Many systems have implemented audit and feedback type programs. In EPIC here at Duke we have provider dashboards which show providers how their patients are doing on a number of metrics overall. This is probably most effective when tied to a pay for performance type program for providers. The other challenge is that the meta-data aren't so useful when a patient sees the provider. This is where patient specific alerts or decision support tools can come in to help alert providers to individual patients who are out of range.

To add to the above, decision support tools should not only alert to out of range values, but provide direction and support to take appropriate action based on the individual patient. This is the real value of CDS.

Polypharmacy is a huge challenge today, and needing to add additional BP meds to the mix will not help. We have good data for using certain classs of anti-hypertensives (e. g. beta blockers for CVD, ACEs for diabetics, etc), but after the first one or two drugs, we don’t have as many studies to advise us. While additional research is in the works, BP management is still very much an art. People respond differently to certain agents and he only way to recognize this is with good communication and close follow-up.

Virtually all EMRs allow you to enter comments regarding current or historic medications, which would be one way to document problems with certain drugs. If it is a basic EMR system with limited options, one could use the allergy section with careful attention to clearly describe the side effects (as it is rarely a true allergy). Regardless, there are ways to update patient reactions to medications. Furthermore, it is an important component of their medical care so that we avoid repeating similar mistakes again. I always implore my patients to carry an updated medication list that also includes drugs that they had been started on previously that were stopped and medications that the patient had a problem taking previously. This is incredibly useful knowledge for ostientsto be able to share with other healthcare providers.

Polypharmacy does sound like a challenge! Dr.Bonnet, you provided us a good picture of how things are done today. Thanks for that! It is interesting that you brought up using the allergies section to document patient reactions to trial-and-error combinations of past medications, which were not quite allergies but close... Is this the unsaid standard way to document patient adverse reactions to medications today?

Also, I wonder - what should things look like in the ideal world?

If someone visits an Emergency Room while traveling out of town, my guess is that the patient cannot readily access their own medication adverse effects history electronically? What essential data fields do we need to show the out-of-town emergency physician what this patient's circumstances were in terms of medication history? What I can think of on my feet include time-stamps, diagnosis events, co-morbidity and acuity, medication trial-and-error results related to evolving co-morbidity, key lab results...The list goes on, then you have information overload...

What would you say are the 2 or 3 most important information which the new doctor should know? Is there a standard way to highlight and update such information, so that the new doctor would be able to better evaluate trade-offs of certain new treatment?

Would clinicians rank patients' tolerance or adverse reactions to various medications in their notes tagged with the top 3 factors to explain such ranking? Will such "rankings" change in the future? Would you say that the direction of changes are purely random, or could be somewhat predictable? Is anything mentioned above nice-to-haves or not-reallys in the between-provider data flows / communication?

Lots of interesting things to think about! While my response is by no means ideal, I have seen providers and patients document allergies as such. In the future, I would imagine that all of this data would be captured on a smart phone/device. For an ER, a list of medical conditions, previous surgeries, active medications, and allergies would be my top four pieces of information for patients to carry with them. Primary care physicians would benefit from additional details, such as previous intolerances/side effects to medications.

While I cannot predict the future, I think that there will be a shift toward patients having acces to, and carrying more of their own medical record with them. This solves the major barrier of poor communication between different health systems and clinics. The patient can carry most relevant information with them at all times to have it available to whoever might need to see it. It will be interesting to see what happens in the future!

Since many patients have smartphones and most providers have EHRs, it is possible today for patients to request their clinical summary from their provider and carry the summary on their smartphone. If the provider offers a patient portal, the ability for the patient to download that summary (as a file) is usually a feature. That file can be viewed on smartphones with viewer apps. iOS phones have this built in since iOS 10. "iOS 10 takes Health a step further by now allowing users to store their health records directly in the app using the Health Level 7 Continuity of Care Document (HL7 CCD) standard. No longer having to track paperwork or use CDs, these health records will be tracked directly in-app with the ability to be imported from Mail, Safari, and other applications." The iBlueButton app is another way for patients to request the summary from providers, especially those who don't provide patient portals.

As our lives become more integrated with technology, home monitoring of multiple metrics (e.g. weight, BP, glucose, etc) is inevitable. We are not from having all of this data appear in the EHRs, which is both good and bad. On one hand, more data generally provides a more accurate picture of how a patient lives day-to-day. Conversely, it will quickly turn into a massive data dump that will require improved ways of filtering and summarizing the information, as well as innovative ways to communicate with patients.

The primary barrier that we currently face is a lack of reimbursement and resources to effectively and efficiently examine all of this data. Right now, combing through this data is all ”extra,” unpaid work for someone or something. Computers could be highly effective at this task but at some point it will require some human oversight/intervention.

Another interesting consideration is that the patients who are most inclined to use mobile health technologies are probably the ones who would benefit from it the least. I was part of a small pilot study at Duke that found just that—healthy people used a higher number of mobile health devices, and also more frequently, than patients with chronic diseases. ncbi.nlm.nih.gov/pmc/articles/... It will be important to identify ways to engage these patients and find meaningful ways to leverage technology to improve health care delivery to them.

The new blood pressure guideline clearly emphasizes using home blood pressures to monitor patient response to therapy. Fortunately, it appears that payment systems may actually be changing to support this. CMS just "unbundled" CPT code 99091 which allows physicians to bill for time spent reviewing home BP results (and other home monitoring data).

EHR solutions can help make this easier - including building dashboards to review/evaluate home BPs either entered by patients or uploaded manually.

See: cms.gov/Newsroom/MediaReleaseD...

I will agree with Jonathan's comment that we will need clinical workflows in place to consume and act on the patient-reported data. Many EHRs can already receive this information, but it is not clear where it should be stored, who acts on it, what the clinician responsibility is related to abnormal readings, etc. Data fidelity is also an issue - knowing what the patient was doing immediately before taking the reading, how they took it, if their home machine is appropriately calibrated, etc. will contribute to the accuracy of the reading.

To me, home BP involves patient education and communication, which is all about data collection between doctors' visits and producing relevant insights, alerts or recommendations at the right time to the right people, which not only include clinicians but also maybe patients and their caregivers at home.

Dr. Bonnet - You brought up many great points. I am hearing data dump concerns, patient education gaps, provider-patient interaction model redesign, financial incentives, mobile health bringing some disparity across individuals of various socio-economic factors. Very interesting! Many of these deserve to be its own separate topic for more in-depth discussion sometime.

Dr. Navar - Thanks for the good news regarding changes in coding and reimbursement. Right now there is no standard way to access home BP readings on the cloud, and to recommend actions for both patients and nurses. You mentioned dashboards, which are visual decision support tools. What are your thoughts on who should be able to access and use the dashboards? Is it mostly for the clinicians? Or the patients and their families or guardians?

Dr. Mishuris - You seem to be very passionate about discussions around redesigning clinical workflows, the implementations and how great ideas actually would get done in the real world...Important stuff! Would you like to elaborate on those a bit more?

I would add: med.stanford.edu/myheartcounts...

Reviewing the hypertension guidelines and research program, as above, presents the context for my personal view. "SNOMED-CT" and "LOINC" may be used for designing machine learning algorithms in an EHR. These are ideal for uploading information from large data warehouses based on clinical practice guidelines.

A good background on the issue: enews.seas.virginia.edu/machin...

Thanks for taking the time to read this!

Some systems will allow for an app or lifestyle prescription, yes. It may be in the form of patient education materials rather than an actual prescription though. Every practice may have a different operational model for engaging and monitoring patients' progress with lifestyle interventions, aside from just the end outcome of improved blood pressure. I have led the development of an integration between an online lifestyle modification program and the EHR - it should be piloted in a clinical setting though to ensure workflow alignment, identify facilitators and barriers to both the clinician and patient, and track outcome metrics.

I cannot overemphasize the importance of attempting intensive lifestyle behavior change before jumping to medications. Many physicians innately assume that patients will be unsuccessful in eating better or moving more, and frequently skip this step altogether! We could not be doing our patients a bigger disservice! Trials such as the D.A.S.H. Diet demonstrate the profound impact that diet can have on disease. The challenge with prescribing lifestyle as medicine is that it takes time and is poorly reimbursed. Applications will continue to make this process simpler but invariably the human connection is still tantamount to effective long-term change. In terms of tracking, Kaiser Permanente and many other health systems throughout the country have implemented Exercise as a vital sign (# minutes per week). This is a crude and oversimplified way to track but is a start. Utilizing health coaches is a relatively inexpensive way to help manage many patients in this process. I would suggest that lifestyle behavior documentation lies as a separate and distinct section in the social history. It does not need to be complex--a simple template could ask, on average: how many hours of sleep do you get per night, how many minutes of moderate-vigorous intensity exercise do you get each week, how many servings of fruits and vegetables do you get each day, do you feel capable of dealing with the amount of stress in your life, do you have at least 2 people you could call in a crisis, do you feel like your life has purpose/meaning? Those 6 questions essentially get you the 6 pillars of lifestyle medicine--what David Katz has coined: feet, fork, fingers, sleep, stress, and love. Approximately 80% of chronic disease and premature death could be prevented by not smoking, being physically active, and adhering to a healthful dietary pattern. Lifestyle is medicine and should be used as such.

Health systems may opt to run a query to identify the new patients considered to have hypertension under the new guidelines in order to assess the impact of the guidelines on their patient population. This effort could help determine the volume of patients impacted and assist in strategizing for the next step (e.g., patient outreach). However, a simple query of blood pressure values will likely overestimate the problem, as a portion of these patients may already be receiving guideline-based care. While it is easy to track prescriptions of anti-hypertensive medications in the EHR, few health systems have implemented a systematic way of tracking the lifestyle recommendations. (e.g., sodium restrictions and physical therapy.) In addition, In most health systems, we will only have the last clinic-based blood pressure value in the EHR. This value will not reflect improvements in blood pressure that may have resulted from a recent prescription or recommendation for diet and physical activity. Thus, sending letters should be done with caution, as it might unnecessarily alarm patients who are actually under control as a result of recent care and/or are already receiving appropriate care. As with all EHR-based algorithms, there are strengths and limitations that must be considered carefully when communicating with patients. One way to improve our ability to use the EHR for ensuring that patients are receiving guideline-based care is to start to systematically document the lifestyle recommendations that are part of the new hypertension guidelines as well as guidelines for diabetes and other conditions.

Correction: "physical therapy" should have been "physical activity."

As Dr. Hirsch mentioned, most major health systems have an EHR that allows for population health queries, such as this. Filters would enable one to identify these individuals (i.e. searching for patients who do not currently have a diagnosis of hypertension but have BP values that are consistent with that diagnosis now). The accuracy of these results would depend on how up-to-date patients' medical records are (i.e. having all current diagnoses listed). Assuming this was the case, one would be able to identify these "new patients with hypertension." The next question then becomes what to do with the people you identify. Should we bring all of these people in for a visit to check and discuss blood pressure or would you be okay with waiting for their annual exam? This depends on how aggressive one would like to be. Another alternative is to send these individuals a letter in the mail, notifying them of the new blood pressure guidelines and asking them to check their own BP to see if they are considered high. This empowers the patient to take control of their own health and see the physician on their own accord when they are ready. The reality is that it will take time before all of these "new patients with hypertension" will be identified. For the majority of these patients, I think that waiting until their annual exam is not unreasonable to me. We have allowed these patients to have their current BP values for years. While it is an important matter to address as the new guidelines suggest, I do not see this as an acute emergency that must be corrected immediately. I think we have some time to identify these patients and get them in when it is convenient for them. The bigger challenge, I think, will be getting physicians to aggressively treat these lower, now considered elevated. I think there is going to be a dramatic increase in hypotensive side effects. Like anything else, it's a risk/benefit analysis that we have to perform on every patient.

There are risk calculators available in our EHR (Epic) - ASCVD risk, for example. The issue is not the absolute value of the calculated risk, but what actions are triggered/taken as a result based on that value and established cut-offs for interventions. Any decision support that triggers off the result of these calculators will need to be reviewed and updated based on new guidelines. Many organizations will have pre-established processes for review of decision support in their EHR - new guidelines is one potential trigger for review.

Agree with Dr. Mishuris's response. I have attached a screenshot taken directly from the Clinical Practice Guidelines report. Our practice uses the ASCVD online tool that is cited in the paper to estimate risk (tools.acc.org/ASCVD-Risk-Estim...). Unfortunately, in my current practice, the calculator does not link directly to our EMR so the data must be manually imported. I previously used EPIC, and it is possible to build a dot phrase that imports all of the necessary values. Some versions have the tool built in as well to provide estimated risk. In an ideal world, this process would happen seamlessly and the risk would automatically be calculated and viewable during every encounter. This would allow one to see how the risk changes over time (due changes in systolic BP, cholesterol levels, age, etc.) so that treatment can be adjusted accordingly.

Dr. Mishuris & Dr. Bonnet, glad to hear your perspectives and to learn about the ASCVD online tool! So I learned that new BP guidelines may present an opportunity to review current intervention strategies for at-risk populations. Risk score calculation may not be readily available if the data on risk factors is not yet directly integrated into EHR systems. It'd be interesting to monitor the low-risk population's risk factors over time to detect any deviations or changes. There are many other important risk factors for heart disease besides blood pressure. Can you share some thoughts on the data implication of the new hypertensive guideline on the importance/awareness or ranking of other risk factors? Do you expect frontline clinicians such as nurse practitioners put more weights on hypertension when screening patients?

Thanks for the question, Cami. Depending on how you look at high blood pressure, it both has risk factors associated with its development and is itself, a risk factor (i.e. in developing CAD or a CVA). The multifactorial nature of most of these types of diseases warrants careful attention to the variety of risk factors that contribute to their development. There are studies that estimate the different odds of developing heart disease or stroke with various risk factors. There are too many variables to include here, but one could easily find and assess the contribution of a single risk factor for a single disease by doing a brief literature search. Regarding the second aspect of your question, I wouldn't expect a significant change in terms of how blood pressure is viewed as a risk factor immediately. It will take time and additional data for us to recognize how impactful the new guidelines are. If the results are promising, I would expect an analogous increase in the amount of attention being paid to blood pressure.