[Reblog] Value-Based Care’s Data Problem

From the 22 May 2015 post at The Healthcare Blog

…

I believe the concept of value-based care is good for healthcare. VBC encourages providers to make changes that put the patient at the center of care, so that different services can be provided across providers in a collaborative way. If all went according to the VBC vision, there would be fewer redundant tests, more emphasis on preventative care, and an effort to keep high-risk patients out of the emergency room. It’s also better for costs, something we desperately need in the US, where healthcare spending per capita is more than twice the OECD average.

But Lisa’s story, at the leading edge of the value-based experiment, is not good at all. ACOs and most other value-based models are new, constantly changing, and unproven. ACOs report on 33 metrics that are supposed to represent the quality of care provided by their networks of providers. While still extremely limited in scope, any more than 33 metrics would have made Lisa’s job impossible. So far, few ACOs have reported any savings. Worse — the metrics are unproven. What if they overemphasize standardized process over patient outcomes? And what if efficiency measures result in neglectful and impersonal care? A lot is riding on Lisa’s testing ground.

---

The administrative challenge

By engaging with and learning from people like Lisa, I have begun to understand the problems frontier administrators face — the same problems countless others will face if we don’t address the administrative burden early on. Here are a few of the top headaches being rolled out in the name of value:

Selecting metrics

For ACOs, 33 metrics are tracked today. Inevitably, these will expand and change as accountable care evolves. There are also countless other systems of metrics encouraged by other incentive programs: the Physician Quality Reporting System measures, Meaningful Use metrics, Agency for Healthcare Research and Quality Indicators, the Consumer Assessment of Healthcare Providers and Systems for patient experience metrics, indicators for each specialty (Stroke and Stroke Rehabilitation Physician Performance Measurement Set, Endoscopy and Polyp Surveillance Physician Performance Measurement Set, and the Heart Failure Performance Measurement Set, to name a few). The document outlining protocols for the Physician Quality Reporting System is 18 pages long, with a mouthful of a title to match: “The 2015 Physician Quality Reporting System (PQRS) Measure-Applicability Validation (MAV) Process for Claims-Based Reporting of Individual Measures.” Got that? A new piece of legislation that passed the House of Representatives last week — the “doc fix” bill — is about to revamp many of these requirements once again.

Collecting data

Lisa had to fumble through different electronic systems and paper charts to extract the relevant data for each patient in her panel at dozens of different clinics. In many cases, it was clear that care had been provided (e.g. an unstable patient had been upgraded from a cane to a walker), but the documentation wasn’t there (to fulfill the “Screening for Future Fall Risk” metric, documentation must state whether the patient had no falls, one fall without major injury, two or more falls, or any fall with major injury.) Therefore, even though care was provided to prevent future falls, the documentation did not meet the CMS requirement and no credit was given.

For the next reporting year, Lisa is designing her own reporting mechanisms for clinics and doctors. She says that her first reporting experience “was invaluable in learning ways to improve the reporting for year 2015 and beyond,” and she is putting processes in place to facilitate reporting next year. But each clinic is different: some need a page at the front of their paper chart with check boxes, and some have templates in their electronic health records. Her new processes may improve the situation, but additional tracking could also cut into time doctors spend with patients and add to the squeeze they already feel.

Integrating data

Lisa integrated all the data from each clinic manually, and this is a problem for small institutions who are trying to communicate and coordinate with each other. Right now it takes a long time and is not very scalable. Even at larger institutions with leading electronic health record systems, the data is locked away within proprietary databases, often in incompatible formats. Clinical data is rarely integrated with financial and patient-reported data in the way required to tie outcomes and claims to reimbursements in a value-based model.

Reporting

After all of her data collection, Lisa still had to submit her data to a third part to produce reports, and she will wait many months for the results. The CMS websites are comically complex ; the instruction manual for using the CMS metric reporting interface is 127 pages long.

---

Putting patients at the center

If these problems aren’t addressed, we’re in for a long and painful healthcare reform. Administrative costs will continue to rise, along with another generation of frustrated physicians and admins. Moreover, value-based care could be deemed a failure not because it’s a bad idea but because of poor implementation. Instead of putting patients at the center of care, it could breed more bureaucracy and force doctors to spend more time reporting on metrics and less time with patients.

We can address these issues and we must — to give value-based care a chance at moving the US toward more patient-centered, less exorbitant healthcare.