metabolites in the patient’s plasma or urine represents the only direct method to monitor therapy adherence. Admittedly, however, application of this strategy to large populations is not feasible, as it is demanding both in terms of costs and time. Therefore, physicians must trust indirect methods to assess compliance with prescriptions, including self-reported adherence and pill counting. Unfortunately, none of these methods is fully reliable. A number of different methods have also been proposed by the 2018 ESC/ESH Hypertension Guidelines to improve patient adherence to treatment [3]. These include a number of interventions at the physician level, at the patient level, and at the healthcare system level (Table 1).

The progressive advancement in information and communication technologies (ICT) has introduced new tools for connecting physicians and patients, starting from simple forms of remote monitoring and transmission of BP records, which became available at the end of the 20th century and represent potentially powerful solutions to improve hypertension management through better patient adherence and reduced physician inertia. This fast-paced technological evolution has paved the way for a new dimension of healthcare management, the so-called digital health (or eHealth) approach, which uses the power of computers and electronic devices to overcome the shortcomings of traditional management and improve its outcomes.

Over the years, telemedicine has evolved from telephone follow-up to web-based and computer-tailored solutions and dedicated devices (eHealth). Since HBPM had already proven to be an effective management strategy for hypertension [5, 6], mainly by promoting empowerment and self-awareness of this condition in patients, the initial application of eHealth in the hypertension field was telemonitoring of home BP values. This strategy was meant as an improvement of the old remote monitoring made by patients through log-book transcription of regular home BP measurements and aimed at using new devices to acquire and transfer BP measurements to the physicians in charge. It was easy to set up, had relatively low costs, and promised to tighten the link between patients and physicians, thereby increasing patient adherence to treatment and fighting physician inertia. Although this strategy had favorable results in clinical trials, it was not enough to solve the hypertension management issue, both because of its costs were not covered by health insurance systems and because of the need for significant compliance efforts by patients in relation to its daily use.

With recent progress in ICT, a new approach to healthcare has developed, termed “digital health”. This approach can be considered a branch of eHealth, and it is defined as the convergence of digital technologies with healthcare aimed at enhancing the efficiency of healthcare delivery and making medicine more personalized and precise. Recently, the widespread, massive diffusion of mobile devices (mainly smartphones), along with dedicated applications (apps), has created a new branch of digital health, which is defined as mobile health (mHealth) and can be considered the innovation frontier in this field. mHealth has already found a number of applications—from patient management to data collection for research purposes.

Indeed, mHealth could be considered a suitable candidate to overcome the well-known limitations affecting the current management of chronic conditions, such as hypertension, diabetes, dyslipidemia, and other cardiovascular risk factors. Digital health, and in particular mHealth approaches, appear to be novel and potentially effective solutions to the pending issues mentioned above, i.e., to promote better empowerment of patients, increase their adherence to treatment, and improve communication between the patients and the treating physicians, thereby also fighting physician inertia [3]. Among its most promising features, the possibility of linking apps to wearable devices with the aim of recording data in real time and in a real-life setting is worth mentioning.

Although an increasing number of studies are being performed in this field, it is unfortunately difficult to summarize the available evidence on the clinical impact of digital health, as this definition encompasses a wide variety of strategies and devices, especially in the field of mHealth, which is characterized by exponential diffusion, although admittedly its use is not yet adequately standardized. In fact, mHealth systems may differ in modalities of data collection (i.e., manual entry and automatic collection), data transmission (telephone based and internet based), methods of interaction (short text messages, e-mail, phone calls, etc.), frequency of contact, presence and type of feedback, and additional features. Such heterogeneity has so far prevented investigators from proposing general recommendations for the clinical use of eHealth and mHealth approaches. The most abundant kind of available evidence, indeed, regards conventional telemonitoring of home BP values, which has been tested in several randomized controlled trials in recent decades.

Over recent decades, several studies have evaluated the efficacy and sustainability of BP telemonitoring, and they have gathered a sufficient amount of data to be organized in large systematic reviews and meta-analyses.

One of the most important meta-analyses [5], published in 2013, evaluated 23 RCTs (total: 7037 patients) comparing home BP telemonitoring and usual care based on regular in-office BP measurements in hypertension management. At a mean follow-up of 6 months, BP telemonitoring was associated with a greater reduction in office systolic BP (SBP) (4.7 mmHg, p < 0.01) and 24-h ambulatory SBP (3.5 mmHg, p < 0.01) compared with a usual care approach. The chance of achieving normalization of BP values was higher with the telemonitoring approach (relative risk 1.16, p = 0.007). Patients randomized to telemonitoring were prescribed slightly more medications (+0.40 drugs per patient, p < 0.001) and had a similar rate of office consultations. The telemonitoring strategy was associated with higher costs (+662.92 euros per patient compared with usual care) due to infrastructure installation and maintenance costs, while the medication cost was similar to that in the control group.

The largest meta-analysis in this field [7], published in 2017, analyzed 46 RCTs with a total of 13,875 patients. It confirmed the superiority of telemonitoring in relation to BP control, with a reduction in systolic and diastolic BP values of 3.99 and 1.99 mmHg, respectively (p < 0.001 for both). In addition, it suggested that integrated interventions consisting of telemonitoring plus additional support (including counseling, education, behavioral management, medication management, and adherence contracts) led to an even higher reduction in BP values (additional reduction of −2.44/−1.12 mmHg for systolic/diastolic BP compared with simple telemonitoring; p value borderline for statistical significance, mainly due to heterogeneity of studies). In addition, the investigators suggested that longer-term monitoring (6–12 months) was more effective than short-term strategies (<6 months).

Another individual-data meta-analysis on self-BP monitoring came to similar conclusions, suggesting that integrated interventions (medication titration by a case manager, education, or lifestyle counseling) along with HBPM could lead to better results compared with BP monitoring alone [8]. In this analysis, 2 months of self-BP monitoring achieved a reduction of 3.2 mmHg compared with usual care, but these BP changes ranged from a 1.0 mmHg reduction in the case of self-management alone to 6.1 mmHg in the case of self-monitoring accompanied by intensive support (Fig. 1). The efficacy of self-BP monitoring has been shown to be largely dependent on the degree of involvement of healthcare personnel.

Impact of self-monitoring of blood pressure on the relative risk of uncontrolled blood pressure values at 12 months in different subgroups of co-interventions. Adapted from Tucker et al. [8] with permission. BP blood pressure, CI confidence interval, RR relative risk

Full size image

An interesting insight into the perception of these new management strategies by patients and physicians came from the Telescot program [9]. This program included seven studies on the implementation of telemonitoring in a primary care setting in the United Kingdom for the long-term management of widespread chronic diseases, such as arterial hypertension, diabetes, and chronic obstructive pulmonary disease. The investigators performed a qualitative analysis in 181 patients and 109 professionals, and they found a high rate of approval of telemonitoring among patients, who judged it empowering, convenient and capable of improving the day-to-day management of their diseases. On the other hand, at first, physicians expressed some concerns related to the initial implementation of the new systems and consequent possible issues in workload and changes in their relationships with patients. In the end, the physicians judged telemonitoring positively, with a minority expressing criticism mainly about specific system design or implementation. Indeed, this result emphasizes the relative impact of the system infrastructure on the outcomes of digital health interventions apart from the intervention itself.

Recently, the results of the TASMINH4 trial were published [10]. The study randomized 1182 patients either to usual care (management based on regular in-office assessment), HBPM, or a telemonitoring system based on automated, short text message-based communications. Telemonitoring achieved a significant improvement in BP control compared with usual care and a faster reduction in BP values compared with HBPM alone-based management. In fact, at 6 months, patients randomized to telemonitoring had a significant reduction in BP values (while the BP reductions obtained through the use of HBPM alone only approached statistical significance). At 1 year, however, there were no differences in BP values achieved with HBPM alone vs. with telemonitoring, though both showed a similar, significant effect compared with the control method. No difference in adherence to treatment was detected between different groups, even though there was a high chance of a ceiling effect, as scores in the evaluation scale were above 95% of the maximum in each group. Thus, telemonitoring allowed a fast response to an evolving situation, while in the long term, this advantage was less marked. While evaluating the results of these studies, it is noteworthy that in the setting of trials, the adherence of patients to prescribed treatments is usually higher than in real life, so in a real-world context, the advantage of telemonitoring might be even higher due to a lower performance in control cohorts. Finally, questions about the cost efficacy of these eHealth-based interventions remain to be answered. A dedicated analysis of the TASMINH4 trial [11] found self-monitoring to be cost-effective, but the role of telemonitoring in achieving better outcomes was unclear. Additional studies with up-to-date devices are therefore needed.

Despite the good results in clinical trials, BP telemonitoring based on services offered by professional providers has struggled to reach wide implementation in daily clinical practice, mainly due to high installation and maintenance costs and difficulties related to older-generation infrastructures, especially in the late 1990s and early 2000s, as remote monitoring required dedicated and expensive systems for the transmission of data.

As stated above, the impressive expansion of the mobile device market has led to a radical change in the concept of digital health and to the development of mHealth. Compared with their previous generations, modern devices are cheaper and more flexible, so they are easily employed in different settings and interventions, with a high propensity for further expansion or readaptation.

Thus, the initial concept of telemonitoring has evolved, moving from traditional phone calls, e-mails, and short text messages to more complex and automated transmissions of data related to different aspects of physical, emotional, and behavioral variables [12, 13]. On the one hand, smartphones can indeed be equipped with a wide variety of sensors (both built-in or connected as accessories) so that they are able to acquire a large amount of data and perform complete remote monitoring of different physiological and behavioral parameters (e.g., heart rate, physical activity, sleep cycles, blood oxygen saturation, etc.) [14, 15]. The extreme flexibility of accompanying software can add a wide variety of additional functions, thus targeting different dimensions of healthcare, not only simple monitoring.

Despite this increase in functionalities, implementation and maintenance costs have dropped due to the lower costs of devices and internet connections, and digital health has also been suggested as a feasible and affordable patient management option in daily care.

The potential role of new mHealth interventions has been addressed even by a recent document of the World Health Organization [16], which has called mHealth a potential promoter of better health conditions in low-income countries.

In 2015, the American Heart Association published a scientific statement on the use of health-related smartphone applications for cardiovascular prevention, covering a wide range of risk factors and behavioral interventions (control of BP values, smoking cessation, exercise training, adoption of a healthy diet, management of dyslipidemias and diabetes mellitus) [17]. A thorough review of available evidence has supported the conclusions on the effectiveness of strategies based on mobile apps. In the end, while praising the potential benefits of this approach, the authors stated that the high heterogeneity of proposed interventions and the lack of standardization of data collection in available studies have hampered our ability to draw strong conclusions. Indeed, there were 12 trials and 2 systematic reviews [18, 19], with a follow-up shorter than 12 months (usually <6 months) and a high reliance on self-reported outcomes, without external adjudication of objective events. Moreover, such a short follow-up prevented the investigators from gathering reliable data on adherence to management strategies, which is well known to drop significantly in the long term. In the most representative trials, self-monitoring of BP and remote support achieved a net reduction of SBP values of 8.3 mmHg compared with 2.1 mmHg with usual care. Longer interventions, the use of multiple behavioral techniques, and proactive interventions aimed at empowering the patients were associated with a trend toward better outcomes. In a trial, mHealth-related interventions were not related to increased costs compared with usual care based on regular in-office visits. A temporary increase in costs was detected in the short term, and it was deemed to be related to the increasing levels of awareness of their conditions in patients. Over medium- to long-term follow-up, however, no differences in outcomes were detected between the two groups [20]. In other fields explored in the consensus documents, the authors found the new mHealth-based interventions to be able to achieve good results in weight loss and to increase exercise training. In addition, smoking cessation achieved good results, with a cessation rate that was almost double that obtained with traditional interventions. However, the authors also highlighted the need for integrated strategies, as single interventions have low rates of success in the long term, and in the case of mHealth, the discontinuation rate was very high.

An updated meta-analysis focusing on mHealth, published in 2019 [21], confirmed a mean reduction of 3.85 mmHg for SBP and 2.19 mmHg for diastolic BP after 12 months compared with usual care. The investigators found signs of even better results in patients with inadequate BP control at the time of enrollment (mean reduction of 5.02 and 4.96 mmHg for systolic and diastolic BP, respectively; in patients with adequate control at enrollment, the reduction was only 1.24 mmHg for SBP, and there was a mean increase of 3.30 mmHg for diastolic BP).

As a consequence of the high interest in the field, the number of ongoing or just-published studies is increasing at a fast pace, opening new perspectives in the management of hypertension. An example is a pilot study performed in the Netherlands on pregnant women at risk for pregnancy-related hypertension and preeclampsia. Remote monitoring of BP and symptoms through a dedicated phone app or web platform was easy for the patients to learn and perform. Rates of acceptance and satisfaction were very high among both patients and health personnel in terms of ease of use and clinical usefulness [22].

Finally, some of the new interest in the mHealth field could derive from the surprising results achieved by two large studies on atrial fibrillation screening with the use of smartwatches [23, 24]. The investigators were able to enroll several hundred thousand volunteers in a short time and to demonstrate the potential usefulness of these wearable devices in detecting atrial fibrillation, especially in subjects without known risk factors for arrhythmias and, therefore, with a low chance of getting a diagnostic workup. These data clearly show the potential impact of these new technologies in both clinical practice and research, such as hypertension, especially once reliable cuffless BP measurement methods finally become available.

Despite the promising results and future potential of mHealth-related interventions, there are still some issues in digital health that should be addressed and that have been extensively described in a previous paper (Table 2) [25].

Most of them are related to the fast-paced growth of this field and to the consequent absence of clear and strong regulations, as ruling authorities are struggling to keep up with the technology. In 2015, the US Food & Drug Administration released guidance recommendations for the developers and distributors of health-related apps to establish specific requirements [26], but worldwide regulations are still insufficient.

First, data security. Health-related data are sensitive data and therefore have significant value. Despite several improvements in recent years, privacy is still a critical issue in the management of these data, as shown by the recent scandals related to privacy issues on social media and the resulting new laws of the European Community on privacy matters. Attention to this topic is rising among the general population, but its improvement is still needed.

Thus, adequate privacy and protection of data must be achieved, both during transmission of results from patients to dedicated services or to the physicians in charge and during long-term storage. Data security also concerns the nature of these data as property: they have high economic value, so there must be clear reference to the right to use and store them to prevent sharing with unwanted third parties.

The accuracy of the scientific content of the services is another crucial topic. Of course, all devices and applications must have proper scientific validation. In addition, smartphone applications frequently provide additional features, such as educational sections for patients or decisional tools for physicians. Therefore, the accuracy of scientific content is a key point. It should be produced or validated by recognized international authorities in the field, and this should be a major element of judgment when reviewing applications. This is particularly important in the case of mobile phone applications, as their proper scientific validity has not been shown to be a predictor of high popularity or better reviews in digital stores [27]. A possible solution could be the official development or endorsement of applications by scientific societies, groups, or experts. An example is the ESH CARE app, endorsed by the ESH and by the Italian Society of Hypertension, which aims to pave the way to a new generation of health-related applications, and it has a solid scientific background and a comprehensive investigational plan. A pilot study showed promising results, with significant improvements in BP control compared with usual care (regular in-office measurements) [28], and large randomized controlled trials are currently underway (Fig. 2).

Rate of normalization of office and home blood pressure control in the “Patient Optimal Strategy of Treatment” (POST group) versus control group (usual care) after 6 months of follow-up. Data from the pilot study of the ESH CARE App project, endorsed by the European Society of Hypertension (ESH). From Albini et al. [28] with permission

Full size image

Of course, costs are a possible limitation to account for. Even though mobile devices are relatively cheap, some interventions may require dedicated software or infrastructures that may still be expensive in terms of installation, maintenance, or training. In addition, time-related costs must be considered, as physicians, nurses, and technicians might need specific training to become proficient, and the early period may see at high rate of quitting, especially by older people who may be less comfortable with technology.

General bias may still be relevant. Nonautomated recording and transmission of BP values might be particularly prone to bias due to the preference for “zero” BP readings or values just below alarm thresholds, which also affects conventional HBPM, even though its real impact may not be significant [29].

Finally, even more advanced systems still depend on traditional, old-fashion BP measurement with a cuff, as cuffless devices are still not accurate enough for daily employment in real-life clinical practice. Therefore, the implementation of the above-mentioned strategies is good as long as the BP-measuring device’s accuracy is checked and the device itself is validated. Validated BP-measuring devices and attention to methodological issues are thus still the mainstay of every strategy in this field.

Digital health is a promising reality and has the potential to bring significant improvements to the management of hypertensive patients. Available studies suggest a beneficial effect on BP control, but the high heterogeneity of proposed interventions and the lack of standardization of available trials are a strong limitation to the formulation of recommendations based on solid evidence.

There is a strong need for adequately powered randomized controlled trials to address the efficacy, feasibility, and cost-effectiveness of these new strategies, especially in the field of mHealth. Close collaborations between industry and academic institutions should aim to develop better digital health systems, more accurate BP-measuring devices (possibly also including wearable solutions) and, in the end, truly useful tools for physicians and patients.