A list of devices which have been validated independent of the manufacturer can be found at the following websites:
hypertension.ca/hypertension-a...
bihsoc.org/bp-monitors/

The model numbers are sometimes slightly different in different countries, usually with the suffix changed.

Specific to the US market, there is no trustworthy resource for finding clinically validated blood pressure devices. The AMA, AHA, APhA, and others are planning to launch a US-specific listing of independently validated BP devices. This can't happen soon enough! Here is the press announcement from AMA Wire:wire.ama-assn.org/delivering-c...

It would be great if there was a "Clinically validated" marking on the device that said it was a clinically validated device. For example the CE marking on a device means the device has passed the tests required.

HI Deepa. It would be great, I agree. CE marking in Europe is similar to UL marking here in the US. It signifies electrical safety and attests to basic performance. At this time, CE mark does not mean "clinically validated" any more that UL mark does. Europe is well-populated with bad blood pressure devices, for example, just like the US. Hypertension Canada has a "recommended device" product labeling program, but my understanding is that such schemes are against the regulations of AMA and AHA as matters of ethics and conflict of interest come into play. So, while a unified validated device "labeling" program would be great (I agree), I think we will end up with an on-line list of validated devices in the near term, and likely no labeling solution.

What does "issues" mean? Lack of testing data available? Poor performance? Please explain your question.

Poor performance describes what I'm thinking of well. Once the device is out in the market and any performance issues arise that affects accuracy or safety, are there mechanisms in place to report this? And if yes, then is there consensus that these mechanisms are adequate?

Josh would know

If a device is not operating appropriately, either in terms of accuracy or safety, the best thing to do is report the issue to the manufacturer. By law, the manufacturer has to investigate the issue and report to the FDA, including whether a serious safety problem arose because of the problem. Patients and clinicians can report directly to the FDA as well, via the MAUDE reporting system, but action is more likely to be taken if reported to the manufacturer.

Dr. Ross is correct. Manufacturers are regulated by the FDA, and complaints regarding after-market device performance are considered "reportable events." There are specific regulated, and audited procedures the manufacturer must follow in response to reportable events. At the end of the day, however, if a non-validated device has been cleared for market by the FDA (many are cleared based on the controversial "substantial equivalence" pathway), there is no "reportable event" that will suddenly make the devices accurate or worthy of use in clinical decision-making. The type of of clinical validity we are talking about cannot be cured through a reportable event, since the low bar of "substantial equivalence" is all that is required in terms of performance. The reportable event procedure can address serious safety issues, for example, but it can't make a bad device a good device!

Just so anyone reading who may having something to report, here is the link to MedWatch for both health professionals and patients/consumers:
accessdata.fda.gov/scripts/med...

Additionally, here is a link to the MAUDE database that Dr. Ross mentioned, which has reports from both mandatory and voluntary reports:
accessdata.fda.gov/scripts/cdr...

Independent validation data on demand. AMEN

Make visible the testing results section from all FDA 510k applications associated with cleared class II and above devices for independent review and inspection. This is a corollary to the above comment.

Can only happen if 510(k) law repealed. Institute of Medicine already condemned it to no avail. Maybe if AMA, AHA, ASH, APhA, nursing organizations, etc, etc band together and march on DC and get Wolf Blitzer involved...........Maybe.

One way of thinking about clinical validity is the path from the original signal capture to end-user reading and interpretation. There's typically a sensor making a physical reading. The device has software that interprets the reading and stores the result in a local database using a data model. There might then be transmission of those data from the device to an aggregator service, then to various applications e.g., analytics tools, EHRs, clinical systems or clinical registries. At each stage in this process, the data are packaged for transmission, transmitted, received, unpacked and then loaded into the application or service in that step in the path.

If the operation of a device's sensors, the device software reading and interpreting the sensor output and the device data model are capturing the right data and metadata, performing measurements in ways that conform to current accepted clinical practice, then those data are more likely to be clinically valid.

As the data flow from point to point in this path, if the data transmitted can be read and understood in the same way by all devices and apps in the path, then this clinical validity is more likely to remain throughout the path.

So what we want to do in IHMI is make it easier for data to be portable, which is the state in which the same data can be read and understood in a clinically-valid manner by all the devices and apps in a path.

Dr. Zeleski, what most people do not understand is that the majority of blood pressure devices are cleared for market by the FDA using a "substantial equivalence" claim. This means they have submitted NO clinical testing data to the agency. So further to your comment, there would not be much "data" to review. Complicating this is the FDA 510(k) regulation itself, which (as Dr. Alpert points out) prohibits the agency from revealing manufacturer submitted testing and/or clinical data. As I stated before, there is no point looking to the FDA to resolve this issue. The only solution is for professional organizations to establish clinical review protocols and infrastructure independently, and make their review results available to the public. That is the rationale behind this AMA/AHA project (see link), and I encourage everyone in this discussion to get behind this effort, and other like it for other class II medical devices that are being used in patient care: wire.ama-assn.org/delivering-c...

At a more general level, we need better ways to track medical devices and relate them to downstream patient outcomes. All Class II and III medical devices have a unique device identifier (UDI) on their label and packaging (which can be linked to the FDA's Global UDI Database, GUDID), but we need the UDI integrated into electronic health records and insurer claims data in order to understand the safety and effectiveness of these devices when used in clinical practice.

Great point Dr. Dhruva. No biometric should exist in the EMR without a device identifier. This will allow inclusion/exclusion control based on clinical validation, and will offer powerful analytics in future years to determine comparative predictive power of different devices.

Since the FDA has a permanent representative AT THE ISO MEETINGS, there is NO REASON for the FDA not to accept portions of the Standard.........that representative was there during the discussion. If he didn't like the sentence he should say so.......Are all aware that ISO IS THE USA NATIONAL STANDARD?

I am not sure I understand the question: is this about data standards, or is this about evidentiary standards (ie, if EMA approves or clears a medical device for market, FDA should abide by that decision and allow marketing in the U.S.)? The latter is much more problematic, and difficult to implement.

Hi Dr. Ross, Let me clarify, as I was thinking of the broader set of evidentiary standards that the FDA currently recognizes.

In this AAMI piece: aami.org/productspublications/...
It's highlighted that the FDA shifted in September 2017 to include more nationally and internationally recognized standards and that the aim was to "reduce the burden on manufacturers from inconsistent standards in use by different international regulatory authorities". The biggest change was "that the center is now required to publish its rationale for recognizing or not recognizing a standard in the Federal Register".

The question is, is this enough for healthcare stakeholders to feel confident in the validation of these devices? Is the process robust and rigorous enough? Are there standards that should be included and are not? This may depend on the type of device and which standards are recognized.

If discussion participants have knowledge and experience of which data standards are currently recognized by the FDA and if there are more that should be added, I think that would also be of value.

I think we need to look past the FDA. Without a new act of congress, signed into law by the president, the FDA is limited in its oversight of consumer medical devices by 42 year-old legislation. Most non-invasive personal biometric devices (such as blood pressure monitors) are listed as "class II" devices subject to FDA 510(k) regulations. Under the 510K any device that (in the eyes of FDA reviewers) demonstrates "substantial equivalence" to devices on the market prior to 1976 must be cleared for market. Independent clinical validation is not required, and direct testing by the FDA is not in scope. The FDA is not allowed to make public any device testing data submitted by manufacturers, and in instances where the public may ask manufacturers for clinical testing data (not all cases apply), the manufacturer has the option (usually exercised) to fully redact any and all test data, which is considered proprietary. FDA 510(k) reviewers are not obligated to adhere to any international standards, and even if they were, it is impossible for the public to inquire or ascertain the reviewer's level of training or expertise related to specific standards and clinical reviews. "Peer Review" in the academic sense, is not called out under the 510(k) - there is no transparency, and there is no accountability process for individual product clearance decisions. This situation is enshrined in US Federal law by a 1976 act of Congress. So when the FDA announces it is going to rely more on international standards, we have every right to ask, "what have they been using in their decisions?" And when they say they are "reducing" regulatory burdens on consumer medical devices, we can ask "were the burdens too high?". In reality, we will never be able to ascertain what's actually going on since transparency is legislated out of the process. As we have learned in blood pressure, the solution is to set up independent peer-review procedures outside the purview of the FDA.

I hope EVERYONE reads and rereads Josh's comments. We on the AAMI Committee have been trying for DECADES to solve some of these issues, but the FDA representatives have hidden behind the 510(k) legislation and prevented any progress. Until the US Congress takes some action (fat chance of that) we are always going to be debating till the cows come home without hope of improvement. WE need to storm the Bastille and DEMAND change, like in 1789.

Several points:

1) Numerous and well-published recommended standards exist and have been validated over many years. These not only include HL7, but the device-specific "flavors" designed specifically for patient care device communication. I'm speaking of the IHE PCD profiles. These are well-defined, have been tested at many of the IHE Connectathons, and promote a common communication framework for interacting with medical devices at the INFORMATION level (not speaking about the physical connectivity level, which is still a bit problematic across many medical devices).

2) Device manufacturers need to liberate their data and make it more free for consumption without requiring high licensing fees for access or restricted access to data repositories. This has been an issue historically with a number of multi-parameter monitoring manufacturers and the practice needs to stop. The data belong to the health system and to the patients, NOT the manufacturers. This is also particularly true of medical devices that produce data at frequencies across the spectrum from discrete data for ad-hoc vital signs (i.e., spot checks) through waveform and alarm-level frequency capture.

3) Clear communication from medical device vendors on what constitutes data use for INTERVENTION versus INFORMATION is an important aspect of FDA clearance and needs to be made transparent to health systems. This is important because if health systems cobble together various 3rd party medical devices and information transport products, they run the risk of creating a "Frankenstein" that may not support their intended function. For example, if a health system wishes to use data for intervention (i.e., calling a code or rapid response), then the responders need to be made aware of the integrity and assured delivery of the data. This is no minor point and one that is not well communicated, in my observation, by manufacturers or understood by health systems. Better education and clarification needed.

There is a need for data standards for medical devices so that there is more interoperability and sending data to monitoring devices can be enabled. If the data is being sent remotely there is an additional need for security standards

There are various e-health standards organizations that do good work in the area of structured data standards and interoperability. None of them that I know of incorporate device validity into their standards development - in other words they "assume" that the device sending data into the network is valid. This is a huge gap and cause for concern. NCQA sets standards for quality measurement and therefore their regulations have an impact on reimbursement. Does anyone in this community have experience working with NCQA? I think the biometric device community and NCQA would both benefit from working more closely together to ensure only clinically valid out-of-office patient data populates the EMR.

I think the most important step is to ensure that only validated devices are available on the market. Those of us who sit on the ISO Sphygmomanometer committee have been working on an updated validation standard, to be released early 2019. Ensuring device manufacturers use this standard (or equivalent ones) to validate their device would be a major step forward.

Recent years have seen increased awareness that the regulatory environment (FDA 510K) is insufficient to ensure the safety and efficacy of blood pressure devices going to market. As they say, the first step is recognizing you have a problem!

The legislation that originally created the 510K pathway decades ago is flawed, and until Congress acts on it, we are "stuck" with it. For those interested in the 510K and what it means for class II medical devices such as BP monitors, this report from the 2011 Institutes of Medicine report to Congress that recommended the 510K pathway be scrapped and rebuilt from scratch.
nationalacademies.org/hmd/Repo...

Congress never took action on these recommendations.

The bottom line is that the free market is "buyer beware". Therefore, the medical community has to establish ground-rules and infrastructure to enable the independent assessment of device performance, and a means of communicating individual device qualifications to the public, and to healthcare professionals. One great example of this is the effort the AMA is leading in creating a Blood Pressure "Validated Device List. wire.ama-assn.org/delivering-c...

This AMA list, to be published in 2019 in partnership with the AHA and others, will provide the USA's first evidence-based, non-commercial resource for those looking to source a qualified blood pressure device. Kudos to the AMA for leading this effort. Similarly, Hypertension Canada recently launched their own recommended device listing. Efforts like these contribute greatly to device accuracy, since it finally places a business premium on validity. Now manufacturers will take clinical validation more seriously, which is a good thing for patients and physicians.

Of course, only accurate devices for office or out-of-office BP measurement should be put on the market, yet to date regulators focus more on safety rather than accuracy. This should certainly change and then – problem solved!

Until that time comes, it is important to inform the doctors, the patients and the public on which devices are properly validated for measurement accuracy according to ‘established standards’ (all devices declare that are accurate).

The first effort to list accurate devices has been the dabl website, which was established by Prof. Eoin O’Brien in 2003 and introduced for the first time the concept of ‘validated devices listing’. The dabl became an enormous success and has been adopted by many hypertension guidelines and societies worldwide. For example, in Canada only devices listed in the dabl were recommended by the Canadian Society of Hypertension. After a decade of offering a uniquely valuable service to the scientific community and the public, the dabl could not retain full scientific control and was abandoned by the BP measurement experts. However, the need is there for doctors and patients, and several initiatives now attempt to fill this gap. The British Society of Hypertension has developed its own listing, the Japanese and the Canadian Society also developed devices’ lists, and also Medaval which is for-profit company.

The listing of accurate devices has become much more complex that in 2003. Too many devices, too many protocols (and versions), and even more devices claiming ‘equivalence’. We welcome the first initiative in the USA by the AMA to list accurate devices. We are also working in Europe to develop an international non-profit organization with the mission to optimise BP measurement technology and methodology. Such initiatives are important for improving the accuracy of BP evaluation, which is step 1 for hypertension diagnosis and management and unfortunately remains an unmet need.

Agree with George Stergiou that it is important to inform the doctors, patients, and the public on which devices are properly validated. In addition, it could also be beneficial to inform what those "established standards" are. For example, the ISO system accuracy standards for blood glucose meters are: 95% of readings within ± 15 mg/dl at concentrations < 100 mg/dL and within ± 20% at concentrations of >= 100 mg/dL. Knowledge of these tolerances may be helpful for the patient in the interpretation of their meter readings (e.g., what's the "real" difference between readings of 86 mg/dL and 106 mg/dL?).

Tong Sheng: while absolute values are important, trends and extremes are often more important. If I am looking at a patient with altered mental status, a possible telltale sign is hypoglycemia of a nature that is unmistakable: < 40 mg/dL. But, I am also looking at whether the patient is diaphoretic. Similarly, with hyperglycemia, seeing a patient with readings upwards of 500 mg/dL or more is fairly unmistakable.

Patients are unuque, and although basic guidelines tell us where values should range, we need to take each patient individually. If a type 2 diabetic normally ranges 120-140 and is compliant with her medications, then she may be doing all right. If we see her glucose suddenly drop — hypoglycemia typically occurs very quickly, whereas hyperglycemia evolves much more slowly — then that is a sign requiring a call to action.

With blood pressure we have similar types of trends. The key difference, though, is understanding how close and by how much blood pressure changes before and after medication. An example would be a patient with a current systolic blood pressure measurement ranging close to 90 mmHg and whose pressure had dropped by 20 points or more within 2 minutes post sublingual nitroglycerin administration. Such trends correlated with causal events are extremely important to know and to measure as intervention is required almost immediately.

The two most important steps are to ensure proper patient technique and cuff size and to ensure patients purchase validated monitors. In fact, devices that have not been validated should not be allowed on the market.

Amen to Raj. This will take time to educate the patients and retest them to insure consistency. Good luck trying to carve out the time to do so. Our system fails miserably because no one will take the needed time. Maybe someone else can invent a way to give clinicians time to do these tasks.

Perhaps establish and grow programs such as community paramedic / community EMT that could facilitate this goal, at least at the patient education level? Have trained individuals who enter the patient home assist in such instruction, education and validation?

That's an excellent proposal! In our hypertension clinic in Edmonton Canada, we ask patients to bring in their monitor so that we can check it. The critical components to check are the tubing, connections, and cuff size. We also do a couple 'same arm simultaneous' measurements, listening while the home BP device takes a reading. In most, cases it's possible to get a reasonable assessment of device performance in that individual. If the deflation rate is fast or if it's a 'stepwise' deflation, it can be more difficult to assess. It's not perfect, but oscillometry can be quite off in some patients, so we at least can pick up those cases and advise to the patient to get a different device (from a different manufacturer, with a different algorithm).

I think the AMA is working on reimbursement to providers for assessing and teaching proper home BP measurement. This would be an excellent way to go, wish we could do this in Canada.

Yep. In those patients I have seen, and who have their own blood pressure cuffs, when the situation was not dire and there was time to talk with them, I have asked about the use of their devices and reviewed measurements with them. Cannot always take the time to do this, but it is a start.

With home blood pressure monitoring we have 2 problems.

One is blood pressure measuring accurate devices, which we discussed dealing with another question put in this group. Measurement accuracy applies equally to all blood pressure measuring devices – office, home, ambulatory and any other. I do not wish to discuss here the issue of the measurement accuracy in ‘individuals’. This applies also to office and ambulatory monitors and there is no ‘established’ method to do and that in practice. Certainly an important and complex issue which at present is not for clinical practice. Needless to say, we check all oscillometric devices (office, home, ambulatory) against auscultation in each individual, but we are a BP measurement research centre and what we do is unrealistic for primary care.

Two is accurate blood pressure evaluation. Some patients take measurements when they have symptoms, others use their own measurement schedule (some take triplicate measurements, others discard the first measurement in each occasion), and eventually ‘select’ which measurements they report to their doctors. Therefore, home blood pressure monitoring is an unstandardized and biased method and, unless this is fixed, it does not appear to be a reliable method on which treatment decisions can be based. Modern technology of home blood pressure monitors can fix most if not all of these issues with minimal additional cost, by using automated memory, automated averaging, implemented home monitoring schedule according to guidelines, etc. It is our responsibility as blood pressure measurement experts to guide and press the manufacturers to develop and put on the market devices designed to meet our needs and help us improve the evaluation of blood pressure in our patients.

Great discussion on patient training and product usability so far. I will add that some of the responsibility should come from the industry side (makers of devices, softwares). For example, account reps/customer success teams should work collaboratively with the clinical care teams who prescribe/recommend these products to patients, as well as with patients themselves (in the form of offering accessible support resources) to ensure patients are trained on how to use the products in a remote setting. In addition, a focus on patient-centered design is also important to optimize usability of products in the intended settings. These approaches may be particularly relevant in the context of emerging SaMD products.

This is all a good discussion. I’ll go back to what I wrote a while ago above: making use of those providers who enter the home to assist in providing education and checking compliance.

I realize this is principally a forum on best practices of home monitoring devices. Yet, I simply think a reminder as to observation bears a little bit of ink.

Precision technology should be expected. But clinical oversight whether via case management including home visits or through other clinical providers is also necessary. Some patients are compliant and want to do the right thing relative to recording their measurements and the context around these measurements. Many others, though, do not. This is not a surprise to those commenting here, I reckon. But, it is important to recognize that even the best technology will be misapplied by some and there is little that the manufacturers or even clinics or emergency departments can do about it save through home visitation and constant verification.

The challenge is very similar, in my observation, to the issue with medication and compliance at home. Ultimately, a human needs to visit with and consult with these patients to ascertain and intervene. People are complex and have many reasons for not being compliant.

Another way to ensure optimal use of home monitoring devices is to engage and educate people with whom a patient lives - family members, friends, or roommates may be underutilized. Of course, this should only be done with the patient's explicit consent.

Other devices that capture patient-generated health data also automate the transmission of readings that may be concerning. For example, implantable cardioverter-defibrillators have mechanisms to automatically transmit data, if the patient is set up for remote monitoring, after a patient receives multiple shocks. Of course, this also means that we must be prepared to receive and act on the data.

For those patients who are motivated and have supportive people around them, all of these ideas are helpful: enlisting family members; remote monitoring services; improved patient education; education of the primary care and emergency staff; in-servicing of the home health aids, family members, neighbors and others.

Technologies that can “phone home” to provide remote monitoring measurements are also useful to a point, but the issues of false alarms and non-clinically-actionable alarms will result, which can become overwhelming if not managed and tuned properly. As many know, false alarms are a significant issue for clinical staff on the inpatient side of continuous monitoring.

Non-compliant patients are a more difficult problem to solve. Patients may stop taking their medications and may cease their monitoring for many reasons, even those who are visited regularly by home health agencies and case managers. Patients who are depressed and feel no desire to live are an example of this cohort and require interventions which are completely unrelated to technological monitoring accuracy. Unfortunately, it is the non-compliant patient who tends to be the long-pole in terms of emergencies, and these are the costliest of patients.

Holistic oversight which enlists multiple methods are needed: monitoring technologies to obtain accurate and objective measurements in order to monitor patients with chronic ailments; caring individuals who can and will check on patients regularly to ensure they are not alone and provide an ear to the lonely and shut-in; local programs to assist the elderly and home-bound with basic issues, from monitoring to grocery shopping to getting to see their care providers when needed; and enlisting volunteers to help fill the gaps and ensure these people are not forgotten.

No easy way to enable best practices, but one of the first steps should be ensuring that any medical device, be it for home monitoring or home testing, is only used for the approved indication or clinical situation. That will ensure that the data generated by the device are most likely to be reliable and valid.

Validation is CRITICAL for ALL devices which estimate BP. Why would anyone want a device which has not been fully validated estimating BP on anyone, anywhere?

I agree that BP devices for home should be validated. Should they also carry a reminder on how often they should be calibrated? Maybe patients should be encouraged to bring them to clinic appointments along with their medication. A kind of medication reconciliation for devices.

Automated oscillometric devices cannot be easily calibrated; their algorithm is a computer program. The manufacturer will tell you how often the device should be serviced. The only way you can check the accuracy is to do a manual auscultation before and after a device reading......they should agree within 5mmHg. The problem with that is office BP IS DONE SO BADLY that this may be more inaccurate.

It seems that there is general acceptance that home-based medical devices should receive oversight for use. I agree with this sentiment. Yet, it is not merely in terms of use that oversight needs to be provided: it is also in terms of context. When I take a patient blood pressure using an automated monitor, I will also occasionally check the measurement using manual auscultation. I trust my ears and my Littmann stethoscope more often than not as many automated measurements obtained from elderly patients (those who are particularly frail) I have had are not reliable. There are contextual aspects of the measurements (regular, thready, strong, etc.) that are not visible through the automated measurements, and depending on the strength and health of the patient, some of these automated measurements are just wrong. Furthermore, general patient health and presentation through observation is an important indicator that needs to provide context over an assessment.

Following up on John's point and CMS's recent push for telemedicine, might it be possible to do such supervised measurement observation via remote video conferencing (at least on occasion and in addition to F2F observation)?

I think this would yield much more accurate and reliable results. Thank you, Stephen.

As with any device, it depends on your goal. If you are looking at a diagnostic device to accurately assess and confirm a medical condition, then yes, you need to have accuracy. However, if this is a screening tool and you are willing to accept a false positive and false negative rate, then you should know that rate and be willing to accept the errors. I think the biggest challenge will be determining if the device or software is meant to be diagnostic/therapeutic vs. screening/monitoring for trends.

Your observation about the challenges of BP measurement in the elderly is significant. Broader awareness of this, especially in the tech community would be helpful.

The devices which are said to be only screening/monitoring........if they are not validated and give false readings, then how will you tell what to do? You need to know how the screening device compares to the diagnostic device. If there is a variable 5-10mmHg difference, how will you know what to do? It might be even worse than that......then what?

I concur. The original thread is regarding accuracy of medical devices...I believe that ALL devices must be accurate (+/- % from gold standard and validated to insure they are measuring what they say they are measuring). A BP cuff or novel new device in a sterile environment simply measuring mmHg on a mannequin may be exceedingly accurate, however when let loose on the real world in a medical clinic let alone the general public is fraught with IMPLEMENTATION concerns. I think we have to break these factors down and take them into account individually (Accuracy, Implementation in context, Validity within a given context). Although I am not an expert in BP assessment and the new devices, I think the accuracy of the devices may be less in question than the IMPLEMENTATION and subsequent VALIDITY. I believe Validation is the culmination of the accuracy of the device employed implemented in a real world population. Similar to research being driven into practice. The research is accurate and the measurements are valid for that particular time, space and population but IMPLEMENTATION in another context is fraught with VALIDITY to which you allude.

As a further question, how accurate and valid are the stand alone BP sitting stations you see at the pharmacy. Those have been around for ages....

Concur with the physicians. If I am to consider administering nitroglycerin to a patient experiencing chest pain, I need to know that the patient is not borderline hypotensive (systolic not less than 90 mmHg). Furthermore, if I am to consider a follow-up nitroglycerin, I need to make sure that the blood pressure has not dropped by more than 20 mmHg before and after the first administration. There is accuracy implied here (as well as precision). But, as I pointed out in my original post, I rely principally on my own auscultations (or check machine auscultations) to verify.

Finally, a real concern with certain elderly patients is accuracy, particularly with frail patients. Sometimes automatic monitors that cannot make determinations very easily on patients with weak pulses will hyper inflate and hurt or injure the patient. This is not such a big problem in young adults, but it is a real concern with the frail. For this reason alone I will auscultate manually in order to save the patient unnecessary pain.

In both of these instances, I rely on the accuracy of the measurements to perform interventions as well as trending. Casually monitoring one's blood pressure trends on a daily basis does not require this level of accuracy if one is simply using this to obtain a general idea of evolution or gross changes. But, if the direction the industry is moving is one which relies more heavily on automation and artificial intelligence, then (in my opinion), technologists should be made aware of the medical ramifications and potential safety risks and necessary limitations associated with poor or questionable accuracy.

So, would I be correct in saying that:
1. The calibration of a device used for diagnostic/therapeutic purposes should always be confirmed against the physician's device during an in-office visit when the remote monitoring begins.
2. This in-office observation should be done periodically [how often and does it depend on the device?]
3. Remote video conferencing done occasionally (if available) might be a useful way to observe the patient's method of taking the measure.

Of the kiosk devices in Pharmacies, as far as I can tell, the only VALIDATED KIOSK is made by PharmaSmart. I have tried to obtain the validation data for 6 other major brands but have been ignored, again in violation of Federal Law.

One important issue which has NOT been highlighted, and maybe is not widely known is that VALIDATED DEVICES ARE ACCURATE FOR A POPULATION. There is NO GUARANTEE that any device will give precise data on an INDIVIDUAL PATIENT. The only device currently available to do individual validation is the iPhone app called the Accutension Stetho. It does automated auscultation (the GOLD STANDARD) and compares the device reading to the TRUE auscultatory value. It is inexpensive and easy to use. Every pharmacy, nurse, and physician needs to use one on every patient to insure that a specific device works on a specific patient.

Having some interest in validating a medical device your expertise and insight is very valuable.

Sorry, but there is NO GUARANTEE that your fancy expensive office device is any more accurate on Mr Jones than his inexpensive home device. Every patient is different; every device algorithm is different. The only way to assess accuracy ON ANY DEVICE is to compare by research quality auscultation or use the Accutension Stetho instead. If your office staff can do research grade testing, following all the AHA rules, etc, then use their expertise.

These are all important issues but let’s not forget Fitbit and other devices that people rely on for health information but that are not regulated. FDA decided that their involvement would interfere with innovation. But what if some of these personal devices provide misleading information?

(Message 1 of 2)
...apologies for the length but the topic is fairly complex :)

Jason, I wanted to reply directly to your question about the accuracy of public-use BP kiosks (many located in community pharmacy). I have been working in this area most of my career, it’s true the modality remains largely misunderstood. For the average physician who does not have time to pour over the research, it can be confusing.

First, as Dr. Alpert discovered and documented in his 2012 published survey, there are multiple manufacturers of such devices, and they are not created equal. pdfs.semanticscholar.org/f604/...

A public-use kiosk may use the same oscillometric technology as well-validated professional clinic devices and has the potential to offer the same level of accuracy. However, the primary technical challenge with a kiosk device is the cuff. Because there is no option to choose a cuff size according to the patient arm-circumference, traditional kiosk devices risk providing false readings for patients with large and small arms.

My company, PharmaSmart, took on the challenge of the wide-range cuff. We patented a unique design that has enabled us to prove validity across a wide-range of adult arm sizes. As Dr. Alpert stated, we remain the only public-use kiosk manufacturer supported by independent, published, clinical validation. To date, it has proven impossible to obtain clinical data from the other kiosk manufacturers.

(Message 2 follows)

(Message 2 of 2)

Many clinicians also wonder (understandably) if readings taken in a pharmacy, even if the device is validated like ours, may be erroneously high due to the environment or methodology. Dr. Raj Padwal (also on this discussion thread) and his team took on this question in a clinical trial that compared serialized BP measurement on PharmaSmart kiosks to daytime ABPM and serialized automated office BP. The PharmaSmart results closely approximated daytime ABPM and were similar to automated office BP, “supporting the use of serial readings from (PharmaSmart) in the assessment of BP.” ashjournal.com/article/S1933-1...

PharmaSmart is proud to be participating in the PCORI funded BP-CHECK study which is a 3-year randomized controlled trial comparing ABPM, Home BP, and validated BP Kiosk (PharmaSmart), for both clinical validity, and patient preference standpoints. We feel this type of research into out-of-office measurement modalities is critical if we are to optimize care models for blood pressure control. pcori.org/research-results/201...

Finally, it is important to consider that modern, internet enabled, clinically-validated blood pressure kiosks offer unique benefits to patients looking for an out-of-office measurement solution, including: zero cost to the patient; a standardized cuff that eliminates errors from mis-cuffing; longitudinal serialized data capture and integration with clinical records; on-device video education and support; and engagement of the community pharmacist supporting emerging collaborative care models.

Thanks for your interest in this area. It’s great to see such a vibrant discussion about device validity and the practical implications of evolving blood pressure technology in healthcare.

(end)

Heart rate on fitness trackers are unfortunately prone to give false readings since they depend somewhat on how they are worn. Now there is a new study of the Apple Watch and arrhythmia. Will see how accurate that proves to be sciencedirect.com/science/arti...

heart rate of minor significance relating to morbidity and mortality in comparison to BP. I can measure my HR easily by palpation; don't need to spend any $. can't do BP by myself

Exactly. I do HR by palpation and still use my second-hand watch to time for the calculation. Everyone around me tells me I need to get an Apple Watch so that I can have accurate HR measurement, but I cannot seem to justify the expense.

Gray hair is associated with wisdom, as is thinning hair. The young have to have a e-object to do their thinking for them, the higher the price the better.

When you are at 250 watts on an uphill I try to keep my hands on the drops and off my radial pulse. I prefer the hear rate strap for accuracy. :-). Wrist watch suffices for travel but heart rate strap during competitions.

Yep. I use the H10 polar for exercise.

I am eager to learn what the findings of the pilot program will be and what guidelines they will drive. It was encouraging to see in the initial working model that the pre-cert review process is dependent on the intended use risk category. The risk categorization definitions may need to be refined over time but I am encouraged that the FDA recognized the need to be more agile in software regulation.

Concur with Dr. Sheng.

At the Connected Health Conference, there was a panel of Pre-cert participants. In addition to talking about their decisions to join the program, they discussed the beginning of an audit process which the FDA is doing. mobihealthnews.com/content/fda...

I have not tried personally but am not surprised by your experience. This brings into discussion the role of the FDA as related to enforcing manufacturer compliance and/or being the centralized provider of this data.

I have tried personally and to no avail. The C-suite talking points indicate that they want to share information with clinicians but there doesn't seem to be people or process in place at the manufacturers to make this happen. One manufacturer customer service rep hung up on me when I asked, and the other told me to look on the internet!! A major disconnect for sure. Still working to solve.