Understanding how to drive clinical and financial value is critical in this new healthcare environment. Dan Michelson, CEO of Strata Decision Technology, and Steve Allegretto, vice president of value innovation and strategic analytics at Yale-New Haven (Conn.) Health System, are using data analytics to develop this core competency at Yale-New Haven.
YALE NEW HAVEN HEALTH SAVES $150 MILLION
Learning to speak the same language on cost and quality metrics has helped clinical and finance leaders at Yale New Haven Health System reduce variation, improve patient outcomes, and decrease overall costs.
Over the past two years, leaders at Yale New Haven Health System, New Haven, Conn., have turned a corner in their efforts to engage physicians and nurses in reducing unnecessary clinical practice variation. Leaders were able to initiate meaningful and productive conversations with clinicians through a simple shift in focus—specifically, looking at clinical and cost variation through a quality lens first.
This focus on quality is at the center of Yale New Haven’s multiyear, $125 million cost-savings plan, designed to reduce the health system’s cost per case by about 20 percent. To get there, leaders at the three-hospital system are relying on a customized set of quality metrics, coupled with a sophisticated cost-accounting system and other tools, to partner with clinical leaders in reducing variation, improving outcomes, and decreasing costs.
Like Yale New Haven, other health systems around the country serving large Medicare and Medicaid populations are looking for better ways to collaborate with clinicians on quality improvement and cost reduction. Finance leaders in particular recognize the critical importance of fully engaging clinical leaders to help their organizations prepare for changes resulting from healthcare reform and the move toward value-based payment models. Yale New Haven’s experience points to strategies finance leaders can use to successfully partner with clinical leaders around cost and quality.
Identify Key Quality Targets
When hospital finance leaders share data on clinical variation, they often face significant skepticism from physicians about whether the data truly reflect their experiences with their patients. Physicians also tend to become impatient when finance leaders talk about cost per case instead of topics that reflect quality concerns, such as waste and the need to reduce clinical complications. Finance leaders are prone to be skeptical as well, often questioning whether quality improvements can lead to improved margins through reductions in labor and nonlabor costs. This skepticism on both sides can obstruct effective communication.
At Yale New Haven, finance leaders have met this challenge by developing a common language around clinical quality and variation that makes it easier for finance and clinical leaders to understand each other. The language is based on metrics they developed called quality variation indicators (QVIs).
Simply defined, QVIs are potentially preventable complications and adverse events that occur in patients. To identify the QVIs, a team of physicians and decision-support nurses spent three months reviewing a comprehensive list of hospital-acquired conditions, patient safety indicators, and other negative outcomes that were associated with significant variation in quality across the organization. Today, the system tracks nearly 30 QVI categories—such as central line-associated blood stream infections, deep vein thrombosis, and iatrogenic pneumothorax (a punctured lung)—for all payers. This approach gives finance and clinical leaders a shared basis for discussing how quality and waste affect both utilization of services and overall cost per case.
The QVIs help clinical and operations leaders of each cost center understand the indicators’ influence on various quality metrics. Each month, cost center and service line leaders receive reports in their monthly financial statements that detail the costs of cases that have a QVI and those that do not. On average, a case with a QVI will cost three to four times as much as a case without a QVI.
Before having the QVIs in place, leaders knew that cost varied within patient populations, but they could not attribute the variation to a particular quality issue. The QVIs allow finance leaders to have meaningful conversations with physicians and other clinical leaders about quality and its impact on cost. They also enable the finance team to measure how quality improvements affect financial outcomes, such as cost per case.
By sharing this information with clinicians, health system leaders hope to drive change that promotes improved quality while reducing costly clinical complications and waste. The leaders are careful not to assign responsibility for a QVI to a particular physician or provider, because a QVI may have occurred at any point during the process of caring for a patient.
The integrated QVI data allow the organization to translate how specific quality-improvement and waste-reduction efforts affect patient-specific costs. In the past, physicians and clinical leaders would spend much time arguing with the finance team about the validity of the data. Now, physicians and clinicians recognize that the data offer an accurate picture of the types of patients they see daily, making them much more receptive to viewing finance as a partner in their efforts to improve patient care.
Find Enthusiastic Clinical Redesign Champions
Reducing variation to improve quality and reduce costs often requires the redesign and standardization of clinical processes. For this reason, hospitals should appoint clinical leaders and teams who can review clinical outcomes and cost data and propose appropriate practice changes.
Leading the clinical redesign efforts at Yale New Haven is Thomas Balcezak, MD, senior vice president and chief quality officer. A physician with a deep interest in healthcare delivery and finance, Balcezak helps guide physician-led clinical redesign teams. These teams, which focus on specific areas such as major joint surgery and sickle cell anemia care, are charged with examining clinical practice, reviewing integrated data, and recommending strategies to reduce inappropriate variation.
A robust cost-accounting system facilitates improved communication between clinical and finance leaders by enabling them to see variation related to clinical processes, outcomes, and cost of care. “We were surprised by how often we saw those three areas overlap—areas where we knew we needed to improve processes of care as well as clinical and financial outcomes,” Balcezak says. “We decided to target those areas of overlap first so we could demonstrate to the institution that targeting costs did not mean there would be any degradation of quality of care. By focusing on those areas, we have been able to convince clinicians to continue to partner with us on our clinical redesign projects.”
When developing a clinical redesign project, leaders at Yale New Haven seek out an engaged clinician to lead the initiative. “When we find the right leader, we show that individual the QVIs and the detailed cost accounting that we have done, as well as the payment arrangements or the quality program that influences how the revenue flows for this patient population, Balcezak says “Then we search for opportunities where we can simultaneously improve the experience, outcomes, and finances for that group of patients.”
Systemwide, leaders at Yale New Haven are focusing their early efforts to reduce clinical variation on three main areas: blood utilization management, palliative care, and care in and outside of the intensive care unit (ICU). More than 30 physician- and nursing-led clinical redesign projects at the three hospitals are centered on improving care related to abdominal surgery, hip fractures, emergency department (ED) admissions, and more.
For example, one physician led the cardiac clinical redesign team’s efforts to improve the appropriate use of bivalirudin, an agent that can reduce the risks of bleeding after surgery in at-risk patients, according to published clinical trials, but that is often used in patients with a low risk of bleeding. At Yale New Haven, an analysis revealed that physicians who prescribed bivalirudin about 90 percent of the time or more for patients with a high risk of bleeding also ordered it 90 percent of the time or more for patients at a low risk of bleeding. Meanwhile, there were instances where physicians actually were prescribing bivalirudin to a greater percentage of patients at low risk of bleeding than to patients at high risk.
The physician leader partnered with the finance team to create monthly reports showing the variation, which has helped to encourage more appropriate prescribing of the drug. In addition, a portion of the clinical department’s performance arrangement is tied to appropriate bivalirudin use. As a result of these efforts, Yale New Haven increased overall appropriate use of bivalirudin by nearly 30 percent, contributing to a reduction in direct cost per case for these patients.
Look for Early Wins and Capitalize on Success
Deciding which areas of variation to focus on can be a challenge for many organizations. Ideally, hospitals should look for projects that can generate results early to help build momentum and clinician buy-in.
One example of an “early win” at Yale New Haven was an initiative for patients with sickle cell disease that improved patient outcomes and achieved a $1.9 million cost improvement during a two-year period. The sickle cell clinical redesign team standardized hydration and infection-control protocols in a dedicated, 20-bed unit for these patients. The team also improved care coordination by giving recently discharged patients around-the-clock phone access to a physician and two mid-level providers with specialized expertise in sickle cell conditions. As a result, the clinical redesign team was able to cut length of stay (LOS) for patients with sickle cell disease by 25 percent in the unit and backfill beds with patients with other conditions. The team also reduced ED utilization by 20 percent and increased primary care visits among patients with sickle cells by 20 percent.
As another example, one of the health system’s hospitals has had significant success in reducing ventilator-associated pneumonia (VAP) through a new protocol implemented in 2013. Specifically, clinical leaders changed the order set in the electronic health record (EHR) to include different sedative medications and early mobilization to get patients moving more quickly. As a result of these efforts, the number of VAP cases in the surgical ICU (SICU) dropped from 11 to four over the period of 2012 to 2013, saving approximately $200,000 in labor and $500,000 in nonlabor costs, including respiratory therapy costs. As VAP cases decreased, the SICU nurse manager appropriately managed staffing ratios to match patient demand, demonstrating that better quality can lead to lower costs.
Having “poster projects” like these has helped leaders at Yale New Haven build momentum for current projects, including an initiative focused on managing outcomes and costs among joint replacement patients. The health system has built a registry to track the functional status of joint replacement patients before and after surgery. By measuring these data and adding other patient outcome measures in the future, the system hopes to demonstrate to payers that it can improve quality and reduce costs in this high-profile population.
With respect to ROI, Yale New Haven’s leaders initially set a goal to achieve an 8-to-1 return on each clinical redesign project. For example, initiatives that require a $100,000 investment (in staff, technology, and the like) need to generate $800,000 in savings that can be attributed to improved performance in at least one of four areas: quality variation, process, revenue, and backfill. Of course, not every clinical project will meet its targeted ROI. But clinical and finance leaders can still learn from these initiatives and keep moving forward.
Help Leaders Become CEOs of Their Units
To drive cost savings at the cost center level, nursing leaders need to be part of the discussions on variation. The reason is simple: Although organizations need physician partnerships to make clinical process changes, nursing and other clinical leaders have the organizational responsibility to translate those utilization and practice changes into real cost improvements. For example, if LOS decreases due to a clinical redesign process, the only way for this change to translate into meaningfully reduced cost is if nurse managers realign labor and nonlabor resources on their unit.
Yale New Haven’s leaders learned this lesson the hard way when a clinical redesign project that reduced LOS in psychiatry failed to reduce year-one labor costs. A number of clinical leaders were not appropriately involved in the redesign efforts, and the hospital was not able to adjust staffing patterns to match the clinical change. This was a major learning experience for Yale New Haven’s leaders, who realized the importance of timely communication and the involvement of all departments contributing to the successful care of patients.
Today, Yale New Haven involves nurse managers and department leaders earlier in the redesign journey and provides them with the meaningful data they need to manage their units and patients more effectively. For example, the finance team has automated a process that runs each unit’s budgeted expenses through a costing system to create a budgeted cost per product or service, which can be compared with actual costs on a biweekly basis.
For years, Yale New Haven has used a flex-budgeting process to match volumes to budget, adjusted for case mix changes. If volume on a nursing unit increases, the budget increases as well. If volume decreases, so does the unit’s budget. However, leaders recognized that managers needed a better way to translate changes in volume to staffing changes. To that end, analysts in the finance department created a tool, called the payroll trend report, to help managers determine whether they are meeting theirstaffing per unit of service goals for their cost center, trended over time.
The tool includes benchmark data, such as targeted worked hours per unit of service, as well as budgeted data and actual worked hours. The color-coded report allows nursing and clinical leaders to quickly determine where they are managing above target and below target. Each pay period, every cost center receives its own report, which pulls volume data from the EHR and cost data from the cost-accounting system.
These integrated data allow department managers access to specific patient population trends over time, including shifts in quality outcomes. By clarifying for clinical and finance leaders the ways quality drives cost, the patient population quality trended data play a critical role in helping both parties understand past expense variation while working together to plan appropriately for future resource needs.
Prepare for Bundled Payments
As bundled payments become more of a reality, health systems need to test new technologies and processes to help them prepare for the journey from volume to value.
To prepare for bundled payments, finance leaders at Yale New Haven are completing the rollout of a new data warehouse and analytics tool that will generate a more accurate cost of an episode of care, thereby providing more accurate utilization and cost data across the continuum of care so clinical leaders can more effectively manage their business in a bundled payment environment. For example, clinical and finance leaders in orthopedics are evaluating the quality and cost benefits of keeping patients undergoing hip replacements a few days longer in the acute care setting, thereby avoiding admissions to skilled nursing facilities, and, instead, discharging patients directly home with home health services. Although this process-of-care change increases the acute care cost, it potentially lowers the overall episode cost while also bringing a strong focus on improving patients’ overall functional outcomes.
In April, Yale New Haven began participating in 11 bundled arrangements through the Centers for Medicare and Medicaid Services’ Bundled Payments for Care Improvement initiative. Clinical and finance leaders at the system are actively tracking QVIs and other metrics, such as LOS in skilled nursing settings and readmissions, to identify process changes that can help reduce episode costs while maintaining or improving quality. The finance team is also implementing its cost-accounting application in the system’s physician practices to track productivity and clinical effectiveness measures, which will create a more accurate view of cost per case for the entire episode of care.
As more organizations like Yale New Haven take on bundled payment arrangements, a greater percentage of their revenue will depend on quality outcomes. In such an environment, accurate cost-accounting data are critical. Balcezak says: “Once you have that data, you can use it to forge relationships between people in finance who understand how care is paid for and the clinicians who ultimately drive the care.” In this way, reducing inappropriate clinical variation can become the common ground for two camps that are still learning how to work together effectively to deliver better outcomes for patients and payers alike.
Steve Allegretto, CPA, MPH, is system vice president, analytic strategy and financial planning, Yale New Haven Health System, New Haven, Conn., and a member of HFMA’s Connecticut Chapter.
Dan Michelson is CEO, Strata Decision Technology, Chicago, and a member of HFMA’s First Illinois Chapter.