Innovation, Primary Care Style

Andrew Morris Singer PCP

On a recent evening at Harvard Medical School, the Primary Care Innovation Challenge and Pitch-Off ,sponsored by WellPoint’s American Resident Project, brought together six finalists, primary care luminaries and trainees, and a host of hangers-on and camp followers for a couple of hours of demos and discussions. The tenor of the evening, which was in many ways a pep rally for primary care – not that there’s anything wrong with that — was best captured by the rhetorical question posed by Asaf Bitton to the primary care practitioners and trainees in the hall, “Are you going to be a playwright or a critic?”

The hoots and hollers in response made clear that these are not your grandfather’s primary care docs. The call to action was echoed by many of the speakers, notably community organizer turned primary care physician Andrew Morris Singer and Dennis Dimitri, both advocating for, well, advocating for primary care.  Bitton’s opening also included the exhortation that proved to be predictive of the winner of the top honors from among the six pitches: Innovation in primary care is not about the technology; it needs to enable better human care.

The projects presented by the finalists ranged (in order from the techiest to the least techie) from using natural language processing of free-text notes in EHRs to improve quality metric capture, to interventions combining app-based data collection with live clinicians and care managers interacting with patients in person or remotely (Heart Failure 2.0 and Twine), to studying “positive deviance” in patients and bottling their secret sauce, to significant re-jiggering of the primary care clinical education experience in medical school (Baystate/Tufts and Beth Israel Deaconess/Harvard). (For the completists among you, here’s the full #PCC14 tweetstream.)

Here are some of the highlights:

The fact that in late 2014 an innovation showcase featured a project based on natural language processing of unstructured data in EHRs should cause, at the very least, widespread wailing, gnashing of teeth and rending of garments. I found it appalling to hear that primary care physicians couldn’t be expected to put all patient data into the structured data fields – they complain, “too much point and click,” according to Howard Haft – appalling because the EHR industry, this far into Meaningful Use, has not figured out how to make the workflow for PCPs work, so that more important patient information may be captured in structured data, or so that the EHR itself may be able to apply some NLP and logic to the free text entries in order to unlock their content for the benefit of the patient.

Twine Health, John Moore’s startup, connects patient and provider – not just the data – thus empowering patients as apprentices, taking an active role in their own care. (“No one ever gave me the chance to be in charge,” a patient using the platform said.) One arresting example of how this works: 100% of patients who were enabled to connect with case managers between appointments using the Twine Health platform were able to control their hypertension within three months. Only 25% of patients without these tools were able to control their hypertension – even after a full year. The connected patients were able to easily review side effects of their meds with their providers, who were able to start them on another med. Discontinuing meds because of side effects is apparently a key factor in low success rates in controlling hypertension.

Cole Zanetti, from Dartmouth Hitchcock, suggested that we focus on the healthy rather than the sick – that we need to keenly observe the patterns of successful behaviors among the “positive deviants,” the healthy cohorts in our patient populations at the far end of the bell curve of health, to figure out what makes them healthy and apply that learning to the care and coaching of the less healthy. He made a compelling case for this initiative.

The winning innovation, in the end, was one of the most low-tech and perhaps the most subversive,  even though its initial roll-out has been supported by a HRSA PCRE grant. It is the Baystate Medical Center/ Tufts University School of Medicine internal medicine residency program, expanded under the grant, and also reimagined. Gina Luciano explained:

Upon graduation, residents are prepared for inpatient medicine but not for primary care (PC) careers in part due to limited ambulatory time and opportunities during training. [Under a] grant we created additional ambulatory time/experiences outside of the core ambulatory training clinic for our PC residents. We eliminated month-long inpatient electives and transformed them into year-long ambulatory continuity experiences. Residents spend 7-8 months a year in ambulatory rotations during which they participate in traditional continuity sessions as well as spending substantial time in subspecialty ambulatory electives and community advocacy projects. Residents commit to 4-5 subspecialty electives during each calendar year. Through extra ambulatory time, trainees enhance their learning of complex disease management from subspecialists, improve their ability to formulate appropriate consults and form collaborative networks with subspecialty providers.

A surgical rotation can happen in a month, because it consists of intensive, in-hospital patient encounters, with some pre-op and post-op care. Primary care – which requires long-term collaborations with other clinicians and with patients to be successful – has long been shoehorned into a training calendar that makes sense for other specialties. Now this is changing. After the grant runs out, the additional residency slots will have to be funded by the training program.

A mere schedule change winning an innovation pitch contest? You may ask why. Well, the focus on the internal medicine residency experience is in fact a big deal, given the historical emphasis placed on training physicians for careers in specialty care.

Marci Nielsen, CEO of the Patient-Centered Primary Care Collaborative (PCPCC) moderated the pitches. I asked her about the import of honoring this entry. She replied:

That the award went to a “low-tech” intervention focused on changing how we train primary care residents speaks to the importance of improving how we train the next generation of health professionals. As was stated in the IOM’s recent report on Graduate Medical Education (GME), we should be training the US physician workforce in a way that is accountable to America’s real population health needs: i.e. training physicians to treat patients and families “where they are,” in community settings, helping to empower patients and their families to deal with chronic illness and one that values all of the health care team (from physicians to medical assistants and all the professionals in-between).  This is difficult to do in the current training environment which links GME dollars to inpatient hospital settings.

I also asked Marci whether the changes in medical education for students interested in primary care will have a real effect on the cost, availability and quality of that care. Her take on this:

Yes, this is an innovative challenge, and one that is sorely needed.  There is a growing body of evidence that when we treat patients and their families in advanced primary care settings (such as patient-centered medical homes) we have the capacity to improve on the care provided, as well as improve on provider and patient satisfaction.  We see cost savings, largely as a result of keeping people out of the emergency department and the hospital.  We also see improvements in overall health outcomes as a result of better managing chronic conditions in partnership with patients (see the PCPCC’s annual report from last year).  But none of this easy or simple.  Asking primary care practices to transform and offer the kind of care that puts patients’ needs first requires changes in reimbursement/payment (away from fee for service and towards fee for value) and — as Gina pointed out — a focus on how and where patients really live, work, and play: in the community.

Eric Weil, Associate Chief for Clinical Affairs, General Medicine, Massachusetts General Hospital, moderated a panel discussion following the pitches. Later, he agreed that innovation is not all about tech, noting that “technology will solve some of the problem – the rest is human factors.” He ticked off three factors vital for the success of hi-tech innovations: they have to be (1) operationalized without negatively affecting workflow, (2) have broad application and (3) made affordable. Further, Weil observed that “it is changes in process and culture, changes in strategy and teaching, that will make the difference in the long run. The models in primary care and more broadly in medicine just aren’t working any more. However, the underpinning values that represent the practice of medicine remain the same. If we want to preserve the values, the models of care delivery need to change.”

The hi-tech entries, in the end, were not seen as being quite as innovative as the low-tech reinvention of physician training.

What do you think?

David Harlow is an attorney and consultant with a national practice focused on health care innovation and health data privacy and security. He lectures extensively on health care law and policy to attorneys and to health care providers. He has served as Deputy General Counsel of the Massachusetts Department of Public Health. He writes at HealthBlawg, a top health care law and policy blog.

7 replies »

  1. But what can we do right now to reduce patients’ waiting times while we wait for our current health system to transform into something resembling sanity?


    We are trying to tackle the problem of redundant medical history taking with an innovative solution called Project C.A.R.E.S. (Communities Aided by Research and Education Solutions).

    In brief, he proposes to provide patients at with automated check-in kiosks that can transfer self-reported health history information directly into their electronic medical records, and measure the effect of this intervention on waiting times and patient satisfaction. If this pilot project is successful, it could inspire other underserved health clinics across the country to do the same. Please read more about the project and consider helping us reach our 30-day fundraising goals to make this project a reality.

  2. Civis Isus,
    Based on how the blog described my proposal I can understand your critique and your frustration that someone would claim what was described as an original idea. I would be right in your camp if my proposal was exactly what was described above. I think that despite the best efforts and great work by Mr. David Harlow to consolidate all of our proposals inevitably portions of the proposals were going to be left out. Unfortunately some of the portions of my proposal that were left out were the key elements that made the proposal distinct.

    I just wanted to clarify a few of these points.
    1) My proposal was not about focusing on those that are healthy, in fact my proposal suggested to focus in on those with high disease severity and high social determinant of health risk factors. This would identify a “high risk and high disease severity” patient population that suffers from a chronic disease. Once this population is identified you would look over a 3-4 year period to identify the frequency of those “high risk” patients being admitted to the hospital due to a chronic disease exacerbation. This would create a bell curve frequency plot.

    2)Once you have this frequency plot of high chronic disease severity and high risk patient population you then identify those that are the “positive deviant” patients. Positive deviant patients are those patients with high disease severity, and high social determinant of health risk factors that have been admitted either zero or very few times due to their chronic disease exacerbations despite having exceptional risk. Essentially these are patients who figured out ways to manage their chronic disease conditions for years despite all of these obstacles.

    3) Once these patients are identified primary care teams would work with nurse navigators along with high utilizer patients to perform qualitative research to identify the patters of behavior and strategies used by positive deviant patients to achieve this success.

    4) Once these behaviors or strategies are identified, the positive deviant patients would teach the entire high risk patient cohort, through simulation based learning, these behaviors and strategies. Our health system is also looking into designating these positive deviant patients as community health workers so that if we pilot the process with dual eligible patients they would be able to get reimbursed for the preventive services they provide and have a sustainable way of continuing the work.

    I would also like to add I am not one to claim the positive deviance approach as my own. Amazing people have been doing this work for decades even within healthcare. Thus far very few people have considered its utility within the patient community and I have been working with the giants that have come before me to work on this within my hospital system. The PD concept came from Jerry and Monique Sternin. Here is a link to their website: http://www.positivedeviance.org/

    I really appreciate your critique and frustration with people claiming novelty or that an idea is their own. I am a firm believer that ideas are a shared concept that wouldn’t exist unless without those that helped shape our reality. If you have any further questions or comments I will do the best I can to respond. Thanks again.

  3. Is Cole Zanetti properly attributing the source of his sage suggestion “that we focus on the healthy rather than the sick” to, at very least, Dee Edington (“keep the well well”), and probably Ostler, if not all the way back to Hippocrates? Or is he claiming he thunk it up on his Health Care Jam Sleepover Party ownsome?

    Yeah, and get off my lawn, too

  4. This is fantastic. I just wanted to clarify that positive deviant patients would be patients with high disease severity, and high social determinant of health risk factors that have figured out was to manage their chronic disease conditions for years despite all of these obstacles. Finding the “Bright Spots” among those patients that are created risk for being admitted to the hospital due to chronic disease exacerbations. This was an amazing event and just the begining of the amazing innovations that are coming out of primary.

  5. Plenty of good points, especially on revising Primary Care training to reflect the needs of patients rather than hospitals. Abandoning a tradition of apprenticeship in favor of bloated hospital based training programs has not served patients or practitioners. I remain suspicious of the PCPCC and PCMH as they have directly led to rising overhead and have done little for the sake of solvency. Outcomes are fishy at this point. There is nothing particularly wrong with fee for service, unless it involves prejudicial undervaluing of Primary Care. Numerous services are compacted into what is known vaguely as “office visit” ; or longitudinal care that is basically taken for granted. Unless reimbursement starts lining up with the realities of overhead, value based care will simply become the new “office visit”, where expanding roles are taken for granted in a flat payment system. Get the cart behind the horse and plenty will change.