CMS and MedPAC Guidelines and Trends

Who’s Afraid of a Little Risk?

July 3, 2018 11:37 am

Is anyone really surprised by the recent predictions that organizations will be stepping away from the accountable care organization (ACO) model in the next year?

I’ve been thinking “I told you so” since mid-May when I read about what appears to be a coming Medicare ACO exodus. The health economist in me says, “What took you so long?”

That said, the organizations that took on the challenge of the Track 2 or Track 3 ACO models are to be applauded. Many have success stories to share, which can provide valuable lessons for their peers. 

The ACO model as written has been the focal point of one of my biggest criticisms of the Affordable Care Act since it was signed into law in 2010. The idea of retrospective attribution of patients to an ACO seemed flawed from the very start. It simply doesn’t make sense to take risk on patients you do not know belong to you until you already have been handed the consequences of their running up costs and ignoring preventive health guidelines. 

I understand why the Centers for Medicare & Medicaid Services (CMS) advocated for the retroactive attribution logic—to minimize incentives for “cherry picking” and “lemon dropping.” That is a worthy goal and one that makes sense. In a downside risk model, where the ACO has risk of losses if costs go above a target level, the incentive is very strong to find ways to encourage those patients with severe, high-cost, or chronic illness needs to seek care elsewhere  (hence the term lemon dropping). Conversely, there are strong incentives for organizations to encourage patients with fewer chronic conditions or needs to seek care in their facilities—cherry picking—when the ACO is taking the upside risk and will receive additional gain-sharing income if costs of patient care come in below target. As long as we have bottom lines to protect, those behaviors—although perhaps ethically challenged—are rational management tactics. However, it might be better to tweak the model on both sides of the transaction.

The population health model promoted by ACOs is turning around the “volume-versus-value” model ever so slowly, and it still has a long way to go. The objective of promoting health rather than promoting additional revenue through additional curative services seems to be getting more attention as the media continue to grouse about high healthcare costs without delving into the underlying causes. That paradigm change involves risk to our bottom lines, but taking some risk is not a bad thing, if done right.

Let’s start with the source of risk—the patient. How about getting a defined population to deal with, rather than the veritable “in and out” attribution model that now predominates in many markets around the country? Defined populations work for the insurer, so why not for the provider organization? Yes, there will be lemons and cherries in the basket that an insurer hands off, but statistically, it should balance out if the basket is large enough. The rule of thumb I used when working with risk contracts back in the 1990s was a minimum of 5,000 people, but 10,000 is better. Just as larger numbers are better for insurers, spreading the risk across more people offers providers some balance against being adversely selected and getting more than their fair share of high-risk patients. 

Now we just need CMS to work with the industry to make the patient attribution predictable. We understand the incentives associated with risk, and CMS is right to try and control those incentives. Why not try offering patients incentives to be part of a specific ACO? Why not try offering a benefit upgrade to beneficiaries to get them to join an ACO plan? Providers should be looking for ways to get healthy seniors to join, and they should consider marketing to patients with something other than nice facilities and short emergency department waits. These strategies have worked for Medicare Advantage plans for decades, and they worked for the Medicare Risk plans I was a part of back in the early 1990s. Such an approach would not constitute a big change, and it could be simpler than getting a state certificate of authority to be a health plan. Many have managed capitated risk for decades with success. Perhaps it is time to dust off those plans and try it again.

The issue of paying providers in the ACO network also is important to address. We all know the incentive of fee-for-service payment: Do more, get paid more. That’s been the provider’s mantra for decades. But partnerships in the days of bundled payments have shown that providers can band together and find ways to get a reasonable level of payment in a prospective fee and split that bundle so everyone is successful. It may be time to restart that conversation.

Maybe capitation is not such a bad thing, as long as the numbers are large enough and providers are committed to cost control. We now have more patient data at our fingertips than were ever available during my days in Medicare Risk. Surely, we could make population health with a prospective payment work again. It is now much easier to identify who our cost risks are, to determine what their needs are, and to get them the resources that mitigate those cost-of-care risks. There are plenty of reinsurance carriers out there that would help manage downside risk if only that risk were predictable. Predictability is the problem now. 

The ACO model of today is having its challenges, but some organizations are making it work.  Why can’t we make it work for others? It’s not all that far-fetched of an idea. This concept can be fixed so that all can prosper. But that will require that CMS work things out with the industry before much of its ACO experiment ends.


googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text1' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text2' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text3' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text4' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text5' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text6' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text7' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-leaderboard' ); } );