Panel 4: Applying Quality Framework to Asynchronous Use Cases

July 12, 2023

ID10125

To CiteDCA Citation Guide

00:00Our last panel today,
00:02which we'll now use to wrap up
00:04the session and then I'll make
00:06just a few closing remarks to,
00:08to close things out,
00:10is about applying Quality framework
00:12now to asynchronous use cases.
00:14And if I have the next slide,
00:18so this panel is going to be moderated
00:20by Doctor Elizabeth Krabinski.
00:23Who is Professor and Vice Chair
00:24for Research in the Department of
00:26Radiology and Imaging Sciences at the
00:28Emory University School of Medicine.
00:29I have to say I I met Elizabeth
00:32through this work process that we
00:34built for the first symposium.
00:35We hadn't met before,
00:37but it's been so delightful getting
00:39to know her and she has such an.
00:40Interesting perspective on this work
00:42and has been so actively engaged
00:44in it from a variety of different
00:47perspectives and has really helped
00:49me to think about the frameworks
00:51around competency in a different way.
00:53So I'm really delighted to have
00:54her here and have her help lead
00:56the the panel for panelists.
00:58So with that, Elizabeth,
00:59I'll turn it over
01:00to you. Great. Thank you so much and
01:03I am accompanied on this panel by two.
01:07Wonderful colleagues,
01:08we have Marcy Bolster,
01:11who is at Mass General Hospital
01:13and Harvard Medical School.
01:14She is director of Rheumatology
01:16Fellowship Training Program at the Mass
01:18General and is associate professor of
01:20medicine at the Harvard Medical School.
01:22She's director, as I said,
01:24of the fellowship training program,
01:26and she's been a Rheumatology
01:29Fellowship program director since 1999
01:31at the Medical University of South
01:34Carolina and Mass General since 2012.
01:36She's committed to facing gaps
01:38in the rheumatology.
01:39Workforces is recipient of a
01:42Rheumatology Research Foundation
01:43Clinician Scholar education grant
01:46entitled Extending Our Reach,
01:48which provides telemedicine training to
01:50rheumatology fellows to provide care
01:53to underserved patient populations.
01:55She received her MD from Duke University
01:57School of Medicine and our third
02:00panelist is Laura Kristofferson,
02:02who is from the Mayo Clinic.
02:04And she is Principal operations Analyst at
02:06the Center for Digital Health at the Mayo.
02:09She has earned her bachelor's degree
02:11from the University of Minnesota,
02:13Twin Cities,
02:14her MB A from Augustburg College
02:17and she will be completing,
02:19although it looks like she completed
02:21her Doctor of Education from
02:23Hamline University in 2022 so.
02:27This is going to be a very interesting panel.
02:29I think a little bit different from
02:31the the ones we've had so far in
02:34that it's all about asynchronous,
02:35which a lot of people don't think of
02:38when they think about telemedicine,
02:39but it certainly is out there
02:41and has been for quite a while.
02:44So we're going to tack this from
02:46three different perspectives
02:47and I'm going to go first.
02:48So if we go to the next slide.
02:50We do have a solid set of objectives.
02:53So first,
02:54recognizing the role that data
02:56and image quality have as a
02:58prerequisite to quality outcomes.
03:00Our second talk by Laura is going
03:03to be demonstrating the use of a
03:05framework for assessing the value
03:06of a digital care plan to from
03:09ideation to outcomes measurements.
03:11And then Marcy's going to talk about
03:14ensuring access to specialty care,
03:16maintaining patient safety.
03:18And recognizing care acuity needs
03:20through the use of Econsults.
03:22So next slide, so as I said,
03:26I'm going to give the first talk
03:27on image quality as a prerequisite.
03:29Next slide,
03:32you know it's asynchronous is all
03:35about data and that can be diagnosis
03:39for treatment and it could either be
03:42images or. Data, files, words and so on.
03:46But it's like I said,
03:47very different from what you typically
03:49experience in a real time encounter.
03:52By nature, it is dissociated in time,
03:55so you don't always have that
03:56ability to reach out and say,
03:57well, what did you mean by that?
03:59Now there's a huge body of
04:03literature on the classic.
04:05Asynchronous modalities, radiology,
04:07pathology, ophthalmology,
04:09dermatology, and so on.
04:11Showing that there's a clear,
04:13very clear relationship between
04:15the quality of the data that's
04:17utilized in that that encounter,
04:20that asynchronous encounter,
04:21the data quality on the diagnostic decision
04:25that's rendered, whether it has
04:27to do with finding something on an
04:29image or taking that information
04:30and deciding on what the treatment
04:32or the next steps are going to be.
04:34Therefore, patient outcomes and patient
04:37care are significantly impacted as well.
04:40Now there's an entire chain of events,
04:44much like there is in a real time encounter,
04:46ensuring quality from the point
04:48the patient enters the system to
04:50the end of the encounter.
04:51Sort of the same thing in in the the
04:54purely digital world of data and images,
04:56you've got an acquisition process,
04:59you've got to transfer that information from
05:01one place to another without data loss.
05:04And then you've got to display that
05:06information in an optimized fashion
05:08to the person who has to look at
05:10that data and cognitively process it
05:12and render some sort of decision.
05:13And then you've got to have some storage
05:16mechanism that you can store the data
05:18but then bring it up again should you
05:20need it for a a future contribution
05:22to your your diagnosis it you know,
05:25a follow up visit or something.
05:27Now there's a lot of professional societies
05:30out there who have developed guidelines.
05:33Practice guidelines,
05:35some even standards for utilizing,
05:38storing forward asynchronous type data.
05:41A lot of them having to do with
05:43quality assurance, quality control,
05:45and quality improvements.
05:47The problem is,
05:48is that there's no single one-size-fits-all.
05:51What you need for radiology is not
05:54necessarily what you need for pathology,
05:56ophthalmology or dermatology.
05:57A lot of common principles.
06:00But it's not gonna fit every
06:02single application in the same way.
06:03And it all starts with acquisition.
06:07And that's why it can get so complicated
06:09is because of all of our acquisition
06:11devices are very, very different.
06:13And in a sense, you need taskers.
06:15I mean,
06:15you can't do a dermatology visit with an MRI.
06:19And you can't do ophthalmology
06:21visit with an X-ray and pathology
06:24requires you know the the,
06:26the raw data from the biopsies and so on.
06:28So the acquisition is really very
06:30specialized in that it impacts the quality
06:33of the information from that point.
06:36On the display,
06:37a lot of people don't even think of
06:39this where when you sit down in front
06:42of the display now in radiology we've
06:45thought about it for almost 30 years.
06:48But in telemedicine, you know,
06:50I can just sit down in front of my laptop.
06:52Now I'll look at that on my phone.
06:54The problem is,
06:55all of these display devices are
06:58not the same and you really have to
07:00consider what is the best way to
07:03display that information to the human being,
07:06their and their perceptual and
07:08and brain systems in a sense.
07:11However,
07:11there is one guideline out there from
07:13the American College of Radiology,
07:15the American Association of
07:16Physicists and Medicine,
07:17and the Society of Imaging
07:19Informatics and Medicine.
07:20That really provides,
07:21in my opinion,
07:22sort of a broad overview of what
07:25display characteristics are important,
07:28no matter what your clinical
07:29specialty or the type of
07:31information, whether it's data,
07:33images and so on, that are required.
07:35And that's a technical standard for
07:36the electronic practice of medicine.
07:38So I encourage people to look at
07:40that when you're considering the
07:42quality of the data as you display
07:44it to the person interpreting it.
07:46Technical standards also exist.
07:48DICOM was developed originally in radiology,
07:51but it's spread to pathology,
07:52to dermatology, ophthalmology and
07:54so on in terms of getting standards
07:58in place for data acquisition,
08:00for data transfer storage and most
08:02importantly for that display of the
08:05information to the diagnostician.
08:07There's also a group,
08:08the International Color Consortium
08:10on Medical Imaging Working Group.
08:13Who is very much concerned with
08:16color fidelity from the point of
08:19acquisition to the point of display.
08:21Again, if you look at you know pathology,
08:23images, dermatology and ophthalmology,
08:26these are color images.
08:28And so the acquisition phase
08:30has to have good fidelity,
08:33good image quality that matches what
08:35that object is that you're imaging.
08:38And then you have to send that all the
08:40way through the train so that the image
08:43you see on your computer display really
08:45does match that original object as well.
08:48Diagnosis often relies on that color quality,
08:52so we have to incorporate it.
08:53There's also some very easy
08:55tools that exist to help you,
08:56such as the color checker,
08:58color rendition,
08:59tart or what we call the Macbeth chart.
09:01So there is a lot of information
09:03out there for the quality control
09:05and asynchronous telemedicine.
09:06Next slide.
09:08This is a really great tool that kind of
09:12exemplifies what I've been talking about.
09:13It's the American Telemedicine Association's
09:16Quick guide for Teledermatology.
09:19And it's it's there's a whole
09:20other document that goes with it,
09:22but it really lays out how
09:25you can help ensure.
09:27Image quality from the point of
09:29acquisition to the point of display again,
09:32and it talks about the camera.
09:34What do you need?
09:35What's the minimum resolution?
09:37How many megapixels will you need?
09:39You know what types of modes do you
09:41need a macro mode? What about flash?
09:43And so on. It goes through specifics
09:45on how to take the image.
09:47Do you look orthogonally?
09:49Should you look at an angle?
09:51You know? How do you point and shoot?
09:52How far away? Should you put the camera?
09:55Should you use flash or not?
09:58What about compression?
09:59All of these images,
10:01whether or it's it's dermatology,
10:03otalmology and so on, can use compression.
10:06But how much should be allowed
10:0820 to 110 to 1? Reversible.
10:10Irreversible.
10:11Lossy.
10:11Not lossy.
10:13All of those are important questions
10:15that affect the final quality
10:16of the image and hence again,
10:18that diagnostic decision.
10:19And what I love about this document
10:22is that it really shows all these
10:25different views and imaging sets
10:26that you should consider when you're
10:29capturing the image to send them
10:32for an asynchronous teleconsult.
10:34And like I said,
10:35this was just for teledermatology.
10:37Ophthalmology has very similar
10:40standards for acquiring views.
10:41A lot of the tools that they
10:43have now just simply acquire
10:45them automatically. Next slide.
10:50So there's also a lot of artificial
10:53intelligence tools that are being
10:55developed to help with assuring.
10:57Image quality, data quality in terms of of
11:01telehealth and and healthcare in general.
11:04Some of them are for detection,
11:06for diagnosis, prediction, workflow.
11:07But there's a lot that are
11:10specifically looking at image
11:12quality across that imaging chain.
11:14And if you look at the
11:16images on on the right there,
11:18the original image on that the
11:21X-ray up there or the the the, the,
11:24the CT image was very, very noisy.
11:26And that got turned into
11:28the one on the right,
11:30which is very more pleasant image to look at.
11:33There's a lot less noise.
11:34That was all done by artificial intelligence.
11:36Same thing on the bottom,
11:38the pathology image was specifically
11:42blurred when it was acquired.
11:46And then on image B there you can see how.
11:49The right and the left version,
11:51one is very blurred and the other is not.
11:52Again,
11:53artificial intelligence helped with this
11:55deblurring to improve the image quality.
11:58And you can see obviously on both of
12:00these the clinician looking at it,
12:02whether it's the radiologist
12:03or the pathologist,
12:04would be far more impacted by a good
12:09image rather than having a noisy 1.
12:11So a I can improve all sorts of
12:14aspects of of the quality of the data
12:17available in an asynchronous visit.
12:19On the user side,
12:21it could certainly help improve the
12:22decisions and we have a lot of very
12:24fundamental seminal studies showing
12:26that image quality and data quality
12:29in all of these clinical specialties
12:32can really impact performance
12:33whether it's the at the acquisition
12:36stage transfer or or display.
12:38The danger of course is you could
12:40manipulate these images if if
12:42you would like and actually the
12:44the examples there on the left,
12:46these mammograms.
12:47Actually show how if someone's
12:50sophisticated enough,
12:51they could actually add or subtract lesions.
12:54So there is a bit of a danger sometimes.
12:57Should someone have access to the
13:00data somewhere along this data
13:02transfer from acquisition to display,
13:05they could manipulate the image to,
13:08in a sense cure the patient
13:10by removing the lesions.
13:12But again,
13:13that is something you have to look out for.
13:15So you know, my final bottom line message is,
13:18you know, garbage in,
13:19garbage out.
13:20And with asynchronous telemedicine,
13:21you really have to be careful of that
13:25and make sure that you pay attention
13:27to your quality control framework along
13:29the entire imaging chain up to the
13:32display and the decision making process.
13:34So next slide,
13:37so we're going to move on to
13:39our second speaker,
13:39Laura Kristofferson.
13:40And she's going to talk about
13:42a framework for assessing the
13:44value of digital care plan from
13:47ideation to outcomes measurement.
13:49Laura Doctor Krapinski,
13:52can you go back one more slide, please?
13:54Thank you. So for this topic,
13:56I'm going to demonstrate how we use
13:58the quality framework to assess the
14:00value of one of our digital health
14:02solutions called interactive care plans.
14:04I'm going to talk about how we
14:06leverage and use this framework from
14:08the ideation phase of a care plan
14:10through to the measurement of outcomes
14:13post implementation along with some
14:14of the other frameworks presented by
14:16my colleagues earlier in panel two.
14:18This framework was developed by
14:20our digital health teams and we
14:22adapted it for use within our remote
14:24patient monitoring program and for
14:26use with the digital care plans.
14:28Next slide.
14:32So before we jump
14:33into discussing the framework,
14:34I'd just like to provide an overview
14:37of the solution to enable you to
14:39better understand the use case and
14:41later discussion of the framework.
14:42So interactive care plans are novel
14:45digital health solution that we deliver
14:47to patients through our Mayo Clinic
14:49mobile app and that goes through
14:51the patient's smartphone or tablet.
14:52They're designed to engage and empower
14:54patients to participate in Selfcare
14:56by providing them health guidance
14:58that's delivered through the app.
15:00The care plan facilitates twoway
15:03exchange of healthcare information
15:05between the patient and the care
15:07team and care oversights provided by
15:09the primary or specialty care team.
15:12Interactive care plans are developed
15:14for patients who are clinically
15:16stable and who've received a new
15:18diagnosis or maybe are experiencing
15:20a limited medical event such as a
15:24surgical procedure or perhaps having
15:26ongoing post treatment rehabilitation.
15:29We've developed several care plans for
15:32conditions such as systolic heart failure,
15:34orthopedic surgery,
15:36and cancer survivorship.
15:39Patients are enrolled in the care plans
15:41by their primary or specialty provider,
15:44and then they receive notifications on their
15:46device when those tasks are assigned to them.
15:48These could be education and health guidance,
15:51symptom assessments,
15:52questionnaires,
15:52and requests to submit physiological data
15:55like blood pressure or heart rate readings.
15:59We do have this solution fully
16:01integrated into the EHR and data can
16:03be viewed by the care team in both the
16:06EHR and through care team dashboards.
16:08We've built decision trees behind
16:10the scenes and logic for each
16:12care plan so that if the patient
16:14generated health data are outside
16:17of our predefined parameters,
16:18patients can get education to facilitate
16:21that self management of their condition
16:24or for serious concerns we can send alert
16:27messages to the managing care team.
16:30So today,
16:30we've developed 17 care plans for use
16:32in primary and specialty practices.
16:35And as the solutions matured and
16:37our teams have gained experience
16:39with implementing the solution,
16:41we've really gained a better understanding
16:42of the value that care plans can
16:44provide to our care teams and patients.
16:46So to benefit from these new insights,
16:48we really did recognize the need to
16:50implement a new quality framework.
16:52That would enable us to assess the value
16:55of new care plan ideas and request that
16:57we receive from our medical practice.
16:591 requirement that we had of the
17:02framework going into development
17:04is that we needed to be able to
17:06produce a weighted score that would
17:08enable us to prioritize care plans
17:10and resources across our medical
17:12practice and with our teams.
17:13Next slide.
17:17So to accomplish this goal,
17:19we leverage A framework and
17:20that's what you see here.
17:22This was already developed internally
17:24and then we took it and successfully
17:27applied it within remote patient
17:29monitoring with one of our other programs,
17:31our high intensity program and then
17:33we turned and adapted it for the
17:36interactive care plan solution.
17:38So this framework is comprised of
17:40three domains that you see here,
17:42desirability, viability and technical
17:45and operational feasibility.
17:48So desirability focuses on the
17:49value of the solution provides to
17:51our patients and our practice.
17:53So the more critical or frequently
17:55reported customer pain points
17:57being solved by the care plan,
17:59the bigger the score will be
18:01in this category.
18:02And for patients,
18:03we're really looking to
18:04improve health outcomes,
18:06improve efficiency of the care we provide,
18:09detect symptoms that we can intervene early,
18:12alerting the care team and then improve
18:14the overall patient experience.
18:16And for our care teams,
18:17we're looking to decrease the care
18:19team burden in monitoring patients
18:22and improve the staff's experiencing
18:24experience in managing the patient
18:27population for that care plan.
18:30One important factor within
18:31this category is volumes,
18:33anticipated volumes for
18:35the potential care plan.
18:37So our goal is to impact as many
18:39patients as we possibly can.
18:41So we do increase the weight of the
18:43score for care plans that are be able,
18:45are able to be used in more than
18:48one practice area like a specialty
18:51and a primary care practice area.
18:53So under the category of viability,
18:56this category focuses on the
18:58business value or opportunity
18:59generated by the initiative or the
19:01care plan that we're considering.
19:03So the higher the business value or
19:05opportunity generated by the care plan,
19:06the bigger the score in this category.
19:09And this is part of the framework
19:11that enables us to consider the
19:13value that can be captured in
19:15the form of revenue generation,
19:17cost reduction and strategic alignment.
19:20So for revenue generation,
19:21we assess whether we're able.
19:23To increase our revenues through billing,
19:26for example, and under cost reductions,
19:28we'd evaluate what we might be
19:31able to save through things like
19:33reducing healthcare utilization,
19:34hospitalization,
19:35ICU admission,
19:36length of stay and some of the
19:39other examples that you see here.
19:42So plans that we anticipate
19:44will generate cost savings,
19:45particularly in multiple areas will
19:47score higher than plans that do not.
19:50And then finally,
19:51we consider where each plan can be utilized.
19:54As as well and how that fits
19:57with strategic alignment.
20:00And then finally we have the category of
20:02technical and operational feasibility.
20:04So this this category provides
20:06some criteria for us to assess the
20:08readiness of our practice teams that
20:10will be implementing the care plan,
20:12assess the technical fit between
20:14the clinical goals of the care plan
20:17and the technology that we use.
20:19And how the care plan idea aligns
20:21with best practices and standards for
20:24care plan use that we've learned over
20:26the years as we've been doing this.
20:29So this category is 1 where
20:30we've had a high
20:31amount of learning over time as
20:33the care plan product has matured.
20:35We've certainly learned about the
20:37importance of an engaged care team
20:39and how that really impacts things
20:41like patient and care team experience.
20:43One example, most care teams leverage
20:45their nursing teams for managing
20:48enrollment engagement and follow up
20:50on alerts from patients who are self
20:54entering their health information.
20:56And so nursing engagement has been
20:58a critical component to patient
21:00and care team adoption.
21:01So this is a great example of how we've
21:04leveraged our lessons learned from that
21:06experience and incorporated that into
21:08this value assessment model with the
21:10hopes of it improving our outcomes.
21:14So the next criteria,
21:16technical readiness.
21:17This is where we assess the alignment
21:18of features and functionality with the
21:21product compared to the functionality
21:22that's being requested by the practice,
21:25who has the idea for the care plan
21:27so that we can ensure that we're
21:29optimally able to support that use case
21:32with the functionality that we have.
21:33So we work with the practice to
21:36understand from their perspective
21:37what constitutes A minimal viable
21:39product for that use case.
21:42This is a critical part of the
21:44assessment just to make sure
21:46that the platform and solution
21:47is able to meet the requirements
21:50for that minimally minimally viable product.
21:53And if not, sometimes we do determine
21:55that the care plan is not a good fit if
21:57the if the functionality isn't there
22:00to meet kind of the the basic or core
22:03components of that care plan use case.
22:07And then finally, we assess the area
22:09of operational readiness and cost.
22:11So this is where we look at other factors
22:14beyond technology and practice readiness
22:16that might impact the feasibility
22:18of a care plan we're evaluating.
22:20A good example of that might be a
22:23practice who's requesting an operational
22:25workflow that may not be supported
22:28with the current functionality.
22:29We do use best practices to
22:33inform the assessment.
22:34So for example,
22:35we've learned that a best practice is to
22:38have a workflow or support in place to
22:40follow up with patients who are enrolled
22:42in the care plan but aren't engaging.
22:44And so to assess operational
22:46readiness with the practice,
22:48we would discuss the workflow and
22:50ensure that they had the resources
22:52in place to support that workflow.
22:54And then finally,
22:55we'd consider the cost of
22:56developing the care plan,
22:58including expenses and then ultimate
23:01reimbursement if applicable.
23:03Next slide.
23:08So once we've assessed
23:09the value of the care plan and
23:11we've made a decision to proceed,
23:13we then leverage what was identified
23:15within the values framework and we
23:17crosswalk that to metrics or data that
23:20would enable us to measure those criteria.
23:22And this is what forms,
23:24what we call the remeasurement plan
23:26that we use with the practice once
23:28they go live with the care plan.
23:30And that's what we use to
23:32assess what the success of.
23:34The implementation and clinical
23:35use of the care plan is so for the
23:39interactive care plan solution,
23:40we have a standard set of metrics that we
23:43typically apply to most every remeasurement.
23:46Plan unless they're not applicable and
23:48these metrics do tie back to several
23:51of the criteria in the framework.
23:53So for example,
23:54for every care plan,
23:55we're always looking at patient adoption.
23:58How many patients are enrolled relative
24:00to those that are eligible and how
24:02many patients are maybe saying no
24:04and how many patients are opting in?
24:07What is patient compliance and
24:09what is patient engagement with
24:11the care plan throughout the the
24:12course of of the care plan,
24:14however long it is.
24:16And then we want to understand
24:18care teams impacts in terms of the
24:20escalations that are coming to them
24:23for patient non compliance and from
24:25maybe calls and secure messages
24:27that the the patients are sending
24:29to the care team for care plan
24:31related questions and requests.
24:33And then we do use surveys and interviews
24:35to understand the patient and care
24:37team experience and satisfaction as well.
24:41And then because each care plan
24:43is different and involves a
24:45different health condition,
24:47the clinical outcomes vary
24:48widely across care plans.
24:49So they're not always the same
24:52exact clinical outcomes depending
24:53on the condition or the use case.
24:57There are some clinical outcomes
24:58that we have included commonly
25:00in our remeasurement plans,
25:02and I've listed some of those here.
25:04Often care plans are used for
25:06early detection of symptoms and
25:07escalation to the care team,
25:08so we would measure the rate
25:12of those early escalations.
25:14We've also utilized questionnaires
25:17like the standard PROMISE questionnaire
25:20to measure functional improvements.
25:23That the patient's experienced
25:24throughout the course of the care plan
25:27and then we've recently implemented A
25:29selfefficacy questionnaire to better
25:31understand the the patient selfmanagement
25:33goal that we have with care plan.
25:36So trying to better understand if we've
25:38moved the needle on on the patient's
25:40confidence in their own ability to
25:42selfmanage their health condition
25:43from when they started the care plan
25:46at baseline to the end of the care plan.
25:49And then we might look at healthcare
25:51utilization or length of stay
25:53depending on the particular care
25:55plan and whether that's appropriate.
25:58So I just want to close by saying that we,
26:02we do view this framework as kind
26:05of a constant working process.
26:07Or in progress.
26:08It's ever evolving and we are
26:10continually learning from every
26:11new care plan that we develop and
26:13implement and then reflecting on the
26:15the framework and making tweaks and
26:18adding things as we identify new
26:21learnings and working with our practice.
26:25Thanks. Great. Thank you so much.
26:26So now we're going to move
26:28on to our last panelist, Marcy,
26:30discussing our third objective,
26:33ensuring access to specialty care,
26:34maintaining patient safety and recognizing
26:37care acuity needs through econsults.
26:41Great. Thank you, Elizabeth.
26:43And I guess I am the final panelist
26:46on the final panel of the day.
26:48So I promise to try to make this a
26:51little punchy and keep everyone engaged.
26:55I think that Elizabeth and Laura
26:58have really defined quite elegantly
27:00to very different aspects of
27:03asynchronous care and I similarly will
27:07address 1/3 aspect of asynchronous
27:09care with the use of Econsults.
27:12Not knowing in the audience how much
27:14experience people have with Econsults,
27:16I imagine that.
27:18Many have had an experience within
27:21their institution or learning about
27:23or reading about Econsults because
27:25they've certainly been well published.
27:29Next slide.
27:34So my hope is to consider Econsults
27:37in the within the under the umbrella
27:41of quality of care and to think
27:45about the utilization of Econsults
27:47for access to specialty care,
27:50keeping in mind the maintenance
27:52of patient safety and recognizing
27:55the needs for acuity of care.
27:56And how do we address this and not only
28:00acuity of care with Econsults but also.
28:03How Econsults address acuity of care for
28:06patients that may displace the need for
28:09an in person or even a virtual visit.
28:13So as my colleagues earlier
28:14in the day talked about,
28:15I think it was Dan Albert who talked
28:18about how often the virtual visit is
28:20used to help determine if and when
28:23an in person visit needs to occur.
28:26And similarly I think the econ
28:28salt can be used to determine.
28:30If a virtual visit or an in person
28:33visit is warranted and would be
28:35helpful and also to determine kind
28:38of the timing of that next slide.
28:43When I think about the use of Econsults,
28:46I think about it in a couple different ways.
28:50There could be a quick question
28:52or the access for patient care.
28:58Could be bridged by.
28:59There's a long wait time for
29:00a new patient appointment,
29:01so the Econsult can help to bridge the
29:06work up and management for the patient.
29:09Let's first address the quick question.
29:12An ECONSULT can be submitted to purely
29:16just seek advice in patient management.
29:19The E council could also be from a
29:22referring provider who is considering
29:24referral to the subspecialty,
29:26but isn't sure if that referral is
29:29appropriate at this time or if further
29:32work up would be helpful to maximize or
29:35optimize the patient's experience and
29:37time spent with the subspecialty provider.
29:40And the other thing is that the the
29:45quick question approach is that the
29:47ECONSULT can serve to provide guidance
29:49in the acuity for the need for referral.
29:52Is this something that this patient
29:54should be seen for within a week
29:55or is this something that with the
29:57following few steps the patient could
29:59then be seen within a month and that?
30:03Brings me to the long wait times
30:05that many of us are facing for
30:07new patient appointments.
30:09And I think this has been
30:10really exacerbated by COVID.
30:11When we had a smaller ability
30:14to see patients,
30:15we had a more narrow scope of being
30:17able to see patients and many of
30:19our practices are now really backed
30:21up for new patient appointments.
30:23And so in this way the ECONSULT in
30:25an asynchronous way can bridge the
30:27work up in the management so that
30:30the referring provider can start on
30:32the evaluation and even institute
30:34some medications or non pharmacologic
30:37interventions to try to assist in
30:43the management and perhaps provide
30:45even more information for when the
30:47patient is seen by the subspecialist.
30:50In terms of access, as I mentioned,
30:53guidance for Acuity or determining
30:55if a sooner appointment needs to
30:57be made and then the other purpose
30:58that an Econsul can provide in an
31:00asynchronous way is to follow up on
31:03questions so that the Econsult can
31:05start the evaluation and management
31:07and then there can be continued
31:10communication between the providers
31:12to get more information and provide
31:14further management decision making.
31:17Similarly if a patient is seen
31:19by a subspecialist.
31:20Then follow up questions could be
31:22pursued through the E Council and
31:25this optimizes the the care for
31:28the patient and addressing in an
31:31asynchronous way ongoing patient care.
31:35Next slide.
31:38I will present a little bit of data
31:41from the rheumatology experience at MGH
31:45and there are many practices that are
31:47perhaps much larger than ours and may
31:49have much more experience with this.
31:51But this is just to provide kind of a
31:55platform to allow you to think about
31:57some things that might engage all
31:59of us in some conversation at the
32:00conclusion of my part of the presentation
32:02and in the question and answer.
32:05In the past six years,
32:06we've done 1400 E councils and I
32:08will say that since March 2020 this,
32:11the trajectory for the number of
32:13E councils is certainly increased.
32:14There are three faculty members
32:16who are participating.
32:17So certainly it's not engaging all faculty
32:20in the rheumatology division and I,
32:23they have gathered data and the most
32:25common questions that are asked in
32:27econsults to our rheumatology group
32:29relate to the patient with a positive
32:30A and A or a positive other auto
32:33antibody such as a rheumatoid factor.
32:35Elevated inflammatory markers for
32:37assistance with the management of gout.
32:40But that's not to say that other more
32:43complicated diseases aren't adjusted
32:44in econsults and you can see even
32:46patients with an unusual condition IG,
32:49G4 related these are systemic vasculitis
32:52may have questions that are fostered
32:55through Econsult interactions and
32:58one thing that I've learned from.
33:01My colleagues here are participating
33:03in econsults at MGH.
33:05Is that typically an Econsult for
33:08the responder for the rheumatologist
33:10takes about 15 to 20 minutes of time.
33:13And it it turns out,
33:14and it makes sense that if more time
33:16is required to answer the question
33:18through an asynchronous econsult,
33:20more likely a formal referral will
33:22be beneficial to the patient.
33:25My experience with the Council
33:28is more as the Econ consult the
33:31referring and asking questions and
33:33as I am engaging now through the
33:36grant that Elizabeth alluded to,
33:38to extend our reach and develop a curriculum
33:42for rheumatology fellows in telemedicine.
33:45One of the things that we're pursuing
33:49is developing an Econsult service
33:52with the Indian Health Service,
33:54which is where we are extending our
33:56reach to an underserved population.
33:59And I think that although we'll
34:01need to use a a new medical record,
34:04one that's not familiar to us,
34:05I think this has the potential to really
34:09benefit an underserved patient population.
34:13With not only asynchronous care,
34:15but very remote asynchronous
34:17care from academic Medical Center
34:20providing care to in an outreach way.
34:23Next slide
34:29you can see that for our rheumatology
34:32experience the conversion to
34:34a new patient referral was
34:36previously on the 10 to 20% range.
34:38That more recently,
34:40because we have less access to care,
34:42many more urgent patients are
34:45being referred by way of Econsul
34:48and then getting a sooner
34:50appointment within the practice.
34:53Satisfaction has been evaluated
34:55on a more global term,
34:57not just in the rheumatology
34:59division at MGH with formal survey
35:01metrics that relate to patient care,
35:04the expediting of referrals and
35:06decreasing wait times and all of these.
35:08Have met with really great
35:11satisfaction levels frequently
35:14there are communications just as
35:16kind of an anecdote that there's
35:19very positive feedback that's
35:20provided to providers who are
35:24participating in ECONSULT information.
35:27And one of the other advantages of using
35:30the asynchronous in the asynchronous
35:33use of Econsults is that then.
35:36Patients who are referred are truly
35:38in need of some specialty care.
35:40And those patients who may not truly
35:43need ongoing subspecialty care but
35:45could have questions answered in an
35:48asynchronous way don't have to make
35:49the trip to come to the hospital or
35:51don't have to take the time or the
35:53provider's time for a full visit,
35:56whether it's virtual or in person.
35:59And when patients were surveyed,
36:01they really liked the fact that
36:03they didn't have an extra visit,
36:05that they got the subspecialists attention,
36:07that they got the subspecialty care,
36:10and that there wasn't the additional copay.
36:12So as others have talked about,
36:14miles saved, dollars saved,
36:17carbon emissions saved,
36:19E councils also fulfill that next slide.
36:27I thought I would present
36:28some data from the literature and there
36:31are a lot of specialties who have published
36:34on the use of econsults and you can,
36:36and I'm not going to read
36:37through all of this for you.
36:38Please feel free to look at the data
36:41that I've highlighted in terms of the
36:43conditions seen and the the metrics that
36:46are used to evaluate the the quality.
36:50Of care that's provided by
36:52Econsults and I chose.
36:55These are each single center
36:56experiences and you can see that
36:58they have been published in 2017
37:00or they're published in 2020-2021,
37:04but the data collections are 2015 to
37:072019 and that's obviously before COVID.
37:10So those who are publishing on
37:13Econsults post COVID experience.
37:15I think we're on the brink of seeing a much,
37:17you know,
37:18much more development in our literature.
37:20And I chose these four specialties
37:23because we have allergy immunology that
37:26spans Pediatrics and adult medicine,
37:29cardiology which is procedural specialty,
37:34but really looked at the quality of
37:37ordering for cardiac stress test
37:40urology which is a procedural specialty.
37:44And neurology,
37:46because neurology has just been so
37:50fantastic and and leading the curve for
37:53synchronous and asynchronous telemedicine.
37:58Interestingly with urology you can
38:00see that about 70% of the Econcil
38:03patients were scheduled for a
38:04clinic visit and as I mentioned in
38:07rheumatology it was more on the.
38:09The nature of 10 to 20% and almost 25%
38:12of the patients were scheduled for
38:15a procedure by way of introduction
38:18through an ECONSULT.
38:20So it really fosters high acuity
38:25and high quality care for patients.
38:29Next slide,
38:34I have a couple slides now just.
38:39Has themes to maybe be provocative for
38:44questions and discussion when I think
38:47about Econsults and acuity of care.
38:50As I mentioned some questions
38:52and patient care can be resolved
38:55completely with the Econsult.
38:57We can get guidance in the pre appointment
39:01evaluation to facilitate even higher
39:03value time spent face to face or in
39:06the virtual visit with the provider.
39:09The referring can also initiate
39:12management options.
39:13We can get a sooner appointment and
39:15we can touch base about further
39:17questions with ongoing management.
39:21Next slide.
39:26In considering quality metrics,
39:27there are a lot of potential quality
39:30metrics and we've heard about a lot
39:32of different ways of evaluating for
39:34quality throughout the day. Today.
39:37I think the quality metrics with econcils
39:40are best adjust through resource utilization,
39:44time to resolution for the referring
39:48provider questions facilitating higher
39:51acuity patients for quicker scheduling and.
39:55Quality also in terms of value and
39:57and that relates to satisfaction
40:00with resource utilization.
40:01We want to optimize the new
40:03patient subspecialty visits and
40:05that can be virtual or in person.
40:08We want to be able to triage higher acuity,
40:11more complex patients to be seen
40:12and we want to reduce costs.
40:15And that's how I would think
40:17about resource utilization.
40:18It's clear with time to resolution
40:20for the referring provider questions.
40:22In urology,
40:23many of those patients were referred
40:25and seen by the specialist,
40:26whereas in rheumatology,
40:28many of the questions can be resolved
40:30without having a referral made.
40:34And then having higher acuity
40:38patients for quicker scheduling
40:39is advantageous for everyone.
40:41And the satisfaction has,
40:42I think there's a lot to be
40:45learned yet from satisfaction.
40:46I think Susan talked about
40:49this early in Panel 1 today.
40:51But certainly preliminary survey data
40:54show high satisfaction with the use of
40:57econsults from referring providers,
40:58patients and the sub specialists.
41:02Next slide,
41:06what's on the horizon?
41:10Well, I think that we all have
41:12the same questions in terms of
41:14whether insurance companies will
41:16provide reimbursement for econsults.
41:18They're highly valuable.
41:19And will facilitate high quality
41:22care and addressing high acuity care.
41:26But I think there's been a lot
41:29of inconsistency and not yet
41:31as much progress in terms of
41:34insurance company reimbursement.
41:36Also on the horizon is how do we
41:38optimize the communication skills.
41:40This is, this is a new skill set I think.
41:44Being able to communicate really
41:46clearly by the referring as well as
41:49by the E consultant will foster even
41:53higher quality care for the patient.
41:56Also on the horizon would be the
41:58possibility of crossing state lines.
42:00This could include inter hospital E
42:04consultation which obviously could be
42:06within a state but could also provide
42:09care to hospitals in more rural areas.
42:12Within or outside of the same state
42:16and then ultimately as a program
42:18director in my hat with program
42:21director is developing resident and
42:23fellow curricula because these are
42:26the new entrants into our workforce.
42:29And I think that Christine Peoples and
42:33her presentation today talked about
42:35and others actually talked to also
42:38about how we were all forced into.
42:40Providing telemedicine,
42:42providing virtual care,
42:43whether it was synchronous or asynchronous,
42:45it was new to many of us in March 2020.
42:48And how do we best prepare our workforce?
42:51By preparing the new entrance and
42:54training them to be able to be quite
42:57facile with Econsults in an asynchronous way,
43:00as well as all of the synchronous
43:02care that we have discussed.
43:07And I think that's it.
43:09I think it's time for Q&amp;A and I'll
43:11turn it back over to Elizabeth. Yes,
43:13thank both of you.
43:14Those are wonderful presentations
43:16and a lot of information.
43:18We do have some questions and comments
43:20before I I get to some that that I've,
43:22I've kind of created.
43:25So Bart Demarsha brought up a question.
43:29So and he thinks it's provocative.
43:31We'll see how provocative you are
43:33at Bart is the econsult of today.
43:36Equivalent or not to the curbside console or
43:40doc of the day duty concepts of yesterday.
43:44So that that's one.
43:45Marcy, you want to kind of address
43:47that or thoughts about that.
43:50Yeah. It's a really
43:52interesting question Bart.
43:53I, you know, I would say that the
43:57curbside console of yesterday was really.
44:03Not very favorably looked upon
44:05by a program director or training
44:07fellows because I always felt that
44:10the patient will deserve to have
44:13the physician see the patient,
44:15see the subspecialist,
44:17see the patient and evaluate the patient.
44:19But I think you're right that the
44:22Econsul edges in on that curbside.
44:24I do think that the Econsul offers
44:27something a little bit higher and
44:29that it's not just somebody stopping.
44:32Subspecialists and talking about a patient,
44:34but it is within the medical record
44:37and and so the E consultant who is
44:41dedicated to the Econsult methodology
44:44is able to not only see as I mentioned,
44:48the importance of the communication skills
44:50by the person requesting the Econsult,
44:52they can also look through the
44:55patient's medical record to
44:56find other valuable information.
44:59And has the opportunity to provide
45:04recommendations for having the patient seen.
45:07But I think you are right,
45:08it really does angle in on
45:12that curbside consult concept.
45:15I just wanted
45:16to add that it's there's
45:17documentation in the electronic
45:19medical record whereas curbside do not.
45:23Yeah, that's right Dan. Exactly.
45:25I thank you for saying that and
45:27it's it's very useful to go back
45:29and see what the answers were to
45:32my curbside counsel, you know,
45:33six months later when I'm,
45:34you know, seeing the patient
45:38to see what the consultant provided for me.
45:41Yeah. I think, you know just the other
45:43point I would make on that topic,
45:44Elizabeth is that it's an
45:47interesting turn in our.
45:49The way we work because it first of
45:52all it professionalizes the ECONSULT,
45:55so it's actually worth something as
45:57opposed to a hallway conversation
45:58that you maybe don't quite remember
46:00when you get back to your office.
46:02And as was just previously mentioned,
46:04there's no, there's no paper trail.
46:06The other thing is that especially if we
46:08monetize these either directly through
46:10reimbursement or through RVU plans,
46:12we are valuing physician time
46:16for considering you know these
46:18these these complex questions.
46:22You know, we've never been paid
46:23for our time in quite that way,
46:25unlike our lawyer colleagues who get
46:26paid every time they pick up the phone.
46:28And there's an expectation around that.
46:31And and one of the challenges, I think,
46:34that we face even before COVID,
46:36but especially now,
46:37is this increasing monetization of activity.
46:41Well, what are you going to pay me for that?
46:42What we going to pay me for teaching?
46:43What are you going to pay me for this?
46:45And the the volunteerism
46:47and the collegiality?
46:49Of Hey, Elizabeth,
46:49can I just ask you a quick question about,
46:51you know, what's the best radiology
46:53study to use to look for a cardiac
46:56thrombus now becomes that, well,
46:57why don't you send me an Econsult,
46:59you know, on that.
47:00And I think we have to figure out
47:03where to draw those boundary lines
47:05and how to make sure that the,
47:08that the interpersonal and interprofessional
47:11communication that is so vital to
47:13our sense of joy of being at work,
47:15our sense of community.
47:17Doesn't get lost if we shift too
47:21much to the asynchronous modality.
47:23At the same time really professionalizing
47:28a more than curiosity related
47:31consultation about our patient care.
47:34Sorry it's a long comment,
47:35but I it's it feels like it's so
47:37much at the center of how we relate
47:39to one another as colleagues.
47:41It is and it goes back to a more,
47:43she said about that,
47:44the communication skills.
47:46Because you've got very
47:46different in the hallway.
47:47You know, I can explain to you,
47:49you know why this imaging
47:50study is better than another.
47:52I can guarantee you that
47:53during in in the Econsult,
47:55it'd be, well, this one
47:57without much of an explanation,
47:58simply because I don't have to.
48:00It's hard to sit there and
48:01type out the whole explanation.
48:02Where is it? We're in the elevator.
48:04Which, by the way, now Econsults are
48:06a heck of a lot more HIPAA compliant.
48:08You don't have to worry about that
48:11talk in the elevator being overheard.
48:14You know, it it it does bring
48:16in into those skills.
48:17I mean, what do I communicate
48:19and how much do I communicate?
48:21Or I've known people who could write
48:23an e-mail that's a novel and nobody
48:25wants to read that one either.
48:27So we really do have to find that
48:30that balance of communication skills.
48:33So, Laura, let me ask you a question.
48:35How do you address inequities of
48:38access to technology for RPM programs?
48:41Sure.
48:42I think
48:44it's it's a challenge and one that
48:46we continue to work on and continue
48:48to ask each other questions about.
48:50I think with the interactive care
48:54plan specifically we've run into
48:57some regulatory barriers in regards
48:59to trying to address some of
49:01those gaps for patients who maybe
49:03can't provide their own devices.
49:06In terms of like Medicare kickback,
49:08so there are some barriers I think
49:10to to work through there to to get
49:13devices in the hands of patients.
49:15Other ways we've addressed disparities
49:19in regards to Internet access and
49:23access to the technology that we
49:27use for remote patient monitoring
49:30is to provide patients with kind
49:32of the full service kit that has a.
49:36Device that is cellular enabled,
49:38so they don't have to have access to Wi-Fi,
49:40they don't have to bring their own device.
49:41So we've kind of we've kind of got the
49:45Cadillac available to us when we want to,
49:47when we want to address those
49:50use cases or when we need to,
49:53that comes at a cost.
49:55So we're always trying to weigh the.
49:59The pros and cons and sometimes
50:01we have to go with the higher end
50:03solution so that we can make sure we
50:06are reaching all our patients when we
50:08would maybe prefer or maybe the low
50:10intensity solution is actually would be
50:13just fine and appropriate for that use case.
50:16We end up going with the higher model
50:18because we can serve more patients that way.
50:21So we've got some challenges to work through
50:23in regards to regulation to figure out how.
50:27How we can,
50:28you know,
50:28use our resources as a nonprofit,
50:31we should be able to provide some
50:34of those services and equipment
50:36to patients and not run into
50:39things like Medicare Kickback.
50:41So there's definitely some
50:42opportunity to work through those,
50:44but those are some of the things
50:45we've done and we've done and
50:47some of the challenges we're
50:49actually working through right now
50:50with interactive care plans.
50:52Grade that kind of leads me to sort
50:54of a tangential question and it was
50:56a kind of came to me when Lee put
50:59in the in the in the comment section
51:01that there's that you know difference
51:04between consumer grade and medical
51:05grade specifically said displays
51:07of the technologies in general.
51:09And when does the FDA come into all of
51:11this and it a question kind of that
51:13comes to my mind and it's come before,
51:15but it kind of applies to all
51:17of our talks is. You know,
51:19what do you do with patient generated data?
51:20Because that's going to be 9
51:22times out of 10 asynchronous data.
51:24They're going to snap a picture
51:25of that funny looking mole.
51:27They're going to take a picture
51:28of their kids ears with these
51:29wonderful apps that come out there.
51:31And there's an app for just
51:33about everything these days.
51:34It gets into the hands of the providers
51:37and whether it's been sent on an ECONSULT,
51:40whether it's a remote patient
51:41monitoring with a Google watch,
51:43I mean that's data that they collect.
51:45It's real time monitoring.
51:46They're sending it to you.
51:48What do you do with the data in terms of,
51:50you know,
51:51sort of deciding is this quality data,
51:54do I have to respond to it or not?
51:57You know, from the perspective,
51:58from my perspective on a lot of its
52:01image base and I've seen them, you know,
52:03come across a lot of the times,
52:06the image is just awful because
52:08these devices can go through all
52:10sorts of approvals and lovely little
52:12clinical studies where the most.
52:14Well trained person has taken all
52:16the images and says well heck I
52:18could put this in the hands of a
52:205 year old and get great images.
52:21We're guaranteed not you're not
52:24going to get very good quality
52:26images most of the time.
52:28So what is the responsibility of
52:30providers and you know and and
52:32at what point do we have to put I
52:35think somebody used the the term
52:37before barriers around this.
52:40So Elizabeth, let me just,
52:41I can't restrain myself from replying
52:43to that question because we've been
52:45spending a lot of time thinking about it.
52:48So we've done a few things.
52:49We've drawn the distinction between provider
52:52initiated versus patient initiated data.
52:55We think about patient generated
52:57health data as the broad taxonomy
53:00RPM remote monitoring is kind of
53:02a bucket within that which we.
53:05Have generally been reserving for provider
53:07initiated medical grade devices or near
53:10medical grade blood pressure cuffs,
53:12activity monitors,
53:14oxygen saturation, glucometers,
53:17devices where we're not too worried
53:18about a thumbprint over the camera lens.
53:20You know when someone's taking a photo.
53:23I think patient initiated data can
53:25come in all stripes and flavors.
53:28It can also contain.
53:30Cyber security threats,
53:32unintentional or intentional.
53:33And so we definitely need to think about
53:37filters and gatekeeping functions before
53:39that information is able to be uploaded.
53:42And I think there's a social contract
53:44that we need to have with our patients,
53:47just like we don't accept the the idea
53:50that they will send us A50 page e-mail.
53:54About their medical care and expect
53:56us to respond or leave us a 20
53:58minute voicemail in lieu of a visit,
54:00right.
54:00We we do need some boundaries around
54:03this and I think ideally we should
54:05come to this together with our
54:07patients and we should be thinking
54:09about alternative interfaces either
54:11peer-to-peer networks or patient
54:13to chat bot or other forms of
54:17automation that can work with patients
54:19to preprocess some of this data.
54:22And then put it in the right bucket,
54:24but we don't want all of it going
54:26into the chart that is for sure.
54:28We at at M GB Master and Brigham
54:30are building a data lake where
54:31we're going to house a lot of this
54:34patient generated health data if
54:35it's not provider initiated.
54:37So we have a place to park it,
54:40but we have a a corral where we
54:42can sort of decide what if any part
54:44of it is a high enough quality or
54:46of reasonable value that we want
54:48to import it into the record.
54:49But it's a really, really important area.
54:52I think I would more of an ocean
54:54than a lake but that's okay
54:56because I think they're going to
54:57get you a lot of information.
54:58But I mean what it also brings up
55:00you know that the whole issue of
55:03sensitivity and specificity and and
55:04the whole false positive situation
55:06and that that's with you know all
55:09the data that we've been talking
55:11about and that's a huge quality
55:13issue that that we really have
55:15to start to deal with as well.
55:17I would add one thing to that we've.
55:20Come across with Remote Agent
55:21monitoring as we've tried to implement
55:24a solution for device integration
55:26with Interactive Care plans.
55:28So one thing that we're working
55:30with our EHR vendor on is.
55:33Making sure that they're able to
55:35distinguish for the clinician in the
55:38patient's chart whether physiological
55:40data that's coming through the care
55:42plan is coming from a patient self
55:45entering that physiological data,
55:46so having a blood pressure
55:48device that's not integrated.
55:50Where the data is not flowing automatically,
55:53where the patient is physically
55:55taking their measurement and then
55:57punching it in with their fingers.
56:00And there's potential for error
56:02there as compared to once we
56:05get device integration going,
56:06Making sure the clinician can see
56:08this was a patient under value versus
56:10this is a value that came from the
56:12patient but came from the device.
56:14So we can be assured that there's
56:16hopefully not a manual entry error.
56:20Now, good, good points.
56:22We're kind of getting out of time here.
56:25A lot of good comments in the discussion,
56:26but I don't think I don't see any
56:29more actual questions. Marcy,
56:31any final thoughts from your perspective?
56:35No, I think it was a fantastic
56:38discussion and I think that it
56:40was great to see three different
56:42aspects of asynchronous care and
56:45we're just on the edge of seeing
56:47even more opportunity for that.
56:50Great, Laura, Any final
56:52thoughts from you then, Lee?
56:53I'm handing it back over to you at 5:44.