May 21, 2012

Patient Engagement, Interoperability Vital to ACO Success
By Greg Goth
For The Record
Vol. 24 No. 10 P. 14

Empowering consumers and flexible infrastructures work hand in hand to create an environment more apt to meet guidelines.

The final rule for the Centers for Medicare & Medicaid Services’ accountable care organization (ACO) definition runs 189 pages in the Federal Register, but any healthcare industry veteran knows the concept is quite simple and has been a goal of sorts for at least 20 years.

The question now is whether there has there been enough change in the financial, clinical, and technical landscape to make ACOs succeed where other approaches—such as network gatekeeper providers and capitation—to cost-effective “accountable” care have had either mixed results or failed miserably.

Cynthia Burghard, International Data Corporation research director, believes the time is right for ACOs, thanks to the availability of rich technical resources and a pressing need to avoid the critical financial ramifications should another attempt at collaborative and accountable care fail.

“The rate of medical inflation is totally unsustainable,” Burghard says, “so there’s more of a financial reality to having to make it work, and the other driver is the technology. In the early 1990s, doctors got no information about their patients, no real metrics by which to measure them, and at the end of a capitation program, their contracting health plan did some mumbo jumbo and said, ‘You owe us X thousands of dollars.’ There wasn’t any exchange of information and no process changes to meet the goals of capitation. It sounded like a good idea, but none of the processes that would allow it to be successful existed.”

“The previous efforts were completely financially based,” notes Bill Fera, MD, executive director of Ernst & Young’s Health Advisory Services. “And frankly, I think it had to be because all you could measure was claims and finances. The thing that has to be different if this is going to work is the availability of clinical data and being able to track that in real time to make sure evidence-based practices are really being followed.”

As for what sort of technical components may comprise an ACO, Burghard says there will be no typical configuration. “There isn’t going to be a single ACO platform,” she says, “and the way the market is responding to date, there are offerings from almost every segment, whether it’s the big IBM or Oracle-type vendors, or the big healthcare enterprise vendors or analytics vendors, or health plans offering their services. There are a ton of different models being experimented with at this point. It depends on your relationship. For instance, a dominant EHR vendor may think about adding care management modules or think about how to bring payer data into analytics modules.”

The drive toward ACOs in both public and private delivery/payment models dovetails quite well with two other overarching healthcare mandates: meeting federal meaningful use requirements and implementing ICD-10. As organizations move to satisfy meaningful use requirements, they will, by definition, be strengthening interorganizational communication capabilities. As ICD-10 eventually becomes the coding and claims standard, both payers and providers are discovering that mapping ICD-9 to ICD-10 and testing the new codes vs. the old ones to achieve revenue neutrality is best achieved in a climate of collaboration rather than isolation.

“Meaningful use is a good opportunity to stair-step in with structured data,” says Joy Grosser, vice president and chief information officer of the Iowa Health System in Des Moines. “In the long run, our caregivers, especially care coaches or point-of-care physicians, will have access to all that information because it is structured in a way that the system can alert them to changes in the health of their population.”

Enlarging the Feedback Loop
Fast forward 20 years from those early days of reform and so much has changed. The federal government has provided the kick-start the market needed to embrace EHRs; the impending adoption of ICD-10 will supply much more granular clinical documentation; and the effort behind the National Health Information Network and regional health information exchanges (HIEs) appears to be gathering momentum.

However, two vital unknowns in those elements have the potential to either confirm the best possible scenario for ACOs or drive the concept aground: How will HIEs find a way to justify an investment in their future? What role will patients play in the structure of ACOs?

“For an ACO to be effective, you have to have a population of patients who are empowered,” says Scott Van Valkenburg, MD, senior manager of Ernst & Young’s Health Advisory Services, “and one of the best ways to do that is to provide them the technology to increase their control of their health, their knowledge of their health, and communicate with their providers across the whole continuum of care.”

One example of such technology, Van Valkenburg says, may be a Web portal with multiple functions, such as biometric monitoring interfaces and communication capabilities.

“We need to do a better job of making it easy for patients to keep their information up-to-date,” Fera says. “I think people want to get to an ATM type of mentality and if we could, we could solve a lot of the interoperability and collaborative efforts that need to occur. A functional, easy-to-use personal health record in so many ways would solve a lot of the interoperability issues—where the patient really becomes the vector for interoperability, which I think would be ideal. But we have to do a better job of auto-updating those types of applications with the appropriate information.”

There are several issues that need to be addressed before this strategy can come to fruition. Both Fera and John Cuddeback, MD, PhD, director of medical informatics for the American Medical Group Association’s Anceta initiative for data-driven shared learning, point out that the industry is still struggling to find a sustainable business case for HIEs, and the drive toward third-party PHRs, such as Google Health and Microsoft HealthVault, has received a mixed reception.

Cuddeback says those concerns are known throughout the industry. In fact, Farzad Mostashari, MD, national coordinator for HIT and a featured speaker at the American Medical Group Association’s recent annual conference, addressed them directly. “He said, if you consider the cost of failure in terms of managing patients with chronic disease and managing referrals effectively, there is enough money in the system to pay for the infrastructure it will take to create the interoperability,” Cuddeback says. “I think the challenge is developing an infrastructure that will then have an ongoing business model. That remains a challenge because every market is different.”

Covered Population vs. the Individual
Complicating matters are the potential incompatibilities that may exist between the goals of combining care coordination across covered populations and individualized courses of treatment with all the subtleties that reside in a clinician’s notes. To bridge those concerns, providers and IT vendors will have to start thinking creatively.

“Meaningful use is pushing for more and more data to be captured in structured form instead of the subtle and meaningful variation we used to have in clinical narrative,” Cuddeback says. “You’ll occasionally see physicians lament the passing of the benefits of the clinical narrative, with nuance in the way you phrased things, but that doesn’t fit into a decision-making algorithm that will allow you to remind the physician or caregiver what the patient needs and make sure those steps have indeed been taken.”

Chris Weiss, president of Dynamic Clinical Systems, manufacturer of an integrated patient-reported outcomes platform, says the company’s developers have been able to program the platform so that 80% of questions that otherwise may generate free text answers can be translated into some sort of structured input. In addition to the fact that structured data are more easily integrated from system to system, he says the form factors into which data are entered also play a role. For example, when the digital content services platform was first introduced, patients used tablet devices “and we didn’t want them to have to type responses,” Weiss says.

Facilitating practice-to-practice data transfer will be crucial to meeting the expectations of the ACO concept. However, the scope of traversing disparate practices is daunting. Cuddeback refers to a 2009 study published in the Annals of Internal Medicine in which the authors found that the typical primary care physician needed to coordinate care with 229 other physicians in 117 other practices. For the cohort of patients with four or more chronic conditions, care needed to be coordinated between 86 physicians in 36 practices.

“In a typical very fragmented healthcare system, this is not just going to happen by magic,” Cuddeback says. “And connecting the EHRs is a necessary but not sufficient solution to that problem.”

Grosser says other sources of healthcare-related data, such as pharmacy purchases and dental care records, can also be crucial components of a patient’s profile. Direct communication with patients, whether it’s through a digital monitoring application that uploads data or via surveys and messages, must be a priority, she notes.

Platforms considered well outside the formal definition of clinical data, such as exercise applications, can factor into a patient’s wellness, which, in essence, ACOs are designed to encourage, Grosser says. For example, in early April, Microsoft, the City of Los Angeles Department of Aging, application consultant Get Real Consulting, and St Barnabas Senior Services, among others, announced a partnership that encourages seniors to get active by using Microsoft’s Kinect gaming technology and logging vitals such as blood pressure and glucose levels into HealthVault.

Grosser sees great potential for applications developers that help bring health information into the broader clinical record. “While our HIT vendors are great about the opportunity to care for patients while they’re in front of us, they’re still a little bit limited by the fact that they only see the patients that exist in their system,” she says. “If I have people in my population I haven’t seen yet or who have only gone to the pharmacy or have an exercise app and haven’t done anything else, my HIT vendor doesn’t have them under the covers. So part of the equation becomes where are the population health vendors who help me take the population of care and help me sort out the highest risks and start there?”

How best to reach that high-risk group may take some paradoxical reasoning, she says. “The 30-year-old patient who has never sought healthcare because they’ve never needed it isn’t the place to spend 80% of my energy,” Grosser says, “but at the same time, my 30-year-old patient who has never sought healthcare because they haven’t needed it is probably the easiest one to start engaging in activities on the wellness side.”

Moreover, that healthy young patient is also very likely the conduit to their higher-risk parent. For example, many children of elderly patients will help send in the results of self-administered tests such as blood pressure readings and glucose monitoring, Grosser says, noting that they’ll also lend a hand with answering patient survey questions and scheduling online appointments. “There may be education that comes to the patient from their caregiver that would be good for both patient and child to go through to understand,” she adds.

With the panoply of new data sources, providers should expect to meet patients with very different ideas of what kind of care they need. For example, a patient who exercises regularly and has high levels of both HDL and LDL cholesterol may balk at taking a statin that may be de rigueur for more sedentary patients.

Cuddeback says there are ways to address these situations, citing Riverside Health System in Newport News, Virginia, as an example.

“They no longer measure just against standardized thresholds, as for hemoglobin A1c in patients with diabetes, “ he notes. “They are also doing their quality measures against targets that are jointly set by the physician and the patient.”

To create these tailored goals, Riverside’s EHR needed customization, but “it makes such a difference in the willingness of the patient to do what’s required because they were part of setting that goal, it wasn’t imposed on them,” Cuddeback says. “That’s a perfect example of the sort of thing you hope is going to grow out of providers becoming more accountable for the overall health of the patient population: That there will be a population view but also the individualization that is important to ensuring good outcomes.”

— Greg Goth is a freelance journalist from Oakville, Connecticut, specializing in technology and healthcare policy issues.