What Primary Care Has to Learn from Behavioral Health
Game changer and 2013 MacArthur fellow Jeffrey Brenner, MD is a primary care physician and founder and executive director of the Camden Coalition of Healthcare Providers. Best known for his “hotspotting” concept, Jeff is creating a healthcare delivery model to meet the medical and social service needs of the most vulnerable citizens in impoverished communities. He returns to the National Council Conference by popular demand. Jeff talks to Meena Dayak and Heather Cobb at the National Council on hotspotting, standardizing, open access, and other disruptive changes in healthcare.
National Council: What are the greatest challenges in primary care today?
Jeff: Primary care has not fundamentally adjusted its operational, clinical, and business model to keep up with change. Right now, primary care docs in America come in and run from room to room to room in meaningless 15–30-minute encounters, which don’t add up to anything. We don’t have the workforce, office structure, protocols, standardization, or data systems we need. We’re so far away from being able to deliver better care at lower cost in our understaffed, undercapitalized, and underperforming offices. The big challenge is to accept that what we’re doing isn’t working and to be ready for the disruptive change that it’s going to take to really do the job.
National Council: What does that disruptive change look like?
Jeff: Primary care docs who are making $150,000 – $175,000 a year are titrating meds for blood pressure, cholesterol, and diabetes. The data says we don’t do a good job of it. We don’t listen very well when we adjust the meds and we don’t give good patient education. So it would be completely legal and possible to have RN and BSN level nurses adjust medications, as they do in inpatient settings and intensive care units.
Routine work in primary care needs to be standardized, protocolized, and delegated. Instead of treating head colds, primary care docs should be treating our sickest, most complex, and most challenging patients. And even in doing that, we’ve got to have structure — and behavioralists and community health workers in the mix.
National Council: So standardization will lead to better care?
Jeff: Yes. Look at car manufacturing. Up to the early 1900s, cars were made by artists in studios. They were expensive works of art and not highly reliable. Then Henry Ford came along and standardized the process and drove down cost by building cars on the assembly line. Craftsmen ended up doing really boring work, like putting a tire on every day. But they were able to drive down the price enough that wages went up and every assembly line worker could now own a car. Later much of the boring work was automated. It took a lot more sophistication and education for workers to not be on the assembly line. And now, front line assembly workers have input on how the cars are made.
If we want perfect healthcare every day for everyone, we’ve got to standardize, protocolize, and delegate. We have to stop the wrong people with the wrong training doing the wrong work and being paid the wrong kind of money.
National Council: If a primary care doc is doing the artisan’s work, is there a point at which s/he’d hand off a patient, say with depression, to specialty care?
Jeff: I would even go further upstream. If the office is running well, the patient has a PHQ-2, administered by a medical assistant, which results in a PHQ-9 administered by another staff member, that triggers a standard protocol where primary care provider comes in and says, “You seem like you’re having a hard time, what’s going on? I’m worried that this is going to affect your health, and I would really like to have my behavioralist come in here and spend some time with you.”
None of this is rocket science. It’s going on all over the country in these new and more advanced models of care where you’ve got behavioralists embedded in primary care.
You would have a series of warm handoffs and clear protocols, clear delegation, and clear workflows, so that we’re not missing stuff. You miss a lot of stuff in healthcare!
National Council: Speaking of embedded specialists, can you describe your work at the Camden Coalition where there is a lot of care coordination?
Jeff: The Camden Coalition was intended to help the city work together to deliver better care at lower cost for patients. There is so much alphabet soup in healthcare and social services and we were not always that good at coordinating with each other. So, we launched citywide monthly meetings where we anonymously discuss cases of complex patients we’re working with to ask for help from all community partners. It’s been a powerful tool for collaboration.
National Council: And care coordination is also intended to address social needs before they become medical problems?
Jeff: Right. If a person doesn’t have a roof over their head, if they don’t have a meal, if they’re a victim of physical or sexual abuse, if their household has a lot of stress in it, if their kids’ school is not safe, then that’s going to impact their health. And I think we see that very clearly in the patients that we work with, that health is more than just the pill that we’re giving you or the hospital that we put you in. It’s all the other parts of your life and whether they’re working in harmony.
National Council: Speaking of people with high needs, what’s the latest on your hotspotting initiative?
Jeff: In hotspotting right now, we’re using datasets to look for outliers, with the hypothesis that if you’re an outlier in a healthcare dataset, you’re probably not getting very good care. For example, if you rank ordered everyone who got a CAT scan at a hospital from those who got the most CAT scans to those who got the least you can look at the outliers, who’ve had an inappropriate number of CAT scans. We found a woman who’d had 79 CAT scans over five years and it turned out she had anxiety disorder. So, that’s an example of hotspotting, of stretching out the dataset, looking for extremes, and then focusing in on those patients. And it’s really a root cause analysis, an indicator that something must be going wrong in healthcare.
In Trenton, New Jersey they found someone who had been to the emergency room 450 times! That’s an extreme outlier. In Camden, we found a building that had $15 million in receipts over five years for the people living there.
National Council: So once you identify that building, what did you do to stem high usage?
Jeff: We’re still figuring that out. In one high-cost building housing 40–60-year-old patients with disabilities, we offered a lot of programming — the Stanford Model of chronic disease training and education, yoga, art therapy, patient education. We even set up an office on the first floor, and it made very little difference in total spend and utilization of the building. Later, we learned that three-quarters of the building never goes to a hospital or ER — in fact, only 30 patients are in and out of the hospital. But earlier, we missed this. What we didn’t have was the list of everyone living in the building. So, you know, if you’re really sick and disabled living in this building, you don’t necessarily come down, take the elevator down to the first floor, and hang out in the room there. So, I think it really highlighted for us that you’ve really got to have all the data to know what you’re doing.
Now, we’re using real time data flows. Every morning we get a list of who’s been admitted to three area hospitals so when one of those patients living in that building gets admitted, we go right to their bedside. Without the data and technology, we’d be flying blind.
We need setups where providers get a list of their hospitalized patients every day. The quickest and easiest way to save money in America is just to focus on who’s in the hospital right now, and catch them every day when they’re leaving. We don’t do a good job of that. Providers don’t even know their patients have been in hospital. Sometimes patients can’t even get primary care appointments in a timely manner. Every hospitalized patient should have an appointment with his or her primary care provider within seven days of leaving the hospital. If we could pull that off, it would be a huge accomplishment.
National Council: Speaking of difficulty getting appointments, you’re an advocate of open access scheduling — how’s that working for you?
Jeff: I think all of American primary care should be open access scheduling because an appointment three weeks from now isn’t very meaningful. If you want an appointment three weeks from now that’s fine, but most people want an appointment when they want one.
We got one large office to switch to open access. And it was really a business operations change management problem. You’ve got to know your baseline data. You’ve got to educate your staff, plan the new workflow, train everyone in the new workflow, and then make the change. Typically when practices switch to open access scheduling, they have 20% excess capacity.
National Council: What can behavioral health providers learn from these disruptive changes that primary care should be preparing for?
Jeff: Actually, our behavioral health colleagues are about 30 years ahead of us. I hope primary care can learn from behavioral health.
When psychiatric care was deinstitutionalized, behavioral health did heroic work to figure out how to deliver better care at lower cost and evolved some creative models. The different tiered interventions provided in behavioral health and ways to engage patients are really remarkable. That is a perfect foreshadowing of what’s about to happen now on the medical side. We’re about to deinstitutionalize primary care.