Tuesday, April 19, 2011

We DON'T need more PCP’s. We need operational efficiency.

I'm getting a little tired of reading articles that bemoan the shortage of primary care providers (PCP's). Don't get me wrong, I completely agree that we're headed for a crisis. I also agree with the issues that are causing this mismatch. We do have an aging population that will require an increasing level of care, and since provider compensation is clearly not aligned with the total healthcare value created, PCP’s are often undervalued and undercompensated relative to their specialist peers. We pay specialty providers much more to replace a diabetic's failed kidneys than we pay PCP's for the much more valuable preventive services that could keep the person from developing diabetes and kidney failure in the first place. With that kind of misalignment, of course medical students (or any other rational being) are going to look favorably on higher paying specialties.

But so what? I mean, it’s one thing to be able to point fingers and say “Wow, there’s a problem here. What a mess!” But anybody who’s familiar with healthcare knows about both of these issues. The question is what are we going to do about it? 

Sermonizing to patients and advocates that the “grey wave” of aging Americans is coming, encouraging them to snap up available PCP panel space, and admonishing them to “be sure to find a reason to get in to see that primary once or twice a year” is certainly not a solution. In fact, if anything, this recommendation will actually exacerbate the problem. The last thing we need is to further burden an already strained primary care system with additional unneeded demand from patients who dread “falling off” their PCP’s roster.
From a more productive standpoint, realigning compensation to promote the preventive activities of primary care would be a tremendous value, and many people are looking for solutions to that problem. As an example, a major part of what ACO’s are being created to do is to realign incentives and promote primary care and prevention. The problem here is velocity of change. Whether you believe that ACO’s are a step in the right direction or not, under the best of circumstances, they will take a few years to show successes, a few more years to expand, and even more time before those changes start shifting medical students decision making about what specialty to enter.

The truth is, there is no way to increase the availability of PCP’s in a timely way. But that’s OK, because what we need isn’t more availability of PCP’s, it’s more availability of primary care services. That’s a subtle distinction, but an important one. Another way to say this is that what we really need is not more providers, but a better way to meet healthcare consumer’s needs, and in order to do that well, the we’ve got to better understand what those needs are.

When a patient wants the healthcare system to manage a complex chronic disease like diabetes, a their needs usually can be expressed something like, “build a trusting relationship with me, know my medical history well, and create an effective long-term care plan”. Traditional primary care often does a pretty good job of meeting these kinds of needs. On the other hand, when the same person wants the healthcare system to manage a itching, burning but minor rash, their needs are different. What they want is more like, “evaluate me now, make sure I don’t have something more serious, get me the medication I need, and make it happen conveniently and cheaply”. Our traditional primary care system is notoriously bad about meeting this set of low acuity, high volume “On-Demand” consumer needs.

The good news is that there is are potential solutions that provide more availability of primary care services and better meet patient’s desire for On-Demand care. These solutions are likely to involve using web-based technology to improve the efficiency of lower cost mid-level providers, like nurse practitioners (NP’s) and physician assistants (PA’s). Let me tell you about the last time my wife had a nasty cold. It had been hanging on for about 8 or 9 days. For the last two or three, she had been feeling worried that it might have progressed to something more serious. She felt anxious and lousy. When she finally decided to go see her PCP, she had to pack up the kids (who were also possibly not feeling particularly well) and get them into the car, drive them across the city to get to the doctor’s office, herd them into the clinic and manage them during the 20-30 minute wait in the lobby. Once she got called back, she had to round up the kids and get them back into the treatment room, and try to settle them down for long enough to have a productive conversation with the provider, only to be told that she really did just have an uncomplicated viral URI, and that there is no prescription that would be particularly helpful (but the Tylenol and Robitussin DM she already had in our medicine cabinet back home would work just fine). At this point, feeling sick, exhausted and frustrated, she herded the kids (who have become increasingly bored and unruly) back to the car, and dragged them and herself back home. All in all, from start to finish, it was a harrowing 2 and a half hour ordeal. The sad thing is that this is actually considered a “good” experience, since she was able to get a same-day appointment with her PCP!

Not only was this a terrible patient experience, it’s also a strikingly inefficient workflow for the provider. In these kinds of situations, the diagnosis is more than 95% dependent on history taking. Assuming that the history is consistent with an uncomplicated viral URI, there are no physical exam findings that will change my diagnosis and treatment plan. I spend 10-15 minutes trying (between child-induced interruptions) to ask the patient a series of routine clinical questions, and then document my findings in the medical record. It’s not a particularly challenging or rewarding professional experience, since the questions and diagnosis are very algorithmic. The one up-side is that I can bill roughly the same for this visit as I did for the poorly managed diabetic with refractory hypertension and dyslipidemia (We’ll talk more about this problem in a future posting…)

Basically, we’ve created a “lose-lose” situation for both patient and provider. We need to stop expending the valuable resource of PCP time and expertise on issues like "sniffly noses", uncomplicated low back pain, and ear aches. PCP visits for these kinds of issues just aren't an efficient use of healthcare resources, and more than half of primary care visits are for these low acuity, algorithmic type of conditions.

Now imagine a slightly different situation. Instead of suppressing her growing anxiety for two to three days because of her dread around the perils of a PCP visit, she goes to a website and from the comfort of the couch at home, she answers the exact same routine clinical questions.  Using basic branching logic, the web program arrives at the exact same algorithmically determined diagnosis and treatment recommendations as I would have in the clinic. The questions, patient responses, and clinical recommendations are passed to a licensed mid-level provider (perhaps on their smart phone) who can review and approve them (or make different recommendations) in less than 5 minutes. When the provider submits their approval, the entire encounter is automatically charted in its entirety with far greater consistency and accuracy than would be possible in a traditional clinical encounter, and the whole transaction takes less than 30 minutes.

Doing just the math, we’re talking about improving provider efficiency by more than 75% (5 minutes vs 20), and shifted care from a physician ($195,000 per year) to a nurse practitioner ($110,000 per year). Now multiply that out times the 30-50% of patient visits that could be managed with this kind of care model. That’s huge healthcare savings. On top of that direct savings, the system has also created value by allowing providers to practice closer to the top of their license. In other words, we’ve freed up the physician’s time to manage more complex patients that their additional training makes them more qualified to manage. Just to put icing on the cake, let’s not forget that we’ve also reducing the patient time cost by 80% (30 minutes vs 2 and a half hours), and improved the patient experience, well, nearly infinitely.

If this kind of solution were put into place (and online tools like the one I’m talking about already exist), and even 10% of patient visits shifted to this type of care, that would increase primary care availability by enough to greatly reduce the projected provider shortages (without increasing the number of PCP’s at all). In the end, there are more than enough resources already being spent on healthcare. In my opinion, it’s time we started figuring out how to allocate them in a more responsible way, instead of just throwing more money at the problems we all know exist.