Article Images

Although much has been written about bending the cost curve, the real health care challenge facing our nation is not costs, but something more significant: value. At Christiana Care Health System, Wilmington, Del., the value proposition for health care is rooted in the fundamentals of good medicine: listening to the wants and desires of patients; partnering with patients in their care; ensuring the highest quality and safety in caring for patients; and taking a hard, objective look at the cost of the care. For Christiana Care, value is making a measurable difference in people's lives through what we as medical professionals do, in ways people appreciate and our society can afford.

Meaningful and Measurable

Part of the Affordable Care Act, the Hospital Value-based Purchasing Program provides $850 million in incentive payments for hospitals that improve the quality of care delivered to Medicare patients. This was a driving factor in Christiana Care's decision to rethink value.

"We are a mission-based, nonprofit organization and managing our resources is critical to advancing our mission of service," says Betty J. Caffo, chair of the system's quality and patient safety committee. Considering that Medicare reimbursement rates are tied to discharges, quality of care and patient experience, "we needed to take a hard look at where we invested our human and financial resources."

Christiana Care's leaders have another reason for their work on measuring value. Delaware demographically represents a microcosm of the entire nation, placing the system's clinicians in a unique position to create models of care from which all providers in the country can learn.

To meet its value-oriented goals, any model that Christiana Care developed had to include the most rational and thoughtful use of resources. It also had to have physician support.

"If you're going to do things that are going to impact the physicians, you have to have them buy into it early in the process," says Joseph A. Lieberman III, M.D., a trustee who serves on the board's quality and safety committee. "You better have them on board from the get-go to make sure the idea is credible and has a chance of being accepted. Otherwise, it will never gain any traction."

Sharing a bunch of statistics that showed where doctors needed to improve was not enough to instigate engagement and change. Rather, the data needed to be distilled. So leaders created a solution that was as elementary as it was novel. In March 2010, they unveiled the value scorecard concept.

The idea was to build on something deeply embedded in U.S. culture: letter grades. Doctors, nurses and other caregivers spend years in academia and the concept of the old-fashioned report card resonates with them. Data collected on numerous metrics could be normalized to a 100-point scale and a standard alphabetical grading system could be implemented: an A+ was a grade of 97 points or higher, an A was 93 to 96 points, an A- was 90 to 92 points and so on.

Critical to the endeavor was that the data be pulled together to produce a single, simple score. Letter grades cut through the complexity. For instance, a grade of B- for the treatment of an illness on a unit is definitive and memorable. More importantly, the grade fueled hospital teams to improve.

"If a clinician gets a score of 212 on some performance measure, it is ambiguous," Caffo says. "On the other hand, a grade of C automatically telegraphs in a clear way the opportunity for improvement. The grade gets their attention and immediate engagement."

Article Images

The Scoring Model

At the follow-up board meeting in May 2010, the following five quality metrics for the scorecards were approved:

Mortality: The percentage of patients who died in the hospital during a given time period;

Morbidity: The percentage of patients who developed complications during a given time period;

30-day readmissions: The percentage of patients who experienced an urgent readmission that occurred within 30 days from the day of their discharge for the original condition;

Guidance compliance: The percentage of patients who received all the appropriate evidence-based treatments;

Patient experience: The percentage of patients who rated the hospital a "9" or "10" for overall hospital experience based on a zero-to-10 rating scale (with zero being lowest and 10 being highest).

Including the patient experience added a special dimension to the scorecards. Traditionally, value is seen as a calculation of cost and quality. Incorporating the patient's voice makes Christiana Care's formula more robust.

Formulas already in existence were leveraged to calculate the measures. Mortality and morbidity rates were risk-adjusted and compared against national rates for both categories, available in a Truven Health Analytics database that contained about 21 million annual discharges.

That database lacked readmission rates, so Christiana Care's current readmission rates were compared with its own rates from the previous year. Scores for each metric were derived by comparing the measure to the risk-adjusted expected rate or established target, similar to an observed-to-expected ratio. To prevent any single metric — mortality, morbidity, readmissions — from having too much sway, a finite range of plus or minus 25 percent was established.

The guideposts for the standards included the Joint Commission, the National Quality Forum and the Centers for Medicare & Medicaid Services. The patient experience metric was derived from HCAHPS, the Hospital Consumer Assessment of Healthcare Providers and Systems survey.

Because costs also were crucial, two more metrics were included:

Length of stay: The number of days spent in the hospital;

Estimated direct costs: The direct costs for patient care based on patient charges.

Board members and senior leaders reasoned that including length of stay would encourage health care teams to use their capacity more prudently. The severity-adjusted expected length of stay was divided by the observed length of stay among patients. That figure also was multiplied by 100 so the score would not unduly influence the overall grade.

The direct costs metric then was determined by using the costs-to-charges ratio, a Truven algorithm that estimates the direct costs for patient care based on patient charges. Board members approved this ratio because it helped to overcome the cumbersome task of developing a new cost equation that accounted for group rates, predetermined discounts and capitation agreements that are connected to Medicare and other third-party payers. Length of stay and estimated direct costs were severity-adjusted and compared against national rates using the Truven database.

The next step was normalizing all those metrics to a 100-point scale. Because the scorecards were measuring value, board members and senior leaders determined that cost should wield significant influence on the overall grades. Therefore, the direct costs and utilization metrics would comprise half the score of the final grade. The other half would be equally represented by the five quality metrics — mortality, morbidity, readmissions, guideline compliance and patient experience. Those metrics were weighted individually to 20 percent of the overall quality score.

"By considering costs in the value scorecards, we were energizing physicians to have a discussion around the efficiency of their decisions," Lie-berman says. "We could no longer be concerned exclusively about quality of care because understanding the costs was also relevant to understanding the value."

A pivotal foothold in the development of the scorecards was gained by seeking physicians' input. From the moment the concept was created, Christiana Care leadership enlisted the help of the medical chairs and directors throughout various departments of the health system in developing and championing the new grading system. As a result, support already was brewing among the front-line health care staff when the value scorecards were officially rolled out in August 2010.

"Doctors know that medicine is just as much an art as a science and they don't want to see the science overwhelm the art by the administrators," Lieberman says. "Since the doctors were part of the process early on, the value scorecard initiative was much more likely to be endorsed after it was launched, and it was going to be endorsed by the people in the hospitals who already were considered the opinion leaders."

Another key was identifying the subjects to grade. Leaders zeroed in on grading the treatment of patients in seven high-volume, high-risk categories: hysterectomy, pneumonia, stroke, heart failure, cesarean section birth, total knee replacement and vaginal delivery birth. The care for those patients generally spans the cardiovascular, neurologic, orthopedic and obstetric/gynecologic disciplines.

That move generated a sense of mutual support. Various specialists and nurses in separate units had a hand in the care of each patient. By categorizing the grades on the patient population's medical problem, health care workers in separate departments shared ownership of a grade and subsequently were compelled to work together. Hospitalists and cardiologists bore the responsibility for heart failure's grade; orthopedists and pulmonologists bore responsibility for pneumonia's grade. The grades also were accessible to the health system's 10,500-person workforce through the employee portal.

"With the grading system, we were able to focus on why one unit had a higher rate of infection than another unit for the same illness," Caffo says.

A baseline was created by using clinical data from 2009. To establish trends, data also were collected from 2008. The results from the original baseline — released in August 2010 — sparked lively discussions and debates.

Grades ranged from a high of a B+ in total knee replacement to a low of a C in pneumonia. Despite scoring differently on various metrics, the baseline value grades for hysterectomy, stroke and heart failure were each a B-. Keeping the measurements from attaining a higher grade were the two cost metrics — length of stay and estimated direct costs. In five of the seven measurements, the pair of cost scores was poorer than expected. Only in total knee replacement did a cost metric — length of stay — perform better than expected.

Unexpected Collaborations

Some staff members maintained that costs should not be included in the scorecards, but senior leaders and trustees remained committed. Results from the scorecard released one year later revealed that the original concept was bearing fruit.

Grades in five of the seven measurements had improved. The grades for both total knee replacements and vaginal deliveries rose to an A-. The hysterectomy grade rose eight percentage points, to a B+ from a B-. The grade for heart failure rose six percentage points, to a B from a B-. What's more, in 12 of the 14 cost categories, the expense of care had shrunk.

Upon closer examination, you could see why. The original value scorecards spawned a number of grassroots initiatives. Orthopedists standardized joint replacement processes at both hospitals that resulted in reduced length of stay and improved compliance with the Joint Commission's Surgical Care Improvement Project. An initiative to share pneumonia patients' data with Christiana Care's hospitalist groups resulted in a reduction of radiology diagnostic tests to a rate 16 percent lower than the national norm. The baseline grades for hysterectomies spurred the accounting staff to share cost data with gynecologists that propelled them to cut down operating supply costs for the procedure by 15 percent, translating to a one-year cost savings of $415,000. Improved cohorting of heart failure patients helped to bring down length of stay by 15 percent, which resulted in a one-year cost savings of more than $204,000. Meanwhile, the scores for both guidance compliance and patient experience among those same heart failure patients improved.

The value scorecards continue to evolve as Christiana Care applies the grades systemwide. The newly established Christiana Care Value Institute, which is designed to focus on research that can make care more effective, evidence-based and efficient, also serves as a vehicle to refine the scorecards.

Ready for Better Reimbursement

In preparing for the federal value-based purchasing program, Christiana Care has added a new component to its scorecards. The value-based purchasing impact evaluates the grades against the value-based purchasing program's formula. This determines how an estimated $850 million in incentives will be distributed to hospitals. Those that perform well will see Medicare reimbursements increase, while those in the bottom half will experience decreases. With value scorecards, Christiana Care already is focused on creating greater value for the patients we are privileged to serve.

Robert J. Laskowski, M.D., M.B.A. (rlaskowski@christianacare.org), is president and CEO, and Sharon Anderson, R.N., B.S.N., M.S., FACHE (sanderson@christianacare.org), is senior vice president, quality and patient safety, Christiana Care Health System, Wilmington, Del.