Burda on Healthcare: I’d Like Some Interoperability with My Value-Based Care, Please

I’ll admit it. I’m not a good cook. Oh, I can do the basics. Grill burgers. Roast a chicken. Boil pasta. Order a pizza. But anything beyond that politely sits in the refrigerator for a few days before it disappears into the garbage can and takes a few disposable containers down with it.

I’m told by those I live with that I’m a bad cook because I don’t have a discerning palate. I’m reminded by those I live with that hot chocolate (chocolate syrup stirred into cold milk and heated up in the micro) isn’t really hot chocolate. Or that my famous “chicken Italiano” dish (chicken breasts with leftover Italian dressing poured over it and baked) isn’t really a thing. It’s just laziness.

I don’t know what goes with what, and I don’t know what doesn’t go with what. I smirk at wine pairings on a menu before I order my vodka martini on the rocks with three regular olives (none of that blue cheese-stuffed olive business, please) because, to me, vodka martinis go with everything.

Do Value-Based Care and Interoperability Go Together?

It should come as no surprise, then, that two things that I thought went together in healthcare — value-based care and interoperability — don’t really, at least not yet, according to a recently published study in JAMA Health Forum. You can download the study here.

You’d think that a hospital or medical practice’s ability to provide the best possible care at the lowest possible cost to a patient would depend on the provider getting a complete medical picture of a patient by accessing the patient’s health information from other sources, including other providers. If you do, you think wrong like I did.

In this study, which to me says a lot about the behaviors of hospitals and EHR vendors, researchers from the University of California in San Francisco and Vanderbilt University looked at the connection between hospitals’ participation in alternative payment models and hospitals’ interoperability capabilities.

APMs included accountable care organizations, bundled payment arrangements and patient-centered medical homes. Being fully interoperable means a hospital queries patient data electronically, sends patient data electronically, receives patient data electronically and integrates patient data electronically into its EHR system from providers outside of its own system.

The researchers’ study pool was 3,928 general acute-care hospitals, and their study period was 2014 through 2018. The researchers analyzed changes in hospitals’ APM participation and interoperability capabilities over that five-year timeframe to see if there was any connection between the two. Again, you’d think that the more interoperable you were, the more you would participate in an APM. Or the more you would participate in an APM, the more interoperable you’d need to be.

APM Participation and Interoperability Rose At the Same Rate

Crunching hospital survey data from the American Hospital Association, here’s what the researchers found:

• The percentage of hospitals participating in any of the three APMs rose to 44.8 percent in 2018 from 31.5 percent in 2014.
• The percentage of hospitals that were fully interoperable rose to 45.4 percent in 2018 from 22.6 percent in 2014.
• The percentage of APM hospitals that were fully interoperable rose to 55.4 percent in 2018 from 33.5 percent in 2014 — or an
  average of 5.5 percentage points per year.
• The percentage of non-APM hospitals that were fully interoperable rose to 37.2 percent in 2018 from 17.6 percent in 2014 — or any average of 4.9 percentage points per year

True, the APM hospitals were consistently more fully interoperable than non-APM hospitals over the study period. But the growth rate in interoperability of both sets of hospitals was essentially the same — 5.5 percent versus 4.9 percent per year. Using some
fancy math, the researchers determined that the difference was statistically insignificant.

And that’s what’s significant. Or, in the words of the researchers: “There was no observable evidence that hospital APM participation was associated with interoperability engagement.”

They offered a two possible reasons.

• One, the APM financial incentives — upside risk only — may not be strong enough to change hospitals’ behaviors substantially.
• Two, there are easier and cheaper ways for hospitals to cut costs and earn bonuses under an APM than hoping for a return on any investment interoperability.

To be fair, the percentage of hospitals that are fully interoperable did creep up to 55 percent in 2019, as we reported in this blog post, “Catching Up on Hospital Interoperability,” about a year ago.

Still, the lack of financial incentives to become fully interoperable could explain the slow progress. In their conclusion, the researchers bemoaned that, at this rate, 100 percent of hospitals won’t be fully interoperable until 2027, not for another five years.

Evidence Suggests EHR Vendors Share the Blame

There’s also another possible reason, and you can find that buried in the supplemental materials that accompanied the published study. (Isn’t the good stuff always in the supplemental materials?) The reason is EHR vendors themselves. We all know that EHR vendors, EPIC particularly, fought interoperability to protect their market shares. That’s why we needed the interoperability and information blocking rules, which took effect in April 2021. The rules implement provisions of the 21st Century Cures Act passed in 2016.

(You can learn more about that in this 2020 commentary, “Healthcare’s Epic Problem & the Audacity of Liberating Patient Data,” by 4sight Health’s David Johnson.)

In the supplemental materials, the researchers published the EHR vendors used by the hospitals in the study pool and the vendors’
share of the market. Here are the top five:

1. EPIC (used by 30 percent of the hospitals in the study pool)
2. Cerner (used by 22 percent of the hospitals in the study pool)
3. Meditech (used by 22 percent of the hospitals in the study pool)
4. CPSI/Evident (used by 8 percent of the hospitals in the study pool)
5. Allscripts/Eclipsys (used by 5 percent of the hospitals in the study pool)

I don’t believe in coincidences, either.

Burda’s Final Bit

I do think value-based care and interoperability do make a good pairing. But I don’t think it’s going to happen unless we make interoperability a condition in a VBC model or make VBC payments — rewards and penalties — dependent on interoperability.

Would you like some hot chocolate with your chicken Italiano?

Thanks for reading.

Listen to the weekly podcast, 4sight Roundup

Dave Burda hosts the best 20 minutes in medicine.
4sight Roundup drops every Friday. Listen here.