As a healthcare organization moves from fee-for-service reimbursement to population health–based care, it must accurately define how sick its population is—not only to take good care of these individuals but also to be reimbursed correctly. If clinicians undervalue the population through the clinical documentation, then the government and payers will follow suit, and that can cost a hospital or health system millions.
Borgess Health, a health system based in Kalamazoo, Mich., was able to uncover more than $6 million in reimbursement by getting physicians to improve their documentation. Chances are that for your organization, it's as simple—and complicated—as that.
Anthony Oliva, DO, CMO at Borgess Health, is no stranger to clinical documentation improvement. In 2004 he was vice president of medical affairs at Bayhealth Medical Center in Dover, Del., where the organization refocused its documentation process by taking a clinical perspective rather than concentrating on primary coding.
Bayhealth added a clinical documentation management program from J.A. Thomas to help get to the heart of assessing and reporting severity of illness (SOI) and expected versus observed mortality rates, to more accurately determine hospital and physician performance. So when Oliva arrived at Borgess Health, it was only natural that he looked at clinical documentation.
"Borgess Health was working to decrease observed mortality through quality care initiatives, but had overlooked another critical point: improving documentation to improve the SOI expected mortality part of the ratio," says Oliva, who adds that the healthcare industry has traditionally viewed mortality as a quality problem, largely ignoring the role documentation plays in outcome metrics.
Having gone through the process of assessing and correcting clinical documentation at Bayhealth, Oliva was intrigued at how changing Borgess Health's documentation approach might influence the organization's metrics. Borgess Health, part of the Ascension Health network, includes more than 120 care sites in 15 southern Michigan cities, as well as five owned or affiliated hospitals, a nursing home, ambulatory care facilities, home health care, physician practices, a cancer center, and an air ambulance service.
Implementing a new clinical documentation improvement program had to be done carefully to ensure that changes didn't negatively affect the system's overall quality of care. Once the J.A. Thomas clinical documentation program was in place, Oliva carefully monitored SOI and mortality for changes.
"If you're moving the severity number and increasing the expected level of severity of patients and your observation indicators, such as mortality, don't move, then you're not having an impact. But if it does move, then you know you're impacting your denominator and not doing anything to impact your quality," he explains. "The first measurements I took after J.A. Thomas was added, there was dramatic move in CC [complications or comorbidities] pairs, and our mortality dropped 30%. That told me the program was working."
To determine how large the opportunity was in the organization’s clinical documentation, Oliva needed to do a measurement gap analysis by comparing Borgess Health’s MS-DRG codes against a benchmark.
"I looked at a couple of different variables and measured the MS-DRG couplet and triplet percentage performance at Borgess Health based on high-to-low severity within 40 MS-DRG groups," he explains. He compared those numbers against his experience and data he'd seen while working at two previous organizations of similar size and make-up.
"That comparison showed me that there was a potential reimbursement increase of about $6 million if we took a more clinically focused documentation approach," he says.
But doing so required Oliva to get physicians to see where they were missing documentation and to work with documentation specialists and coders on accurately coding clinical work. No easy task, since it meant that physicians needed to spend extra time to do this, but it was the physicians’ outcomes that were the key to moving this program forward.
"With the data I'd gathered, I could show the physicians how putting their clinical documentation into a language that coders can understand would affect their outcomes," Oliva says. "With CMS and Healthgrades now tracking physicians' individual performance data, including individual physician mortality rates, seeing this information really hit a nerve."
Once Borgess Health physicians recognized they needed to document that their patients were sicker in order for severity adjustments to apply, it sunk in, Oliva says: "If you don't get the information into the documentation, it doesn't accurately reflect the patient that was treated. And if that physician is compared to another physician that is [documenting accurately], then that doctor's outcomes will look worse. Eventually we'll be paid for outcomes. So I tell them, 'Get the documentation right and you’ll get credit for what you're doing to care for your patients.'"
Physicians were not asked to memorize any codes, just to work more closely with the clinical documentation specialists when more information on the clinical notations was requested.
"We asked them to be part of the team that’s there to help them get their outcomes to where they should be. Over time, the physicians will learn where they need to put in more information to do a better job documenting," explains Oliva. "This is helping encourage a conversation between the coder and the physician."
Coders and clinical documentation specialists were also put through two weeks of clinical documentation training to grow their knowledge and improve the relationship between these two departments. "You have to build a team between the CD specialists and coders and the physicians so they don’t see their worlds as separate," says Oliva. "This is all really basic stuff, but it's effective."
The financial and clinical results speak for themselves: Since the implementation of the program in August 2011, Borgess Health has picked up over $6 million in reimbursements, and with a 25% improvement in severity-adjusted mortality, which has placed the organization in the top 10% performance category of the Premier, Inc., outcome database.
"In healthcare, it's a positive to identify patients on a more granular level," Oliva says. "We need accurate information if we are to manage populations in the future."
Karen Minich-Pourshadi is a Senior Editor with HealthLeaders Media.
