Last November, chief analytics officer Charlton Park got his five seconds of fame in an NBC Nightly News segment on how the University of Utah Health Care decreased costs by 30 percent during a period when other teaching hospitals saw their costs rise.

Few healthcare IT leaders end up on national TV news. But then few have helped solve two of healthcare’s biggest mysteries: How much does it actually cost to deliver patient care? And how do the costs relate to the quality of care delivered?

Hospitals have historically allocated costs in a siloed fashion by department or service—for example, by nursing unit or operating room—making it difficult to determine how much is spent on patients across their hospital stays. As hospitals assume more financial risk under bundled payment and other value-based contracts, they will need a more realistic understanding of their costs so they can improve inefficiencies that could contribute to financial losses and negative patient outcomes.

University Hospital of University of Utah Health Care
University Hospital of University of Utah Health Care

In 2012, the University of Utah Health Care turned the cost question into an organizational imperative. “Our senior vice president, Dr. Vivian Lee, along with her executive team said, ‘We need to get our arms around this,’ ” Park says.

The result: a cost accounting system that can determine precisely how much it costs the health system to provide care, for example, in the emergency department (82 cents per minute) or in the ICU ($1.43 a minute). “Cost accounting allows us to get a single cost number for all the care involved in a patient encounter and compare it to the payment we receive,” Park explains. “If the margin is tight, where are the areas where we’re spending the most?”

Physicians can also benchmark their costs, which are adjusted for severity of illness and other clinical issues. “We can compare two providers for the same procedure and see who is more expensive,” Park says. “We're then able to drill into those costs and understand what's driving the variation? Is it the utilization of drugs or is one provider keeping patients in the hospital longer?”

To help physicians determine how to reduce costs without harming patients, quality data is provided alongside cost data. IT staff is developing scorecards for 50 high-cost or high-volume conditions, including sepsis and C-sections. “The scorecards combine the cost trend over time with anywhere between six to a dozen quality metrics,” Park says, adding that his team meets with providers and the quality department to see which outcomes are most imperative to track.

The healthcare system decided to build its own cost accounting system rather than go with a vendor application. “We used to have a vendor-based system, but it didn't give us the detail we needed,” Park says. “Ours is home grown and very data driven, so we have complete flexibility. We wanted the most detailed data collected at the patient level through the electronic medical record.”

A recent electronic health record (EHR) switch to Epic has helped the health system pinpoint costs down to the minute. “It [the EHR] takes a lot of time stamps when a patient enters or exits a room and when a nurse is doing this, that and the other. That data allows us to allocate cost based on how many minutes a patient is in the operating room, or on a certain nursing unit,” Park says.

University of Utah Health Care started building its cost accounting solution with a major advantage: “Our data warehouse is almost 30 years old, so we've got years and years of clinical, financial, operational, and all kinds of data,” he says.

Turning all that data into easy-to-understand scorecards comparing costs and quality involves numerous moving IT parts, Park explains. First extraction, transformation and load (ETL) processors pull the data out of the EHR and other source systems at regular intervals and feed the data into the warehouse. These data are then directed into dozens of data marts (e.g., supply, labor, financial/general ledger, clinical) that were built by the organization’s data warehouse team.

“We're able to relate all those different data marts to each other and apply different rules or methodologies for allocating costs,” he says. Then a SAP business intelligence tool analyzes and present the data. “That’s our front-end analytic tool that sits on top of this data foundation we created, and we use it to make all the pretty charts and graphs and drillable reports.”

The most time-consuming part of the build was determining the cost allocation rules. For six months in 2012, Park was part of a team of about 30 staff—including physicians, finance, and IT—sequestered off campus to determine these rules. “This core group was focused on building the infrastructure with all the methodologies,” he says. “We would request operational leaders from the operating room or the nursing unit or pharmacy to meet with us so we could question them about how their costs worked. Then we would design their methodologies to be accurate and appropriate.”

As Park explains, some costing rules are bottom-up methodologies and others are top-down. Which approach is used often depends on what cost-related information is available. For example, hospitals know how much they pay vendors for supplies. Thus, a bottom-up methodology might begin by identifying all the supplies used for a hernia patient, for example, and allocating the costs of those supplies to the patient’s cost of care.

In contrast, a top-down example might begin with the knowledge that the OR only assigns one OR technician to each surgical suite, which would help determine an average per-minute OR technician labor rate. That rate could then be multiplied by the minutes the hernia patient was in surgery to determine the OR tech labor expense to allocate to the patient’s cost of care.

“It’s a complicated process, Park says, adding that the health system’s cost accounting effort would not have been as successful without the hands-on attention of senior executives and a multidisciplinary approach. “To get this done right, it takes a lot of resources and attention,” Park says. “I think that's one of the downfalls of a lot of costing systems: It’s all owned and operated by the finance team. There needs to be clinical and operational input to make a difference.”

The University of Utah Health Care now runs its costs once a month after the accounting month closes. IT staff then conduct an audit before pushing out the information via various reports. “We make sure the costing is tying back to the general ledger,” Park says. “If it doesn't, or we're not in the allowable margin, we dig into the issue, identify what's going on and then re-cost it.”

When asked what IT leaders should consider when building or buying a cost accounting system, Park stresses the need to relate costs to quality and outcomes. “The costing data alone won't engage the physician.”

Because costs and quality information is transparent at the University of Utah Health Care, physicians are making decisions with costs in mind,” Park says. “More and more doctors are accessing and becoming familiar with cost data. It’s been very interesting to meet with them and see their reaction when they see the cost of an implant or a pharmaceutical used in a case. It can often be a big surprise for them.”

This transparency has led to a number of improvements that have significantly reduced costs, including a 30 percent reduction in the annual cost of total joint replacements.

Park offers one more piece of advice: Ensure the cost accounting system is flexible enough to allocate costs in various ways. “Our solution allows us to do it [allocate costs] in whatever way we feel is most accurate. Then we're able to quickly adapt those methodologies and make them more accurate if there are new data sources that become available.”

Looking to the future, the University of Utah Health Care’s tool may be available commercially. The health system is talking with some development partners about this possibility, Park says.

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access