Special Report: Focusing on Performance

Don’t just collect data — improve outcomes with process improvement

Goal is to use data most effectively

While all rehab facilities regularly check process data, the odds are that very few are using the information as effectively as they could with regard to performance improvements. By instituting a program to improve the use of quality, safety, patient outcome, and other outcome data, a facility could improve costs, efficiency, and regulatory compliance.

"Over the last three years, we’ve really focused on making sure that we’re using data in an effective way to drive program, process, and operations," says Joan Alverzo, PhD, CRRN, chief clinical officer of Kessler Institute for Rehabilitation of West Orange, NJ, a Mechanicsburg, PA-based Select Medical Co. organization. "We measure outcomes 90 days after discharge to make sure the rehab process is continuing to work to help patients lead an independent life, and we do that with all patients," Alverzo explains. "What we’re focusing on now is we’re really trying to look at data in a way that drives decision making about how to improve things."

The decision-making process is crucial

Making the most of the decision-making process is crucial to continued organizational success, says Sue Kida, PT, MHA, assistant vice president of administration. "A lot of organizations collect data, and many compare it against some type of benchmark; but what most organizations struggle with is improving data through a process change," Kida says. "How can you look at data, share it with staff, involve staff, and make process changes that affect functional outcomes of patients?"

Alverzo and Kida discuss the best strategies for improving an organization’s performance improvement process with these tips:

• Understand your data.

"We collect quantitative and qualitative data," Alverzo says. The quantitative data area center around the functional improvement measures (FIM) scores on admission and discharge, she says. "One of the important things about collecting that kind of data is to make sure it’s reliable and accurate," Alverzo explains. "So we have two people score the patient at the same time, and we see if the scores match."

This results in giving administrators confidence that the data accurately reflect how a patient is performing without bias from the tester, she adds. By daily checking FIM scores, administrators can see how efficient process improvements are, Alverzo says. "We look at rehab length of stay to see if the intensity of services and patient improvements are happening with the highest amount of efficiency compared with other patients with the same measurements," she says.

"We participate in a national database called e-Rehab data that provides us with benchmark data for the region and nation for all rehab hospitals participating in that database."

Qualitative data also provide important information that can be used when improving processes, Alverzo explains. "With qualitative data we look at patient complaints and other patient feedback and attempt to identify trends in complaints by department or based around the admitting or discharge process," she says. "We value this data in terms of letting us know where we’re falling short in the eyes of our patients, and we can make changes based on that," Alverzo notes. "Patient focus groups are administered each month, and these really open up a free-flowing conversation about how their stay is going," she continues.

Staff take notes on patients’ comments, grouping these into positive and negative comments and sorting them into categories in which some process might be improved, she says. "So we have trends over time about whether a process improvement has changed feedback from patients," Alverzo adds.

The right staffing mix

• Focus on staffing efficiency.

"We look at staffing effectiveness variance in our two major clinical provision areas: nursing and therapy," Kida says. "For therapy, we look at it based on several factors, such as census and case mix." There are various patient diagnoses that can be assessed for frequency and intensity of services needed by patients, she explains. "We use various formulas for determining percentage of staffing each day, and we have a benchmark level, a threshold where we anticipate staffing at a certain percentage," Kida says. "By using it prospectively, we can look at future staffing."

With five years of experience in managing staffing in this way, administrators are confident that so long as the staffing levels are within a pre-determined range, then patients will not be placed at risk of falls or other problems during their stay, Alverzo says. Five years ago, CARF determined a target nursing care standard in rehabilitation of 5.5 nursing care hours per patient day, she says. "We intuitively came up with variations from that standard based on the different groups of patients we treat," Alverzo explains.

Designing a pilot project

Some patients will require more than 5.5 hours per day and others less, so administrators divided patients by diagnosis into six different categories and developed a yearlong pilot project around those calculations, she says. "We measured staffing and patients in each category to get a sense of whether or not the formula worked," Alverzo notes. "At the end of the year, we made some changes; and we’ve been using these same hours of care standards for the past four years."

Occasionally, they’ll tweak the numbers. For example, administrators decided ventilator-dependent patients require a great deal more nursing hours during their first three days of stay than after that, so the staffing expectations were adjusted accordingly, she says.

As an offshoot of the nursing care standards, administrators also developed a therapy care standard, Kida points out. "We developed formulas of care treatment frequency and intensity necessary based on the number of patients we’re treating, as well as the diagnoses because certain patients require more or less occupational therapy or physical therapy," she explains. "We trialed our formulas for a year, and once we were comfortable with staff hours, they intuitively met the standards."

Kessler is a pioneer in this area of developing therapy staffing standards, according to Kida. "These variance numbers have aided us in determining how to ensure safe and effective treatment of patients on a day-to-day basis," she notes. "And they’ve helped us proactively look at budgeting therapy needs, basing these on clinical intuition and intensity of therapy needs per patient," Kida explains.

• Learn effective data analysis.

Before data can be effectively analyzed, a facility must make certain the information is accurate and efficient, Alverzo stresses. "So if we’re looking at FIM scores, we want to know the ages of patients, and we may want to know more specifically what their comorbidities are and additional information," she points out. "So when we’re evaluating the data, we have enough supportive data to answer the questions that are going to come up first," Alverzo explains.

For example, administrators may want to know why functional improvement at one site is different than it is at another site. Is this because the ages of patients are different, or does one site has more comorbidities or diagnoses that interfere with therapy? Or is it because the patients’ conditions are more severe at one site than they are another?

"We have all of that data in front of us so we’re answering questions instead of just asking questions," Alverzo says. "One of the major challenges that many organizations face when they analyze data is to do it in a way that’s legitimate and not to bend toward the tendency to want to explain away the data or misinterpret data," she continues. "So we applied some discipline to data analysis to make sure we’re not analyzing data unless we have sufficient sample size, enough patients," Alverzo adds.

Also, administrators look at data over time to obtain a sense of normal variation, she notes. "There’s a technique in quality management where you’re taking data over a period of time and looking at how variable it is, and you’re basically creating ceiling and floor numbers," Alverzo says. "That helps us avoid focusing on minor changes that may not be meaningful," she notes.

Studying the process

Also, it’s important to go beyond coming up with reasons why data are different in one situation than another, Alverzo advises. "We try to get beyond that question and say, How does that process work at that facility?’" she says. "Perhaps the data are really telling us more about the fact that we’re treating patients in a different way and getting a different outcome," Alverzo contends. So it’s possible that patients are treated differently at one site, and if another site were to use the same processes, the outcomes would be similar, she explains.

• Set goals.

The next step is to determine where to focus attention and set goals in the areas where improvements are most important, according to Kida. "In quantitative data, we look at the discharge disposition and percentage of patients discharged into which environments, such as home, skilled nursing facilities, acute care facilities," she points out. "Our goal is to discharge back to the home." Administrators saw a trend of higher than desired discharges of patients to subacute facilities, Kida says.

Quarterly updates

• Monitor and track outcomes.

"Most of the data we monitor are standing indicators," Alverzo says. "We have added additional indicators because of concerns, and we’ll monitor those for four quarters before we make any changes," she explains. It’s important to look at a substantial data set each quarter because this will reflect the key elements of a program and patient quality indicators, and it provides data consistency that is necessary for observing trends, Alverzo points out. "Finding the right data is a combination of having a team of people who understand the clinical program and processes that are the core of the organization," she notes.

Since the staff are essential to both the collection of good data and the implementation of process improvements, it’s also important to share findings with them, Alverzo says. Each quarter, administrators give staff, patients, physicians, and others a stakeholder report that shows some of the indicators that have been addressed and what the data indicated about the facility’s quality improvement and patient outcomes, she notes. "This can be used as a marketing strategy to increase referrals to our facility," Alverzo adds. "We try to make it user-friendly so a patient or family could read it and understand it, but it also has important information a physician would be looking at."