Case Study: Stroke

Overview

"Public reporting and pay for performance for stroke is around the corner," said Dr. Lee Schwamm, Director of Acute Stroke Services at Massachusetts General Hospital and a member of the American Stroke Association's Advisory Committee. "So it's time to get on this."

Providing high quality, comprehensive stroke services offers hospital administrators that rare opportunity in life – doing well by doing good. Evidence-based guidelines and performance measures allow well-managed stroke programs to save lives and improve outcomes. But a movement towards pay-for-performance and other accountability measures in this area means that improving stroke performance is key to adding to and maintaining the bottom line.

There are good business reasons for doing so. A well-managed stroke program should be able to shorten lengths of stay to 5-6 days, mostly by preventing complications that can drag out hospital stays, Schwamm said. This allows hospitals to capture more revenue from these patients. That's particularly true now that CMS has approved a new DRG for treating stroke patients with clot-busting drugs that nearly doubles reimbursement for this procedure, to about $12,000.

"If you have a stroke system in place that really functions well, that's enough to swing a stroke patient from a break even patient to a pretty profitable one," Schwamm said.

In addition, stroke patients typically have systemic cardiovascular disease, which means they may require additional cardiac procedures that can boost the bottom line.

There are also competitive reasons for improving stroke care. As regional and national efforts to set up and certify high-performance stroke centers intensify, hospitals may need to keep up so they don't lose patients. This has already happened in Massachusetts, where the threat of diverting patients to state certified stroke centers resulted in nearly every hospital in the state setting up an acute stroke program.

Because stroke has an evidence-based set of standard guidelines that can be boiled down to 7 or 8 measurable steps, it lends itself to a pay-for-performance approach, Schwamm said. There is currently an effort on the part of the American Stroke Association, JCAHO, and CDC to create a consensus version of their various stroke guidelines that would be standardized across the nation. That effort should be finished in February and once it is, the guidelines will be submitted to the National Quality Forum.

From there, Schwamm warns, it's a short step to pay for performance programs or public reporting of hospital performance on stroke outcomes.

"In the few conditions where we have a well-evidenced set of guidelines, actionable items with well-defined performance measures…there's a clear movement to respond to that," said Schwamm. "That's the climate we want to create. We want to create competition for the highest quality."

This should translate into a significant number of lives saved and patient quality of life improved. Each year there are about 700,000 strokes in the U.S., about 200,000 of which are second attacks, according to the CDC. Stroke is the third leading cause of death and a leading cause of permanent disability. In 2006, stroke is expected to cost nearly $58 billion in healthcare services and medication, along with lost productivity.

Hospitals have three opportunities to improve stroke outcomes if they implement care guidelines and manage their stroke programs well. One is to maximize the use of clot-busting drugs that can reverse the effects of a stroke and prevent long-term disability. These drugs require rapid, coordinated acute stroke care to insure they are delivered within the narrow window after a patient arrives at the hospital. The second is to prevent complications in the days immediately after a stroke by implementing care guidelines that call for performing swallow tests to prevent aspiration pneumonia or efforts to prevent blood clots from forming in the legs. The third area is to start patients on whatever preventive regimens they may require to prevent a second stroke, like better management of diabetes or high blood pressure. About 40 percent of stroke patients go on to have a second stroke within five years but that risk can be reduced by 19-34 percent by giving them aspirin therapy, said Ann Pianka, Stroke Program Manager at Beverly Hospital in Massachusetts.

"It's a matter of teaching people there are things we can do now," said Pianka. "When I went to nursing school 20 years ago, there wasn't anything we could do for stroke patients. They came in and we put them to bed and if they survived we got them into rehab. Now we have research that shows there are things we can do but we're 20 years behind the times in terms of setting up the systems."

Setting up those systems requires a commitment to long-term monitoring of performance measures and a willingness to keep experimenting until you design a system of care that really works, said Schwamm. But that all starts with good data, properly analyzed to make sure performance targets are being hit. At the moment, that's often a barrier to designing effective care systems because the burden of gathering all that data and analyzing it strains hospital resources, Schwamm said. Many hospitals in the American Stroke Association's Get With the Guidelines program are struggling just to fill out the 133-field form the program uses. He hopes that form will be pared down to just the essential data elements when the consensus guidelines are finalized. That would lessen the burden of data collection and free up resources to actually improve systems of care.

"We're still early in this effort and we haven't matured yet as a healthcare system to better align the way we document the care we'd like to provide with the way data is abstracted," Schwamm said. "There needs to be a balance between the burden of data collection and the goals of quality improvement. "

"I think it's really important for hospital administrators to realize the data issues need to be addressed and they need to provide systems and support for clinicians who are trying to do the right thing," said Schwamm. "Any investments they make in acquiring data more efficiently will pay them back in spades because of the current environment in terms of quality improvement and pay for performance in stroke."

Case Study

The Institute for Health Metrics is easing the data collection burden at hospitals who use it to manage their stroke programs, freeing up time and resources to better manage patient care.

Ann Pianka, Stroke Program Manager at Beverly Hospital in Massachusetts, used to sift through hundreds of paper files to find the 40 or so stroke and TIA patients she needed to track. She also used emergency room and admission logs as well as computer runs to help identify patients. But IHM data allows her to find patients with specific diagnoses in a single place, cutting down on the detective work and helping her identify patients she'd missed.

"There are so many codes we're looking at in so many different places," she said. "Having the added tool will help streamline a lot of that. It has certainly highlighted places in my system where I've lost patients for whatever reason. It's plugged some holes so my numbers are better than they were before."

Pianka is optimistic that will cut down the time she spends on data collection significantly. For Lynn Scannell, who handles data collection for the stroke program at Emerson Hospital in Massachusetts, IHM has already reduced the time she spends on data collection almost in half. Before she got access to IHM's data, she was a familiar site at medical records, where she had to root through paper files to fill out the 133-field form required by the Get with the Guidelines program of the American Heart Association. Not all the data was in the paper records so she frequently had to consult the hospital's computer systems as well. But that was time-consuming because the hospital's system simply didn't track data the way it was organized in the Get with the Guidelines form.

"The measures Health Metrics pulled were sort of the ones that were the highest priority," Scannell said. "Things like was the patient being treated with aspirin on discharge…well, now that's easy to find. It has just made data collection so much easier because it pulls things together in a logical way."

Over time, Scannell has worked with IHM staff to customize the product to better serve her needs. In addition to the stroke-specific measures IHM tracks, Scannell has learned to use it to narrow down her search for other information. Because IHM provides a summary of the patient's history and medical record that is searchable, she can easily identify much of the information she needs to track. If the information isn't there, it often helps her figure out whether she needs to search a paper record or the hospital's computer system. In a fragmented system, it's the closest thing to one-stop shopping for medical information that she's found.

"When you're struggling to pull all this together, it's a tremendous resource," she said.

It also helps her check her work and easily pull up information about the hospital's performance in treating stroke. Because IHM data sets are pulled from the hospital's own computer system and then organized in a way that makes it easier to track quality of care trends, she knows the data is accurate.

"It's just a great validator," Scannell said. "If somebody wants to know the average length of stay with our stroke patients it's right there. It's just made the process so much more efficient."

Though Pianka has only been using the IHM data for a few months, it has already helped her spot an important trend. Not all stroke patients at Beverly Hospital were being treated using the hospital's stroke orders, an important glitch because when stroke orders are used the hospital's rate of hitting stroke performance measures rockets up.

"I was beating my head against the wall for a while," Pianka said. "I kept looking at the data sets to see if the problem was an individual or a location and it turned out to be a location. IHM helped me confirm that."

It turned out stroke patients admitted to the hospital's cardiac care unit instead of the stroke unit were being treated with cardiac care orders instead of stroke orders. Now she's working on getting stroke orders implemented in that department. She expects it to improve the hospital's performance significantly.

She also hopes to use IHM to help her identify individuals who may be failing to fully implement the hospital's stroke orders so she can target them for individual attention if necessary. She hopes to broaden her use of IHM as she learns more about the product.

"Maybe we'll see trends we didn't expect to see," Pianka said. "I think the potential is boundless."

Tips

If you're interested in starting or improving a comprehensive stroke program, here are some tips from the experts and those already doing it:

Work in real time. Because data collection is so difficult and time consuming, many hospitals struggle just to put together a retrospective review of patient records to monitor performance. This can help spot larger trends but it typically provides answers too late to do anything about a problem. The biggest improvements in performance come from catching problems while patients are still in the hospital so they can be fixed. Some hospitals assign a nurse or other stroke coordinator to pull patient records daily and check in with staff to see that performance measures are met. Another solution that some hospitals are developing is to create a computer "dashboard" that provides staff a daily look at where stroke patients are in the system and what care they've received. "That's the best chance of cracking the 100 percent barrier," said Dr. Lee Schwamm, Director of Acute Stroke Services at Massachusetts General Hospital and a member of the American Stroke Association's Advisory Committee. "Decision support in real time will enable you to track adherence to these guidelines in vivo and intervene in time to make a difference."
Give data collection its due. Many hospitals try to pass off data collection to an overworked staffer who handles data for several other programs as well. Or they try to have a stroke program coordinator handle both data collection and implementation of care guidelines. But data collection takes up so much time, it tends to eat away at the ability of staffers to focus on figuring out how systems of care need to be changed and on making those changes happen. Stroke programs seem to work better when hospitals put enough resources into data collection to free up a staffer's time to focus on the process of implementing changes, Schwamm said.
Get someone who can drive the process. You need to put someone in charge whose focus is on changing systems and who has enough clout to get hospital staff to cooperate. What seems to work best is to assign a physician to work with physician peers, said Schwamm. But it may mean dedicating at least one person full-time to the process. Handing off responsibility for stroke services to an already over-worked ICU manager doesn't work, experts say. "The best advice I have is what administrators don't want to hear,'' said Angie West, Neuro/ Stroke Program manager at Long Beach Memorial Medical Center in California. "You need to have a dedicated person who's passionate about making this happen. I really think it's a two person job. You need one person to drive it and one to do the paperwork."
Get out of the office. Sometimes data doesn't tell the whole story or can only hint at a problem. It may take some detective work to figure out what's going on and that means freeing staff from paperwork so they can walk the floors and talk to those directly involved in patient care. West couldn't figure out why her patients weren't getting enough stroke education until she spent some time on the nursing floors and realized nurses simply didn't have time to do it. Her solution? Train physical therapists to do it because they have more time to spend with patients and can talk to them while giving therapy. She's also found it helpful to attend patient support groups because patients have given her a lot of feedback about their experiences in the hospital. That's pointed to deficits she's now trying to fix.

Understand the cycle of quality improvement and be patient. When you first implement a quality improvement program for stroke, you're likely to see big changes relatively quickly. But there's a natural human tendency to backslide into old habits, so you'll need to consider how to keep the momentum going for the long haul. That means going beyond monitoring of performance measures to develop a system of constant small improvements. You'll have to get quite sophisticated about the ways you use data to look for the small improvements that can add up. You'll have to train staff and provide support to help them analyze data, generate hypotheses about what's going wrong and then test and fix the problem, said Schwamm. "The low-hanging fruit will give you a big bump at first, then you see a period of stagnation even though you're doing a better job and it make take eight to twelve months to see improvement,'' he said. "System change takes a while. "

For more information about how IHM can work with your hospital, contact IHM today:

      Jen O’Brien | jen.obrien@healthmetrics.org
      Institute for Health Metrics
      +1.781.328.3010
      One New England Executive Park
      Burlington, MA 01803
      http://www.healthmetrics.org