Save 3%-5% a year with integrated outcomes

Combining data gives you complete picture

Organizations willing to take a coordinated approach to analyzing data will find many opportunities for routinely improving outcomes and reducing costs by 3% to 5%, according to a researcher who has studied 200 successful outcomes projects. However, he cautions, introducing benchmarking data into the improvement process too early can be counterproductive.

"There are incredible pockets of fiery success, places that are just hotbeds of activity in implementing integrated outcomes," says Steve Shaha, PhD, research director for the Gartner Group, Healthcare Advisory Services in Wakefield, MA, which offers health care information technology and advisory services. "The trick is the process that underlies integrated outcomes because if the process is implemented correctly, then the success is a natural outgrowth. It's called `integrated outcomes' because it simultaneously addresses and optimizes financial, clinical, and satisfaction outcomes."

Shaha has visited or monitored more than 75 health care organizations and collected outcomes results from more than 200 projects. He says the average initiative will lower annual costs by 3% to 5% for each disease process for which integrated outcomes are implemented. Some organizations have achieved cost reductions from 7% to 12%.

Based on his experiences and those of other health care organizations that achieved the same results, Shaha offers the following suggestions for your benchmarking programs:

1. Align your intent with the goal of improving care. "The intent of collecting all the data and information should be to help people do better," he says. An organization that intends to use benchmarking data and baseline information to penalize or scrutinize physicians and employees will not design an integrated outcomes process that succeeds in saving money over the long term.

"If an organization uses a decisions support system designed primarily for administrative feedback, then they will fail to get savings or to engage the people who can make the real difference."

Health care organizations with this scrutinizing approach tend to turn the process into a punitive one, in which physicians and employees resent changes and refuse to buy into the process. On the other hand, if the organization's intent is to make the process customer-focused, with the data collection and benchmarking seen merely as a means to learn about new ways of doing their jobs, then positive changes will occur naturally, he says.

2. Market the process to practitioners and staff. This is where the physician or staff buy-in will occur, unless managers introduce benchmarking into the process too soon, he says. For example, if physicians are presented immediately with benchmarking data from outside organizations, they may become defensive and take a dislike to the whole process.

Shaha recommends organizations make a presentation to physicians that describes the intent, methodology, and examples of successes at other organizations. It could show how to improve medical outcomes, patient satisfaction, and costs by participating in an integrated outcomes project. If the presentation is done right, a few physicians will enthusiastically embrace the idea and emerge as champions, he says. Those champions become the impetus for the first implementations.

3. Prioritize the processes to be improved. Shaha says organizations should start their efforts by focusing on those processes with the highest likelihood of having a high return on investment. Five criteria for prioritization are:

· high-cost processes;

· high-volume processes;

· high risk to patients, providers, or organization as a whole;

· problem-prone areas, in which things keep breaking down;

· an initiative with system-level or strategic impact on the organization.

Those criteria should be analyzed for each integrated area: clinical, administrative, and operations. In the clinical area, for example, the hospital might focus on hip replace ment surgeries or coronary arterial bypass grafts because of their high costs, volumes, and risks. On the administrative side, the organization might target accounts receivable days to see if the billing cycle could be shortened. And under operations, the organization could look at billing practices, patient waiting times, admission, or registration processes.

"There is benchmark information available in each of these three areas from various sources," Shaha says. Once the top priorities are selected, the organization needs to assess how many resources are available to invest in the integrated outcomes venture or initiative. If the organization can spare only enough time to develop one team, it should select one or two processes to improve rather than spreading the team too thin and failing to achieve desirable outcomes on all of them, Shaha says. "The availability of a champion should influence the final priorities," he adds.

4. Ask physicians/staff to select outcomes to study. Start by asking the more enthusiastic staff or physicians to begin the pilot process, Shaha advises. They will start to collect data within their areas of specialty and compare each physician's information at the monthly meetings.

This first effort may take three to six months because the team is defining the process as it goes. Once they've completed the data collection for one area, it will be easier and much faster to do it for a second. Here's how it typically works for a clinical initiative, Shaha says:

Physicians take the lead in identifying processes and outcomes to implement, including cost, clinical, and patient satisfaction. The resource team begins to use those measures primarily through existing data sources but adds others, such as satisfaction surveys, as needed. The team examines the data to discover where the outcomes variations exist. At the same time, it identifies current baseline performances. The comparisons raise questions that lead to process improvements. "If you're looking around the room, and seven people are doing the same thing and getting different outcomes, then they will ask why and want more information," Shaha says.

Physicians whose results are not as promising as a colleague's ask the colleague for more details about how the better results came about. Then the physicians agree to study more details and meet monthly to compare them. After comparing their ways of handling cases, Physician A, who has poorer outcomes than Physician B, might decide to try some of the techniques used by Physician B. At the next meeting, Physician A reports whether those changes helped improve outcomes.

Shaha says this type of process doesn't have to be written or mandated because by nature physicians are scientists hungry for data. "People learn from variation, not from similarity."

As the physicians refine their treatment methods, improvements will occur in clinical outcomes, patient satisfaction, and costs, he adds. "You don't get lower costs by telling doctors how to practice," he asserts. Rather, administrators should curb their natural tendency to try to control people's behavior, especially when it comes to physicians. "They should establish data-driven feedback [without results] as the impetus to getting the best results," Shaha says.

5. Introduce external benchmarks into the process. This is the tricky part because there could be problems if benchmarks are introduced too early in the process. First, physicians may resist and view benchmarks as punitive if they are thrust upon them. Second, it's difficult to judge whether the benchmarks are meaningful to your organization's data. For example, a hip replacement might be listed as costing $3,500 at one hospital and $6,800 at yours, he says. But is that because your figure includes costs after the patient was transferred into transitional care, for example, whereas the other hospital did not include those costs?

That's why the benchmarks can't be introduced in a critical or punitive way, he says. "It's an enormous danger because people are being held accountable for benchmarks of which we know too little." Instead, benchmarking data should be presented as simply another source of information that may be useful to the integrated outcomes process. The best way to do that is to wait until the physicians ask for external benchmarking data, Shaha says.

"It has been my experience that in most organizations, the physicians will ultimately request benchmark information during the process. And they will say, `What is the national average for that?' or `What are others outside our organization experiencing?'"Then the administrator may respond, "Let me do some research, and I'll get back to you with the information."

Those kinds of steps also may be taken with other departments and employees. "Use the same approach but aim it at the people who impact the process," he advises. "If you want housekeeping costs to go down, then start with housekeeping people: Get their input; build systems around the people who make a difference, and improvements will happen."