CHAPTER EIGHT
Here, There, and Everywhere
A CFI Project Showcase
Innovation that works is a disciplined process.
We will use Chapter 8 to share specific projects from three of our four platforms: Mayo Practice, Connected Care, and Health and Well-Being. (The fourth platform, the Innovation Accelerator, was described in Chapter 6.) In this chapter, we’ll take you on a short flight through a series of CFI projects.
We start with a low-altitude flight through the Mars Outpatient Practice Redesign project as an example of scanning and framing, experimentation, and prototyping to transform the Mayo Clinic outpatient practice. No, “Mars” is not an acronym. Rather, it’s a from-the-ground-up construct to answer the question, “What if we were to start from scratch and set up a new medical practice on Mars?”
As part of Mars, we’ll take a more detailed tour through a subproject called microConsults. microConsults use electronic tools to incorporate specialty consultations into a single visit, avoiding a second appointment and trip to our facilities. You’re at the Mayo Clinic for an orthopedic treatment, and the orthopedist recommends a neurologist consult? Done, then and there, on site, in real time, electronically. No referral appointment, no second visit. Estimated savings? Over 38 minutes per appointment for microConsults-enabled appointments—that’s part of a 30 percent reduction in outpatient cost at Mayo Clinic, which is the overall financial goal of the Mars practice redesign—and more important, the benefit to the patient of a far smoother and more effective health care experience.
From there, we give shorter, higher-altitude summaries of other key parts of our four-platform vision. For example, as part of Connected Care, we look at eConsults, which leverages digital technologies to extend care beyond Mayo bricks-and-mortar facilities. We examine our Optimized Care Teams project, reinventing care team staffing, and the adjacent Wellness Navigators project from our Health and Well-Being platform. Finally, we wrap up with the Mayo Clinic Patient App project, which came through our CoDE incubator. Although they originated in separate CFI platforms, you’ll see how these projects connect with each other and how they play into our greater vision for transforming the 21st century model of care.
A Flight Plan for Our Projects
Before taking off on our showcase tour, let’s review how we map our projects into the 21st century model of care vision and how we get these projects off the ground.
Revisiting the 21st Century Model of Care Vision
First, here’s a quick refresher on our 21st century model of care vision, which can be stated as “Always be there for me”:
“When I need to come to you”—Mayo Practice (Mars), “Here”
“When you can come to me”—Connected Care (eConsults), “There”
“When I didn’t know I needed you”—Health and Well-Being (Optimized Care Teams, Wellness Navigators, Mayo Clinic Patient App), “Everywhere”
Project Presentation
Here is a quick recap to prep you for the flight.
Scanning and framing is the first major step of any project or program. Here, we scan the world around us, examine the situation, do our initial research, and frame the opportunity or problem to be solved in the broadest context.
Next, we develop and document our research path, which includes a sequence of experiments and prototypes designed to test our insights and the tools, processes, and technologies that we think will solve the problem and that may lead to a deeper customer understanding or a reframing of the problem in many cases. The scientific method really takes over at this stage, and the exact course of a project may change radically here, especially at the experimentation phase.
Then, to maximize the value of our learnings as well as to demonstrate them to constituents inside and outside Mayo Clinic, we summarize the findings of both our experimentation and prototyping efforts. Our findings can lead us to more prototypes and to implementation as well as serve as a vehicle to document concepts that can be used in adjacent projects or platforms.
Finally, once settled on the mechanics of a project, we develop an implementation plan to guide our practices to adopt it. Typically these program write-ups include descriptions of the new tools and processes and their impacts, a description of the staff needed for implementation, and a complete, typically self-paced training guide to walk the practices through it.
From here our showcase tour will give you important examples of where we’ve gone and how we got there. To start, we’ll take you on a rather long flight to Mars and back to present an in-depth example; then we’ll take several shorter flights to round out your tour of key CFI projects.
A Journey to Mars
Think about the last time you went to see a physician. Chances are, if you’re relatively healthy, you had some measurements taken upon entering the office, and then you had a brief consult with the doctor, you received advice and maybe a prescription or two, and you moved on.
Such simple, one-stop visits happen all the time, every day, all across the country. But suppose your situation is more complex. You have multiple complex symptoms, or you have something the doctor would like to have checked out by another specialty. Can that happen during today’s visit? You’ll get a referral, and most likely you’ll have to make another appointment, come in for another visit on another day, and go through another round of discussions, lab tests, and treatment planning.
And you still might not be done. Perhaps a third specialty needs to come into the picture. The visits start to add up: time spent in waiting rooms, time documenting medical records, and worst of all, from your point of view, time spent in repetition of the same information and the same questions. All in all, for you, these are seemingly endless trips to a clinic or hospital. Medicine, especially the treatment of complex diseases, is a complex thing. At Mayo Clinic we are famous for and proud of making things happen during the same visit day but, even here, many times this is not possible.
The Background
So we set out to address this complex patient experience. Our interest, of course, centered on the patient, but really, the same inefficiencies that were blanketing the patient experience were also wreaking havoc in our practices—in this case, our outpatient practice. Patients would make the trip to Rochester and schedule an appointment, only to find out that they needed another appointment, which we would try to get scheduled on the same day but which could be days out. More inconvenience, more effort, more documentation, more check-ins and checkouts—we just wanted to start over and create the “perfect” practice for these patients.
Very quickly, we started to call this project “Mars”—what if we were to scrap everything here on Earth and create the perfect multispecialty outpatient practice on Mars? What would we do? How would we design the practice from the ground up—the physical spaces, electronic tools, administrative support, the staffing—to get the job done with minimal inconvenience to our patients and most efficiently and cost-effectively?
We had a strong sense that something big could be done here. With the right research and combination of innovations, we could streamline the patient experience. We could react better to the patient condition known before and after the initial visit. We could avoid costly and disruptive stops and starts in the treatment process, reduce repeat visits from referrals and separate inputs from multiple specialties, and save our payers a bundle of money in the process.
Project Mars was born. We set a specific goal to improve the patient’s experience as well as the provider’s practice experience and to reduce practice costs for Mayo Clinic. We created a team and let it proceed.
Understanding User Needs
So with these strong hunches in mind, we set out to scan and frame our trip to Mars. What would the perfect practice, created from the ground up, look like? How would it work? How would it address the nagging issues out there? And how would our discoveries and prototypes in the Mars project, part of the Mayo Practice “When I need to come to you” platform, support the Connected Care and Health and Well-Being platforms?
We started this practice redesign project with big expectations of discovering efficiencies and applications of technology that would help across the health care spectrum. From the outset we “went big” with the discovery process to develop a complete list of insights through hours of external scanning, detailed observation, and listening.
A Patient Care Continuum
We developed a patient continuum to describe why most patients seek medical help (Figure 8.1).
FIGURE 8.1. PATIENT CARE CONTINUUM
For Mars, our specific focus was on the more complex patients who would require the care of multiple specialties. The more complex patients using our outpatient practice generally fall into the two left-hand categories, so-called complex or chronic patients. Complex patients are the most challenging to schedule and move through the process due to the combinative nature of their diseases and the care required to address them. In some cases care is delivered when specialists can consult together; others can be handled in discrete consultations. But in all cases, the need to consult multiple specialties produces a logistics challenge.
Chronic care patients may or may not need to see multiple specialists, but they do require repeat visits, so easy and efficient scheduling is also important. For chronic patients, as well as others across the continuum, we began to have a notion that progress could be made by looking at the big-picture Mayo Practice—that is, Mayo Clinic as a central practice containing the specialties sitting behind Mayo, or the non-Mayo but partnered with us, primary care facilities local to the patient. A chronic care patient could get treatment locally with an electronic consultation with a specialist in a Mayo outpatient practice—if the logistics surrounding scheduling, medical records, and interactive technologies could be managed.
The upshot: we realized that Mars discoveries would be significant throughout the patient experience for all kinds of patients. A large-scale research effort was warranted.
Practice Research
We began a major practice research campaign to observe and discover what was really happening in our outpatient practices and to develop insights from that research. The campaign was conducted by CFI designers over a period of eight months in our practices. Specifically, the exploration included 200 hours of interviews and observations across the 50+ practices and specialties. We talked to patients, physicians, nurses, clinical assistants (C.A.s), and other staff members. We conducted 35 ethnographic patient interviews in three cities to understand patient needs, goals, and motivators. We also completed a thorough external assessment of best practices and trends, summarizing the state of the industry with “trend cards” (introduced in Chapter 4).
In all, it was a from-the-ground-up, start-from-scratch exercise in understanding what was really going on in our practices (the scan) and where we could best direct our innovation efforts toward solving the most important problems (the frame). Of course, the research didn’t stop here—it just got us started! Experimentation and further research would get us closer to true customer—explicit, tacit, and latent—needs.
Practice Insights
From our deep observation, we identified a list of 238 insights in 12 categories to describe the current state, the key challenges and issues, and general observations about what occurred daily in the practices.
Here are the 12 categories. You may find these categories also apply to your organization’s main challenges:
Mayo Organization, Culture, and Beliefs
Access
Coordination of Care
Predictability and Preparation
Variance and Flexibility
Care Team Dynamics
Systems and Practice
Communication
Patient Experience and Relationships
Care Models
Billing and Financial Issues
Framing the Problem
Even when the insights are grouped into 12 main categories, it took us a while to synthesize the body of research—that is, the 238 insights—into an actionable set of problems. Eventually, we did boil our outpatient practice redesign down into seven encompassing issues and problems to be solved:
Lack of standardization. We maintained different processes for scheduling, treating, and managing patients, as well as different team compositions, even though the institution desired standardization.
Only one type of patient visit offered. We essentially had only one offering for all patients: a face-to-face appointment, although we could vary it in length and provider type.
Patient complexity. Scheduling needed to classify people as each having one condition, and additional needs were visible only after the patient arrived.
Lack of timely access. Different departments and providers had better access than others; access time ranged from a few days to a couple of weeks, months, or even a year.
Patients not transitioning back to primary care. Once patients became part of the subspecialty outpatient practice, they rarely transitioned back to primary care.
Lack of systemic recovery. If the needs of the patients changed or if they were not originally scheduled correctly, care teams found it hard to readjust the patients’ schedules.
Clerical burden. Providers and care team members had to use, and move between, multiple, mostly fixed, IT systems, causing significant clerical burdens.
Creating a Research Path
Fast forwarding just a bit, we felt that we could define the basic problem areas as (1) identifying correctly the care needed, and thus the care process and schedule, of a patient before the visit and (2) more effectively reacting to true patient needs during and even after the visit. By applying new technologies, including synchronous (real-time) and asynchronous (not real-time) interconnectivity, predictive modeling, and even the promise of genomics-based disease modeling, we could design and create “an intelligent adaptable system to provide an unparalleled experience.”
This led to identifying the main lines of experimentation that we felt would lead to the creation of an intelligent adaptable system and ultimately the redesign of the practice. Here are the seven Mars research paths, or “families,” grouped in four larger categories—know the person; provide the right service, in the right place, at the right time; optimize services and experiences; and create awareness and flexibility:
1. Know the person
Pre-visit question sets. Establish an automated and standardized process for obtaining patient clinical and psychosocial information at the time of the initial appointment request and/or prior to a patient’s appointment.
2. Provide the right service, in the right place, at the right time
Customized education. Provide education to patients in a format, location, and timeline that is more convenient and better prepares them for their visit and to create care team and provider efficiencies.
Shared medical appointments (group visits). Create appointments for small groups of patients in situations in which meeting together would support the medical and social needs of the patients and that would support the availability and needs of the care team.
Remote follow-up. Conduct follow-up or return visits with patients remotely in an asynchronous or synchronous manner.
3. Optimize services and experiences
SmartSpace: reducing the clerical burden. Automatically record conversations between providers and patients, and auto-fill clerical information (billing codes, notes, and orders) to allow providers to be more efficient and reduce clerical resources.
microConsults. Enable one provider with a patient to connect with another provider to answer a focused question.
4. Create awareness and flexibility
SmartSpace: increasing our situational awareness. Use a Wi-Fi–enabled mobile application to provide near-real-time, geospatial views of the care team members, panels of patients, and workflows in a space throughout a day. This will allow us to respond more quickly to changing conditions, potentially including more day-of or week-of appointment scheduling, real-time triage, and systemwide “air-traffic control” with visibility and oversight to the system.
These research paths, or families, ultimately became subprojects under the Mars umbrella, based on the results of further experimentation. Figure 8.2 summarizes the research progress toward identifying the seven initial experiment families. From here, we will drop down into one of these paths—microConsults—and give a tour of the specific experiments and research done on this emerging subproject.
FIGURE 8.2. MARS: THE DISCOVERY BEGINS
Systematizing Practice Flexibility with microConsults
The microConsults project is designed to allow a practice to schedule an electronic consult from another specialty based on need discovered during a patient visit.
The scenario described at the beginning of the chapter serves as an example. A patient arrives for an orthopedic exam or treatment, and the scheduled physician decides that a neurological consult is in order. That consult could be scheduled and delivered in near real time using electronic means—a short, perhaps 15-minute video consult, with shared charts, texts, X-rays, and electronic medical record (EMR) data. This construct was the basis of our set of experiments.
The microConsults Concept
From our observations and initial data collection, we realized that it was common, especially for complex patients, to require a secondary consultation from another specialty. As we’ll see later, that percentage varied by practice; through our research it averaged about 31 percent of patients with a range of 25 to about 50 percent.
Further, we realized that in many cases, again varying by specialty, patients were forced to leave the clinic and come back another day, to receive the secondary consult. That consult could happen the same day but rarely did through a formal scheduling process. If a secondary specialist saw a patient at all, it was usually through an informal contact and a “curbside” visit—usually a drop-by visit or a consultation from the primary to a secondary provider by phone, if, by chance, it happened to work out.
With microConsults, the vision was that a primary provider, upon deciding that a secondary provider opinion and/or treatment was in order, could systematically see the availability of a secondary provider and contact that specialist, in real time or near real time, to deliver a short, electronic three-way consult between the patient, the primary provider, and the secondary provider. The orthopedist could call in the neurologist; the neurologist, if available then or later that day, could review the patient case with the primary physician in a real-time, video chat. If successful, a treatment or treatment plan could be developed by both physicians together, usually in the presence of the patient too. The secondary physician could be called in without the time consumed and interruption of a full face-to-face visit.
While the business case is not always clear in the early stages, in this case it was very strong for Mayo Clinic with the move away from fee-for-service accounting to total-cost-of-care and bundled payments accounting.
Designing and Scoping the microConsults Experiment
With this idea in mind, we began a deeper exploration of the basic need and workflow concept. We began our exploration by holding co-creation workshops to understand the challenges and identify the opportunities. We hosted and facilitated four co-creation sessions over a three-week period, involving 19 provider participants across different medical departments and specialties.
In these workshops we identified different patient types who could benefit from increased integration including acute and nonacute complex patients, simple patients, and surgical patients or those who are considering surgery. In the workshops we also outlined a workflow that walked through a way to connect two providers and the patient. At this stage, we gave this workflow the name microConsults and developed a definition:
microConsults is an integrated care model where multiple providers meet with a patient simultaneously in a clinical space using technology that allows them to virtually connect and collaborate. A microConsult will be typically effective for a focused question that leads to diagnosis and adjustment to treatment. Many take 10 minutes or less. A microConsult can take place as a short scheduled consult or as a real-time unscheduled event.
After scoping the type of interaction and appropriate patient types, we then wanted to understand the potential scalability of this idea in different departments. We completed a retrospective review with five practices to understand how many patients could have been appropriately served with a microConsult.
We evaluated 154 appointments, our sample being composed of new and established patients, and we determined that 31 percent of the appointments could have been handled as microConsults. This study helped us understand that there is a significant opportunity for microConsults across Mayo Clinic.
Two Iterations
The next step was to design the experiment. The team laid out two phases, or iterations, to move forward. The first iteration was designed as a live proof-of-concept exercise to confirm the value of the microConsult concept to the practices and to the patients; the second was to develop the operational model and confirm that it could actually work in practice without CFI involvement or interaction.
Iteration 1: Proof of Concept
Iteration 1—essentially a formal, electronic curbside consultation—was designed to confirm the value of microConsults to both patients and providers; to confirm the anticipated efficiencies in reducing repeat visits, scheduling overhead, and so forth; and to confirm the overall viability of microConsults to Mayo Clinic. The experiment began by creating a set of design questions that defined what we wanted to learn in the context of the challenges identified prior in our research (Figure 8.3).
FIGURE 8.3. MICROCONSULT PROCESS (DRAWING BY NICHOLAS BREUTZMAN)
Here are some examples:
Does a microConsult effectively reduce referral appointments?
Is a microConsult appropriate in some situations in which the redundancies of the rooming, history, and exam are not needed for each patient?
Is there value and satisfaction in this interaction to both patient and the initiating and receiving providers?
We also decided that Iteration 1 should not include a test of different technologies. Instead, we would go with Apple FaceTime as the primary interface tool plus additional enhancements to practice scheduling systems. A complete set of User Requirements (too detailed to present here) was created.
Partnering strategies and concepts were a big part of Iteration 1. Our initial studies led us to work with General Internal Medicine due to the relatively high concentration of potential microConsult-eligible cases in this practice. We decided to add in a test of a specific group of chronic patients—acute dialysis patients.
We recruited individuals from primary and secondary practices to participate in the experiment; overall, there were 12 initiating and 37 receiving specialist providers.
Iteration 1 Metrics
Our metrics included the following:
Appointment length. The length of the microConsults
Referral status. Whether the patients were being referred following the microConsults
Impact and acceptance of the technology. Captured qualitatively through debriefing interviews with both providers and the patients regarding their impressions of the technology from a performance and overall acceptance perspective
Value. Captured qualitatively through debriefing interviews with both providers and patients regarding the value of the interactions
Patient and provider satisfaction. Captured qualitatively, and in most instances through numerical rating, with both providers and patients, regarding the satisfaction and overall thoughts about the interactions
Iteration 1 Results
Initial microConsult measurements were encouraging. In the initial iteration, we completed 27 microConsults over a period of six weeks. The average length of the microConsult session was 9 minutes, 25 seconds.
Obviously, to shed more light on potential patient and practice efficiencies, we needed to quantify the savings in time spent with patients and in administering the visits. We compared the time utilization between the microConsult and what it would have been in a typical referred consult. The results were substantial: over the 27 microConsult appointments, we measured a savings of 1,035 appointment minutes (17.25 appointment hours), or just over 38 appointment minutes per appointment.
But that wasn’t the whole story. What about the patient experience in terms of the time savings? In all we estimated a total savings of 118 itinerary days—that is, days spanning the initial and secondary consults as they would have happened, or about four days per patient. That’s a lot of travel and hotel dollars saved for the patient, not to mention anxious time waiting between appointments, family member travel costs, and so forth.
We also measured satisfaction through surveying for overall satisfaction with the experience, on a 1-to-7 scale:
The average ranking for initiating providers was 6.1.
The average ranking for receiving providers was 5.7.
The average ranking for patients was 6.2.
We also collected initiating provider, receiving provider, and patient comments:
Dialysis, initiating provider. “We were able to schedule an upper endoscopy for the patient. This allowed patient care to move forward quickly, and it satisfied the patient who was significantly impacted by the condition.”
General Internal Medicine, initiating provider. “The microConsult expedited the process a great deal, and after confirming the diagnosis with Endocrinology, the patient changed his travel plans and opted for surgery the next day at Mayo Clinic.”
Gastrointestinal, receiving provider. “This is excellent. Far superior to a face-to-face consult with the patient because the dialysis nurse practitioner was present, and she could help interpret the relevant medical information. There is great potential for this process. We could potentially do six of these in an hour, saving everyone time and money.”
Employee Community Health, patient. “The most positive thing is that we are all immediately on the same page. My doctor knows what the specialist said and can ask questions I wouldn’t have thought of. I don’t have to try to remember for my next appointment. Everyone already knows the story.”
Moving on to Iteration 2
Iteration 2, which is in progress at the time of this writing, is designed to fully automate the process, removing CFI from the loop, and to make it possible or easier for an initiating provider to schedule a “scheduled” or even a real-time microConsult directly and independently of CFI and administrative help. The team is presently recruiting additional partners. Metrics are expected to stay the same.
Upon completion of this iteration, the expectation is to “go live” with a version of microConsults and to eventually expand it to most practices, and to many initiating providers outside the outpatient practice.
From the example, you can see the Fusion Innovation Model at work. There was plenty of design thinking and co-creation—in this case, the patient and the practice customers. You could see the blending of scientific method (hypothesis, experiments, detailed measurements) and project management (phasing, reporting, communications) inherent in the effort. And there was constant iteration through the various stages of research and project development.
Being There and Everywhere with eConsults
From here, we want to summarize some of our other more illustrative projects. To respect your reading time, we offer these next “showcase” reviews at a “higher altitude,” more of a big-picture summary.
As you probably noted, the Mars Outpatient Practice Redesign project covers a wide gamut of issues related to the physical destination practice. Among other things, we introduce asynchronous and synchronous connectivity to achieve greater practice flexibility and to deliver medicine beyond the walls of the clinic by bringing in specialty consultations through microConsults.
The principles and technologies of microConsults were not so new; for Mars, the application to achieve secondary consultations in near real time, without a follow-up appointment, was new and transformative, particularly for the patient experience. Really, microConsults borrowed principles and technologies already developed in our Connected Care platform, conceived to connect patients (and other providers) to Mayo physicians from “there and everywhere”—that is, without coming to a Mayo Clinic facility.
A Triple Win
Since 2009, CFI has been researching and developing “connected” models to deliver Mayo primary and specialty care without patients traveling to a Mayo facility. These models transform patients’ lives by bringing Mayo Clinic specialty expertise conveniently to them. Furthermore, the e-models reduce the cost of care and help Mayo leverage its provider capacity in a manner that can help the most patients. The combination of patient experience and enterprise benefits makes these Connected Care initiatives especially important. In our lingo, we would describe it as a “triple win.”
Among several projects within Connected Care is eConsults, an asynchronous electronic consultation model allowing remote secondary consultations to occur behind a Mayo or non-Mayo local primary provider.
Developing eConsults
The eConsult model offers an efficient way to access specialty consultations when a face-to-face visit with the patient is not required. The project is in an expanding implementation; we have defined 170 medical conditions appropriate for an eConsult, and we have completed 14,000 eConsults since the model was fully launched three years ago.
Problem to Be Solved
This innovation was developed out of the opportunity to electronically deliver cost-effective specialty care to patients with established primary care providers, in place of traditional face-to-face specialty consultations. This model would then allow remote patients to receive Mayo specialty care, providing an affordable and convenient option and open access for patients most in need of face-to-face care.
We looked at two delivery approaches: synchronous (real-time dialogue with the care provider) and asynchronous (non-real-time—answered quickly but not instantaneously). It had to be easy to use, particularly in that non-Mayo users in non-Mayo primary care facilities had to be able to access it and receive results seamlessly.
Understanding User Needs
We began our work by collaborating with our largest commercial payer, Blue Cross and Blue Shield of Minnesota (BCBS-MN). We worked together on the eConsult Model of Specialty Care, which would be more convenient for patients, would strengthen our relationships at the local community primary care level, and would reduce the cost of care. We selected a BCBS-MN–affiliated clinic, located in Duluth, Minnesota, hundreds of miles from Mayo Clinic Rochester, to develop a pilot model.
We collaborated with providers in Duluth and with our Mayo Clinic primary care and specialty physicians to understand the clinical situations for which an electronic consult (eConsult) could replace a face-to-face consult. Just as a technical investment analyst might “backtest” an analytic model, we reviewed medical records to further define the scenarios of use, and we surveyed our physicians to understand their needs and willingness to use this type of model. Through this initial research, we estimated that about 30 percent of secondary referrals could be handled through an eConsult and that 85 percent of the physician specialists surveyed believed that an eConsult would be feasible.
We also worked to understand the patient experience and how best to deliver care at a distance, and we built a model that incorporated key features of the Mayo model of integrated and patient-centered care. We felt that the savings in patient time and travel by avoiding a second appointment and trip would be obvious, but we did make a point of getting patient feedback and testing eConsult response times to ensure minimal latency in getting a specialist response.
Experimenting and Prototyping
Our initial research proved so positive that we decided to launch a large experiment, really in scale, a full prototype, with the BCBS-affiliated Duluth clinic. The experiment initially covered 120 specialty consultations with the external practice group over a seven-month period. It then expanded to a larger scale prototype across Mayo over a two-year period with participation from 39 specialties and for 158 medical conditions.
We flowcharted the current process of face-to-face consults, including the patient touch points, operational aspects, and interactions with the electronic medical record. We then prototyped an eConsult delivery method and tested it to ensure the quality of the process. We piloted it with a few primary care providers at the BCBS-affiliated clinic in Duluth, and we started with one specialty practice (cardiology). Once proven to be effective, we gradually spread the eConsult model from the Duluth practice to our own primary care providers at Mayo Clinic and specialty practices (Figure 8.4).
FIGURE 8.4. ECONSULT IN PROGRESS
To properly execute the experiment, we had to take on some important issues. First, since electronic consultations are generally not covered by existing payers, that is, insurance or government payers, we had to establish payment through BCBS. Naturally, we also had to develop easy-to-understand patient communication materials that the primary care providers could utilize when discussing this modality of care with their patients.
We worked with specialty practices to understand the conditions for which eConsults could be provided and the prerequisites for specialists to render the consultations. For example, if the eConsult is for endocrinology/osteoporosis, it would require a bone mineral density completed within the last six months, a serum calcium level, and a list of current medications.
We developed a tool to assess the satisfaction of both the primary care providers and specialists with this new model of care. We also undertook a similar exercise for patients to compare this eConsult to the face-to-face consultative model.
Experiment Findings
The large-scale experiment told us a lot:
Volumes. By September 2011, we had completed 6,253 eConsult orders to Mayo Rochester. About 65 percent were adequately completed using the eConsult method; 35 percent required further face-to-face consultation.
Time for completion. Typical eConsults took 15 to 20 minutes, about a third of the time of a face-to-face consult. About 27 percent were complete in the same day, 83 percent were completed in one business day, 93 percent in two business days.
Quality. Through random sampling and registered nurse (RN) review, we learned that about 98 percent of the consults requested were appropriate for both the condition for which it was ordered and the clarity of the question in the requesting physician note.
Physician satisfaction. Based on physician surveys taken during and after the experiment, 84 percent of referring primary physicians who had ordered an eConsult were either very or somewhat satisfied.
Patient satisfaction. Detailed surveys and interviews were conducted with patients, and they reported high levels of satisfaction—over 90 percent very or somewhat satisfied. Not all physicians were early proponents of this type of care, and engaging with early adopters to demonstrate and lead the way was critical to its acceptance. Fear of not having the personal relationship or personal touch was a component of this model that needed to be overcome. It helped to ease this concern by having the specialists be involved in determining the appropriate conditions for eConsults.
We should also note another factor responsible for the model’s success: the passion and commitment of a CFI physician, Dr. Rajeev Chaudhry, who led the project and an engaged interdisciplinary team of designers, project managers, physicians, nurses, financial and systems analysts, and staff members from operational areas across Mayo Clinic. The team won the revered Mayo Excellence in Teamwork Award for the eConsult project. This project was co-creation at its best.
Implications
It was pretty obvious early on that eConsults offered a new modality for providing specialty care. This is especially appropriate for situations in which knowledge transfer is needed between primary care providers and specialists. With key health care challenges including the increasing complexity of care (especially with new tests and medications), excessive costs, fragmentation, a projected shortfall of physician capacity, and emerging paradigms of patients as consumers, there is value in providing remote collaborative tools for primary care physicians and specialists to work together.
As of the end of year 2013, we had completed over 25,000 eConsults. In 2013, we performed almost 8,000 at Mayo Clinic Rochester, almost 2,000 at our other Mayo practice sites in Florida and Arizona, and about 1,800 in our Mayo Clinic Care Network and international affiliate practices. In 2014, volumes had increased 40 percent from 2013.
The Nurse Is in Today: Optimized Care Teams
The Health and Well-Being platform is our third platform in the “here, there, and everywhere” vision. The essence of this platform is a broadening of the definition of care to include health maintenance and enhancement beyond the traditional physician visit. Major thrusts of this platform include the Community Health Transformation project, which we’ll explore further in a moment, and the Healthy Aging and Independent Living (HAIL) projects targeted to seniors and the chronically ill and managed through our HAIL Lab.
For our showcase, we’ll take you on a high-level tour of another project that intersects closely with the Mayo Practice and Connected Care platforms: Optimized Care Teams (OCTs).
Background: Care When and from Whom You Need It
Optimized Care Teams is really a subproject under a larger Community Health Transformation umbrella project. The central idea of the Community Health Transformation is to provide the right mix of care based on need, geography, and cost. In our thinking, three themes intersected to bring about this project.
First, we recognized that today’s care model is linear and centered on a break-fix model—you need care, you call a doctor, you schedule an appointment, you visit the doctor, you get a treatment, and you’re done (and possibly repeated over several iterations). This conveyor belt model (described in more detail shortly) generally fails to provide care or make care available during periods of wellness. It also centers on a physician visit, which is more expensive and less convenient to the patient than many of the alternatives.
Second, we recognized that this model works best in large cities, where there is a “critical mass” of hospitals, primary physicians, specialists, labs, and other services located near each other. What about the many remote areas or smaller towns and cities? Can these folks get the care they need without traveling back and forth to the Twin Cities or Rochester? Or do we have to build small and inefficient hospitals in remote places, fully staffed and equipped, to meet patient needs?
Third, the health care world is moving rapidly toward the pay-for-value and the accountable care organization (ACO) models we noted in Chapter 2. With pay-for-value, we can’t afford to keep full-time physicians in place to serve locations with low population densities, but we could afford to keep professionals there who had lesser credentials and were connected electronically to Mayo Clinic in Rochester or one of Mayo’s other locations.
We took these three issues and formed this hypothesis (the scientific component of our fusion method): “By developing a flexible, team-based model of care with a convergence of the right technology, access points, facility design, and team approach, we could deliver the right care at the right time in the right place by the right person in the right way—all at the right cost.”
Thus, the Optimized Care Team model was born for further research and implementation as part of the larger Community Health Transformation project.
Defining the Care Model
We had a sense that the current model of care could be improved, but we wanted to approach it in a way that was inherently measured, scientific, and demonstrable to the current constituents.
We felt it extremely important to get to a visible, working prototype very quickly (Think Big, Start Small, Move Fast). As such, we worked fast to develop an alternative care model that could be tested in practice, exercising a care team concept and providing a real patient care experience. In this case, the experiment was used to do most of the research.
The first step was to identify the “current” care model and evolve it to a future model for testing in research and in practice. The current model is illustrated in Figure 8.5. This model is recognizable and doesn’t need much description—it is face-to-face, it is generally but not always with a physician, it is reactive, and it happens only when someone is sick and needs a remedy.
FIGURE 8.5. CURRENT MODEL OF CARE: A CONVEYOR BELT
The “future” care model would “wrap around” the patient in all phases of sickness and health, with all levels of medical disciplines in contact at the right time and in the right context. In this model, care would be delivered by a team, either in preventative or reactive mode (Figure 8.6). The “team” would include a blend of M.D.s, R.N.s, L.P.N.s, and R.N. care coordinators, nurse practitioners, medical residents, community health workers (C.H.W.s), and later on, “wellness navigators.”
FIGURE 8.6. THE FUTURE MODEL: WRAPAROUND CARE
While the details are beyond the scope of this book, the operating model included a group care team calendar with team “huddles” to discuss patients, to triage care, and to assign care to the appropriate team member.
In Sight, It Must Be Right
Kasson, Minnesota, is a small town in Dodge County of about 6,000 residents with a typical Midwestern downtown, a unique stone-based water tower built in 1895, and an annual Festival in the Park celebration. It also just happens to be 15 miles west of Rochester, and it also happens to have a clinic that is part of the Mayo Clinic Health System.
To get quickly to proof of concept, we constructed a series of on-site experiments at the Kasson Clinic. We wanted to see what would happen if we created a care team and managed the patient flow from all of Dodge County according to the needs of the patients and the skills and locations of the providers. We wanted to see how the flexible care model would work, and we wanted to measure who actually saw the patients, what care was given, and how the team and the patients felt about it. We learned from our experiments that there was significant potential for utilization of the broader health care team. Patients in many instances could be better served with nurse-only visits or nonvisit care or with a more integrated care team experience (Figure 8.7).
FIGURE 8.7. RESULTS: WHO NEEDS TO SEE THE PATIENT?
In these pilots, as Figure 8.7 illustrates, we found that, set up properly, the flexible, or optimized, care teams allowed our staff members “to operate at the top of their licensure,” that is, to give care according to their skills and the patients’ needs, all of which could be decided on in the huddle prior to the patient arrival. By making the best use of the R.N.s and N.P.s, only 6 percent of the incoming patients needed to see a doctor initially. There was also less paperwork, and since much of the flow was worked out in advance, there was a smoother patient flow. That was because there were fewer “stop-and-starts” as patients had to pass on from one care provider to another.
As is the norm, we also collected qualitative feedback from care team members and patients, to make sure the data wasn’t steering us in the wrong direction:
“What may seem like a complicated patient may not be if you know the person well.”—M.D.
“I’m glad my doctor didn’t need to see me. That means I’m not sick enough! I know they sent the right person to take care of me today.”—Kasson patient
“When the larger team is involved, it really gets the physician to think more broadly.”—M.D.
“I’m really enjoying hearing the doctor’s conversations and learning more about how they think. It will really help me to pull out useful information during the rooming process.”—L.P.N.
“It feels like there is a team of people working in my best interest. It feels efficient and like there is good communication.”—Kasson patient
From these initial experiments, we learned that there was more work to do to optimize the role of the clinical assistant (C.A.). C.A.s are central to the operation of the team and the scheduling of resources; with more tools and training at their disposal, we could streamline the process. As we “morph” the model into more of an ongoing care (versus acute care) model, we will pay more attention to scheduling nurse contacts with patients in “wellness” as well as in sickness—and so-called nonvisit care. Finally, we designed a workspace ideal for the OCT and the team huddle (Figure 8.8).
FIGURE 8.8. OPTIMIZED CARE TEAM WORKSPACE
As we explored the spectrum of “well” care between acute sickness events, and how to provide the best, most cost-effective ways to achieve this, an idea popped into our heads: What if we set up some volunteers—community-minded folks connected with our clinical staff—to help our patients maintain and improve their health without medical intervention? We felt that a “wellness navigator” could provide a contact point for these individuals and point them to good community resources, including our health resources where needed. The wellness navigator would provide someone to talk to, sort of a “health concierge” available for community use. We developed the Wellness Navigator project, and we fully piloted a new wellness navigator team member who could eventually be a core team member.
We (and our constituents) were very encouraged by these pilots, and we could see some real merit to this approach. There’s a lot of potential synergy between the OCT model and our Mayo Practice and Connected Care platforms—all three connect well to deliver our greater vision.
Guiding the Implementation
Finally, we put together a 64-page booklet, called Optimized Care Team: Implementation to Adoption Toolkit 1.0, to assist any clinic anywhere in the system to implement the Optimized Care Team model.
Mayo Clinic Patient App
Finally, we finish up with a case example—the Mayo Clinic Patient App—that demonstrates the “here, there, and everywhere” vision—and happens to be a CoDE project originating in the practice and expanding beyond. Here’s the story.
In order to meet the growing expectations of Mayo Clinic patients and to enhance the overall patient experience, the Center for Innovation gave a CoDE award to create a unique, simple, portable electronic environment for Mayo guests and visitors. Two areas at Mayo came together to apply for the award—Public Affairs and IT Workstation Support Services—to create an app, which would be an electronic tool that would act as a Mayo Clinic concierge to enhance patients’ visits by helping them navigate around Mayo.
The app was made available in a soft launch in the Apple Store in May 2012 (Figure 8.9). In its initial launch, the Mayo Clinic Patient App provided an easy-to-use tool to navigate a visit while at a Mayo Clinic campus. The Version 1 app included instructions for finding one’s way around the clinic and hospitals, community information, and directions to local restaurants and entertainment. On the first day, the app was downloaded 1,000 times.
FIGURE 8.9. MAYO CLINIC PATIENT APP
In typical CFI fashion, at the post-launch, the small CoDE team continuously tracked comments about the app on social media. On the day after the download, a patient made a comment on social media that caused the team to dig deeper and learn more from him. The comment was this: “You guys really missed the boat on this one.” The team reached out, and it turned out that the patient was blind, and he called out some serious flaws that prevented the app from being usable by those with impairments. Some behind-the-scenes IT magic and a few weeks of effort was what it took to fix it up and offer a platform that would be accessible for everyone.
The team kept working in the manner so common for the CFI. How can we keep revising the product or service to help more patients? What additional user needs are out there that we can try to serve? Additional features were quickly added to the app, including the ability for the first time for Mayo Clinic patients to have 24/7 access to their personal medical records, their lab results, appointment schedules, and other services using their individual Patient Online Services account (easy to set up—a patient just needs his or her Mayo Clinic patient number). The team further expanded the functionality to include the latest news, publications, and health information such as videos from Mayo Clinic. All of these improvements made it a true “here, there, and everywhere” cross-platform tool. Just as important, the establishment of the Mayo Clinic Patient App and its successful patient- or person-centered navigation have set the stage as a standard for other Mayo apps to follow.
It even made a splash at Apple. Tim Cook, Apple’s CEO, introduced the 2012 Apple Worldwide Developers Conference on stage by highlighting the Mayo Clinic Patient App. Since then, the app has expanded to the Motorola DROID and to virtually all other mobile platforms, and it has been downloaded almost 200,000 times. It is Mayo Clinic’s most downloaded and used app!
Hopefully these snapshots have given you a good sense of how we approach our projects and our 21st century model of care vision. Instead of summarizing this chapter here, we’ll take you straight to Chapter 9, our final chapter: a “user’s guide” to implementing successful transformative innovation in complex enterprises—from our experience and our hearts.