Cuffless Blood Pressure Measurement: How Did Accuracy Become an Afterthought?

In many aspects of life, accurate measurement is of critical importance. One cannot successfully pilot an airplane, follow a recipe, or construct a bridge without careful measurement. This ostensibly simple concept seems neglected in, of all places, the field of blood pressure (BP) measurement. Given the substantial clinical importance of BP measurement, it is bewildering that the importance of ensuring accurate readings is often treated with indifference. Two main reference standards for BP measurement exist.1 The first, applicable primarily to the critical care setting, is intra-arterial measurement. The second, germane to most other aspects of clinical care including the diagnosis and management of hypertension, is standardized, simultaneous, blinded, two-observer auscultation. These two reference standards produce physiologically different measurements and are not interchangeable. If a BP monitor is intended to be used for the diagnosis and management of hypertension outside of the critical care or research setting, its accuracy should first be confirmed by comparing it to two-observer auscultation using an internationally recognized BP measurement validation protocol.1–3 Over the past decade, rapid growth in activity within the “cuffless” BP measurement space has occurred, and these devices are projected to seize an ever greater share of the broader multibillion dollar BP monitoring space. Most cuffless BP devices are wrist-based, although other sites (e.g., ear, forehead, finger) are also used. A variety of BP estimation techniques are employed, including tonometry and pulse transit time (PTT). PTT, the most commonly used method, is calculated from electrocardiographic and peripheral photoplethysmographic (PPG) signals, as the time interval between the R-wave of the QRS complex and either the foot, midpoint, or peak of a simultaneously recorded PPG signal.4 PTT-based BP estimation is based on the principle that PTT and BP are highly and inversely correlated (although the degree of correlation varies).4 PTT-based devices estimate BP using mathematical algorithms, which can be simple equations, regression models, or more complex methods like neural networks.4 Initial calibration to conventionally performed cuffed BP measurement(s) is required; the algorithm is subsequently used to estimate BP. Manufacturers of PTT-based devices often emphasize the cuffless and continuous BP “measurement” features that can enhance user comfort and convenience. To say, however, that PTT-based devices “measure” BP is a misnomer because a pressure value is never measured. Cuffless BP measurement has several challenges (Table 1). While rigorous validation protocols exist for conventional (cuffed) BP devices, they are not yet available for cuffless measurement. This is a major barrier to development of accurate cuffless devices and results in a paucity of proper guidance for investigators and manufacturers. The Institute of Electrical and Electronics Engineers (IEEE) standard is the only one currently available.5 This validation protocol has adopted similar thresholds to conventional blood pressure measurement validation protocols, including the stipulation that device-under-test should estimate BP to within 5 mmHg of the reference standard to be considered acceptably accurate.1 A major limitation of the IEEE protocol is that the reference measurement procedure is not described in sufficient detail to ensure that it will be performed in standardized fashion. A new version of the IEEE standard is under development that hopefully addresses these limitations, as is an International Standards Organization validation protocol that covers specifically the validation of continuous automated BP Devices. Limitations of cuffless BP measurement Abbreviations: BP, blood pressure; IEEE, Institute of Electrical and Electronics Engineers; ISO, International Standards Organization. Limitations of cuffless BP measurement Abbreviations: BP, blood pressure; IEEE, Institute of Electrical and Electronics Engineers; ISO, International Standards Organization. Most cuffless BP device studies reported in the literature do not follow a proper validation protocol and do not use a proper reference standard (instead a conventional oscillometric device is used). While an initial calibration procedure is often described, the possibility of drift and need for recalibration is not considered. This prevents the ability to make definitive conclusions about validation and accuracy. An additional challenge is motion artifact and change in the position—these effects on BP estimation can be substantial,6 This is clearly problematic for a class of devices that are touted as delivering “continuous” (beat-to-beat) BP measurement. Accordingly, it is essential that manufacturers of PTT-based devices intended for use in ambulatory individuals ensure that BP is measured accurately beyond the rested, motionless state. The Somnotouch-NIBP (Somnomedics GmbH, Randersacker, Germany) is an example of a PTT-based device with validation data.7 However, testing was performed after 15 minutes of rest and only while motionless. When a 24-hour ambulatory monitoring comparison was performed (Spacelabs 90217, Spacelabs, Snoqualmie, USA), the Somnotouch-NIBP produced mean BP levels that would not meet conventional validation standards (mean systolic and diastolic BP levels were higher by 6.8 ± 10.5 and 6.5 ± 7.5 mmHg, respectively).8 The concept of proper validation testing is crucial. Ignoring it is ultimately detrimental to device users and clinicians, who are largely unaware of the nuances involved. Cuffless device makers are often start-up companies that are critically dependent on venture capital. In stark contrast to the clinical world, in the business world, the ability to secure capital relies primarily on a promising business plan—accuracy data are often an afterthought. Devices are sometimes labeled as “for recreational purposes only,” which is a meaningless phrase and only serves to generate further confusion. Recently, an iPhone app (Instant Blood Pressure App, AuraLabs), purported to measure BP by positioning the phone in simultaneous contact with the chest wall and an index finger, was found to be highly inaccurate when independently tested—but only after nearly 150,000 units were sold.9 This finding resulted in a United States Federal Trade Commission lawsuit and eventual settlement. In an analogous example, Fitbit, maker of a popular wrist-based wearable device, recently settled a class-action lawsuit alleging that the sleep tracking function was inaccurate. The negative consequences of inaccurate BP measurement are much greater than that of inaccurate sleep tracking, given, in the case of BP tracking, the potential for initiating inappropriate pharmacotherapy, incurring unnecessary costs, and precipitating avoidable adverse effects. Manufacturers, clinicians, policy makers, regulatory authorities, standards committees, investors, and device users must pay more attention to accuracy and validation assessment. Otherwise, the advent of cuffless measurement may prove to be a curse rather than a blessing for individuals with known or suspected hypertension who are understandably just trying to get a more complete assessment of their true BP. I am a Co-Founder of mmHg Inc., a start-up focusing on innovations in BP measurement, a member of several BP validation standards committees, most actively the International Standards Organization sphygmomanometer committee. Any views or opinions are those of the author(s) only and not of the AJH or its Editors.