Certification Puts UTHealth Researchers on Path to Studying Texas Medicare Claims Data
Quality Feature — November 2017
Tex Med. 2017;113(11):47-53.
By Joey Berlin
It's more than just a bunch of data ― it's an opportunity, and one that a University of Texas health data center had to go for. Even better, it's given physicians a seat at the table.
The Center for Health Care Research Data at The University of Texas Health Science Center at Houston's (UTHealth's) School of Public Health is on the path to receiving Medicare claims information that could shed precious new light on cost, practitioners, and utilization across Texas ― information such as where the illnesses are, where the needs are, and where the high-performing doctors are. In a state where some people go to their local emergency department more days than they don't, Medicare claims data can be an illuminating treasure trove.
Last June, the Center for Health Care Research Data announced it had earned "qualified entity" certification from the Centers for Medicare & Medicaid Services (CMS). That certification means it will have access to claims data from Medicare Parts A, B, and D. The center will be the first CMS-certified organization to focus on that data specifically for Texas.
Cecilia Ganduglia-Cazaban, MD, codirector of the center, says the information will offer a more comprehensive look at patient activity than is available from one health system, such as UTHealth's.
"Having access to Medicare claims data allows us to track patients not just [by analyzing] one visit, but you can track them across time and across providers," Dr. Ganduglia-Cazaban said. "Even if they went into one hospital and then got into a different ER from another system, and then into another hospital, you can track these patients and see them as they move across providers and across time."
How the Data Can Help
Texas' car-clogged cities, sparsely populated counties, and locales in between all have their own stories to tell when it comes to health care quality. Drilling into the Medicare data can help reveal objective truths about those stories and perhaps drive health care policy change.
CMS-certified qualified entities are required to use the Medicare data they receive to produce reports on practitioner performance. CMS certified 18 qualified entities from 2012 to Nov. 3, 2016, according to its website. While some of the qualified entities focus on data from the entire nation, others focus specifically on one state, or even on a small number of counties in one or a few states.
"If you look at the map of the qualified entities across the country, there's a big gap in the South, and certainly in Texas," said Trudy Krause, PhD, who codirects the center with Dr. Ganduglia-Cazaban. "So we wanted to be the ones that would pick up that role and be able to do that process."
Dr. Ganduglia-Cazaban says the metrics of a given area with a high diabetes prevalence, for example, can reveal whether that population is getting the right diabetic preventive care.
"You can measure overall in Texas, but what we can really do with this data is identify areas in which either education programs or preventive programs could help solve situations and improve the care for that population," she said. "The same thing would go for other diagnoses, like cancer diagnoses: Do we have areas in which we know that we have a higher prevalence of a specific type of cancer, or cancer overall, and what care are these patients receiving there? Can we do something to improve their care in that area?"
Bradly Bundrant, MD, an emergency physician in rural Ballinger, near San Angelo, says the Medicare claims information should be able to reveal facts about patient acuity, as well as costs related to emergency department patients.
"We see a lot of people that come into the ER with high acuity, and they should be able to monitor what happens to them if they get admitted to the local hospital versus if they get transferred, and how many people get admitted to the local hospital and then get transferred. Those are all gauges of quality," said Dr. Bundrant, a member of the Texas Medical Association's Ad Hoc Committee on Health Information Technology.
He says the data also should be valuable for comparing rural hospitals like his ― one of two in Runnels County. Both are critical-access facilities. They're about 15 miles apart, but the closest hospitals otherwise are 35 to 45 miles away, he says.
"I suspect that there is a strong desire to reduce the number of critical-access hospitals, and the first criteria probably will be distance," he said. "If you have two critical-access hospitals that are close to each other, then one of them probably is going to be closed. That's just the bureaucratic pressure. And one of the best tools to decide which one should be closed should be information like this."
The Medicare numbers should tell some revealing stories for counties with much larger populations, too. Houston internist Lisa Ehrlich, MD, president of the Harris County Medical Society, says the county has "superutilizer" patients who in some cases visit the emergency department as many as 250 times during a one-year period.
"It's absurd," said Dr. Ehrlich, a member of TMA's Council on Health Care Quality. "But if you can identify those, and you figure systems out that can mitigate that, whether it's giving them a place to stay and having telemedicine … or social worker outreach to make sure they're taking their medicine, it's totally worth it ― because 250 visits to the ER is superexpensive."
A 2013 National Institutes of Health study of more than 8,000 outpatient emergency department visits during 2006–08 found the average charge per visit was $2,168, with a median charge of $1,233.
A regional map of the annual cost of care for commercial insurance in 2011, provided by the center, shows an example of what UTHealth researchers can do once they get the Medicare data. (See "Using the Data.") The center previously had obtained 2011 Medicare data for another project, and the variations in that data provided a glimpse into Texas' health disparities.
"It changes by whether you're in a metropolitan area or you're in a rural area or you're in the Rio Grande Valley, and some of those things just logically should not change for those kinds of reasons," Dr. Krause said. "What it really brings home is that there may be some very important social determinants of health that have to be considered. Especially when we're looking at moving more toward the value-based health care system; you have to look at all of the issues that impact health."
Although the center already has been certified, in September it still was in the process of completing two more phases of application to demonstrate to CMS that it has the proper security requirements in place to house the claims data. It was hoping to complete that process by late October and hear from CMS again by December. The center is seeking data from 2012 onward.
What Another State Learned
As it begins its work, the Center for Health Care Research Data can look for guidance from other qualified entities that have benefitted from the certification for several years. The Oregon Health Care Quality Corp., an independent nonprofit also known as Q Corp, has been certified since 2012.
Q Corp integrated the data it received from Medicare into both its public reporting system for consumers and its private reports for practitioners. Q Corp provides its reporting to most of the primary care clinics in the state, which allows those clinics to drill down into the quality measures.
"They can look and get patient-level detail and provide us feedback on whether patients have had particular services. So we have that information, and then that feeds into our Compare Your Care website, which is a website that is basically designed for your average consumer," said Cindi McElhaney, senior health care analyst for Q Corp. "The intent is to allow consumers to get in there and look by clinic at quality scores and, hopefully, make more informed decisions on where they seek their care."
The infusion of new data sheds light on how Oregon's Medicare fee-for-service population performed differently than either its Medicare Advantage or commercial populations. Ms. McElhaney says Medicare Advantage plans tend to perform well on quality measures such as A1c testing for diabetes, often performing better than commercial plans. She says the data reflecting Oregon's Medicare fee-for-service population came largely from people in rural areas or those who are dual-eligible on Medicare and Medicaid.
When Q Corp included the Medicare fee-for-service data, it saw a decline in several quality metrics. Drilling down into the measures revealed Medicare fee-for-service patients "tend to perform better than Medicaid, but worse than commercial populations," Ms. McElhaney said. For example, Q Corp's overall breast cancer screening rate, which it had been reporting for years, took a dip thanks to the Medicare fee-for-service patients.
"We always suspected this population was unique. In terms of quality-measure reporting, they do perform very differently than either the Medicare Advantage or the commercial population or Medicaid. They're kind of their own unique little subpopulation," Ms. McElhaney said. "From my perspective … that is one of the reasons we need to continue to include them in our data warehouse to have some transparency into both the cost and quality that that population receives and generates."
Q Corp Senior Director Meredith Roberts Tomasi says if Medicare fee-for-service data revealed that "Texas was really low in one area of utilization or really high in one area of price, there might be some probing questions about why that might be or what a community can do about it."
Hearing Doctors' Voices
In preparing for qualified entity certification, the Center for Health Care Research Data did something academic researchers often neglect to do: It asked Texas physicians for their input on how to best use the Medicare data.
The center's preparation brought physicians and other interested entities from across the state, including TMA, together in "think tank" sessions. The center presented a variety of approaches before a consensus developed to focus on variances in cost, utilization, and quality measures.
"They have enlisted physicians' input as to what sort of things would be valuable to us, which is somewhat unusual for an academic endeavor," said Dr. Ehrlich, who was part of those think tank sessions. "It was really collaborative, and I think that's why it's going to be more beneficial and give us more information that we can actually utilize, because we had physician input."
The center will study variances not only by geographical region, but also among different payer systems, such as Medicare, Medicaid, and commercial insurance. Although CMS allows qualified entities to sell Medicare claims data to practitioners, health plans, and others, Dr. Ganduglia-Cazaban says the center won't "go into the business of selling the data." But it will use the data for other research projects and will partner with other organizations.
Although measuring practitioner performance is a stated goal of the CME Qualified Entity Program, the center's work won't be "calling out" specific physicians for high or low performance.
"We're not reporting on any particular providers at this point in time, which a lot of qualified entities do; they actually name the providers. We're not doing that," Dr. Krause said. "We're looking more at variation across the state, and where [there are] areas that potentially need improvement, so we can really focus policy on those areas."
Dr. Ehrlich cautions that the Medicare data has limitations.
"We don't want to overstate it, because there may be a very small county that only has 500 patients in it, and [the quality measures can look] horrible because of one outlying utilizer. But because the mean is more influenced by more populated areas, the way that the statistics are done, they may be at an advantage or a disadvantage," she said. "For instance, all they have to have is one really sick patient who overutilizes in a small group to make that county look like an outlier."
Still, she calls the project "exciting and new."
"They can go as granular as by ZIP code or as large as by region, and it is really exciting information to see," she said. "They can definitely show disparities."
Joey Berlin can be reached by phone at (800) 880-1300, ext. 1393, or (512) 370-1393; by fax at (512) 370-1629; or by email.
November 2017 Texas Medicine Contents
Texas Medicine Back Issues