I've been taking a tour through the world of value-based care, which is the most recent place where the Centers for Medicare and Medicaid (CMS) have changed the rules of the game for reimbursement at the organizational level.
Value-based care is defined as as positive health outcomes divided by cost.
It's a new and interesting body of knowledge to learn about, and as with most changes in healthcare, there are abundant pitfalls. Each time the goalposts move, the rules of the game change substantially.
Understand and consider the rules and their underlying incentives, and you stand to come out ahead in the game. Ignore them at your peril.
The legacy system for financing health care in the US is the fee-for-service model. Medicare reimburses each provider based on the service performed, not necessarily on whether that service improved the patient's health.
Our fee-for-service system has led us to a place where providers are incentivized to deliver as many services as possible, but are not held accountable for a) the cost those services incur to the system as a whole, nor b) whether those services improve health for the patient in question.
CMS is trying out new models that explore whether better health can be obtained at lower cost, as many other developed nations have managed to achieve. One of several models being tested is the Medicare Shared Savings Program (MSSP). Under MSSP, a provider-led organization is tasked with caring for a population of Medicare patients.
These patients are predicted to cost Medicare a set amount based on a review of the prior codes used to bill services for their care. Those codes are used to "risk-adjust" the expected cost for the patient, where those whose codes demonstrate that they are sicker are predicted to cost the system ore than those whose codes suggest their baseline level of health is less ill.
The total population attributed to the provider organization is risk-adjusted and then benchmarked to an expected cost. To use an example, his new primary care doc documents on his initial encounter that Joe has heart failure, diabetes and hypertension. Medicare expects to pay $10k per year for someone his age with those illness codes - that's Joe's benchmark.
If the provider organization cares for Joe and spends only $5k, under MSSP, Medicare and the provider organization get to split the difference between the actual spend and the expected spend - the shared savings - and everyone wins.
If the provider organization instead spends $15k to care for Joe, they have to pony up the additional cost - a two-sided risk arrangement where the organization has skin in the game.
The codes used at the time the patient is attributed by Medicare to a particular provider organization are critically important, because the fine print rules of Medicare use only this initial set of codes to predict what the patient will cost the system.
If Joe is coded more carefully by his new primary care physician (whom we will assume is documenting truthfully) to have heart failure, pulmonary hypertension (a consequence of his heart failure), diabetes with nephropathy, and hypertension with arteriosclerosis, Medicare might risk adjust Joe's benchmark and expect to spend $15k per year for someone this sick.
If the provider organization spends less, their take is bigger, and if they overspend, they swallow the risk.
The upsetting reality about Medicare rules is that from the moment Joe's illness is coded by his physician, Joe or any other patient is benchmarked to the expected cost associated with illness severity at the time a patient was attributed.
What if Joe's doc was rushed on the first meeting, and codes more carefully on the second encounter? According to Medicare's rules, they stick with the initial risk adjustment. You want a "do-over?" Better go back to the school yard.
The stakes are high and the rules are Byzantine. You get one chance to do it right, and if you flub it your income is adversely affected going forward.
Comments 7
I started med school in 1981, the year of the HMO and the year the HMO began to fail. This version will also fail because it’s just more cost shifting. It’s advertised if you choose your patients wisely you have a chance to win. You have no chance, in the aggregate.
A chance might be developed however using AI to always choose the highest probability healthy patient, (as in my Galton board arguments) otherwise the game is rigged against you.
Author
Gasem,
Nixon signed the HMO into law the year I was born, in 1973, and they’ve been a reality ever since.
I get a little more excited about value-based care than I did about HMOs because there’s a triple emphasis on outcomes/quality, cost, and patient experience that seems to align goals better. HMOs by and large focused on cost alone, and there was certainly some shifting from the patient spend to the infrastructure spend there.
Value-based means, among other things:
1) The patient gets to rate the experience, so it reorients care to the patient rather than the provider’s convenience
2) The patient has to get healthier / not get sicker
3) You are incentivized to do both by addressing gaps in a fragmented system
4) There’s a two-sided risk element in a shared-savings arrangement. You do well, you get paid better. You do worse, you eat the difference in cost.
The strategic thinker in me likes the challenge in that.
An example might be: grandpa keeps missing his doctor’s appointments. Doc’s admin calls grandpa, no answer, efforts end there and grandpa gets sicker / has a prolonged ICU stay with sepsis later that year.
For an ACO or other value-based care system, grandpa keeps missing his doctor’s appointments. Doc’s admin calls grandpa, no answer. Social worker visits grandpa, sees issue is he can no longer drive but still functions well. Social worker meets with case manager, who arranges transport to/from doc’s office and reschedules appointment. Grandpa makes it, he’s happy, doc catches UTI before grandpa gets septic, system spend is down, doc is paid better, Medicare saves money, grandpa likes his new benefit of rides to and from his appointments.
Don’t get me wrong, there is plenty to critique and no shortage of things to improve. But the effect, by realigning the incentives, has the potential to deliver incrementally better, more patient-oriented health outcomes.
I see promise in that, even if I may be in the minority.
Fondly,
CD
Value-based care is a step in the right direction for the patient, agreed, but it is absolute trash for the quality of life for the primary care physician. How do you code everything to the umpteenth degree AND explain why grandpa doesn’t need a lumbar MRI AND order all the vaccines AND get a great patient satisfaction score AND do it all in 30 minutes? 20 times a day? Oh, and now a good portion of your income hinges on hitting those vaccine numbers, satisfaction numbers etc. Hope you don’t mind charting til midnight.
I’d love to think the ACO administrators would take a pay cut so their physicians could all have 60 minute new patient visits for Medicare (for example). I’d love to think that we, the exhausted PCPs, would all walk out one day and demand better working conditions. But we can’t and we don’t. So, value based care increases the burden of documentation on PCPs, increases the non-medical cognitive load on PCPs who are trying to follow all the ‘rules’, increases the financial risk to PCPs, and just generally makes practice harder. Perhaps other people’s practices have figured this out. Mine had not, at least not prior to my departure.
Author
KV,
Thanks for the insight of what the unintended consequences have been for PCPs. It sounds like an idea that sounded reasonable in theory fell completely short in practice.
We’ve certainly experienced the increase burden with documentation in the ER, as well as the added cognitive load of trying check the right boxes to avoid getting dinged (for us it involves things like explaining fallouts that deviate from sepsis guidelines or fail to meet time-sensitive door to groin times for ST elevation MI’s or thrombolysis for acute stroke).
Try the “If you were emperor” question: What, with your understanding of the front-line experience, would make it work better? Is it an unsalveageable idea, or is it a decent concept with poor execution?
Genuinely interested in your answer, if you have the time and inclination.
Thanks for letting me see a little more clearly through your eyes,
CD
Better late than never?
A few thoughts, all on time.
The first is the physician’s time. If PCPs could bill value-based systems by the minute (like my lawyer relatives do), undoubtedly those systems would invest in smarter EMR development that made the administrative work more efficient. Gmail can practically write my emails for me, why can’t the EMR offer a code for diabetes with CKD for me instead of making me search?
The second thought is the patient’s time. Patients highly value quality time, undistracted, focused time, with their doctor. A great deal of this Dr-patient time (like early childhood education!) looks basic on the surface, like nothing exciting is happening, but put in the time and you will get great dividends in the future. Relationship building takes time. Trust takes time. Value-based systems should be increasing, not decreasing, contact time between doctors and patients.
My final thought is our time together, as a profession. I have given and received the gift of time from colleagues to come and take a look at a rash, consider a strange lab value, discuss a difficult patient etc. But this is reserved for rare cases/emergencies only in a time-based billing system. I truly believe we’d all provide better care if we continued routine case-sharing, brainstorming, collaboration, teaching etc. We’d also be less burnt out. Outside of the academy, I haven’t seen this implemented routinely.
Author
KV,
I genuinely appreciate your taking the time to share those thoughts. Reducing the friction caused by the EMR to make it intuitive and simple, and giving you back control of your time (with patients and colleagues alike) sound like reasonable goals deserving of support.
Thanks,
CD
Beautiful comment as “value” based care does not consider the primary care physician sho has to spend the time implementing such a system or the patient who healthcare is affected by the system to be of any value. The value is translated as greater profit to an insurance entity or greater efficiency in a government agency. Meanwhile the cost of healthcare keeps rising along with the profits of the insurance industry while the relative value of a primary care physician’s salary drops along with the lifespans of the patients.