Computerized outcomes — One size doesn’t fit all

Beware of glib software sales pitches

Nearly every vendor worth its weight in data claims its outcomes measurement system is the perfect fit for agencies gearing up to integrate the Outcomes and Assessment Information Set (OASIS) into their patient assessment system.

They all are a perfect fit. Yeah, that’s it. That’s the ticket. And they’re inexpensive, too. And, yeah, we’ve got some oceanfront property in Arizona to sell you.

With your hospital executives or your agency director pressuring you to fish or cut bait, you may make a hasty decision that you’ll regret later when the Medicare Conditions of Participation require agencies to use OASIS.

Having a computerized outcomes system is not a requirement of the current proposal that would require the use of OASIS, says Mary Vienna, co-director of the Center for Hospital and Community Care in HCFA’s Health Standards and Quality Bureau. However, "it’s unlikely that agencies would easily be able to develop outcomes measures without some sort of electronic capability," she tells Homecare Quality Management.

So here are nine tips from the experts for choosing software that will make your job as quality manager more like a day at the beach and less like being lost in the desert:

1. Don’t rush.

Fools rush in, and you’re no fool. Resist the strong-arm sales tactics of some vendors. "You don’t have to jump into anything right away," says Vienna. Home care doesn’t have to comply until 1998. Take your time.

2. Research the family before the marriage.

Saying you have to choose a reputable vendor may sound obvious, but you’ll need to pick a vendor that has a reputation for longevity and a commitment to electronics in the home care field, says Vienna.

3. Take it for a test drive.

Require a demonstration of the system, says Kathryn S. Crisler, MS, RN, senior research associate for the Center for Health Services and Policy Research in Denver, which runs the Outcomes Based Quality Improvement (OBQI) project, including OASIS. "Some vendors will tell you they’ve integrated OASIS, but they’re really not there yet," she says. When you see the system, "check to see that the OASIS items are actually there and that they’re word for word." One indication of that may be the copyright notice.

4. Look for adaptability.

The system should be easy to adapt, Vienna says. OASIS, which is already in its second official version, will be tweaked and will grow as time goes on.

5. Look for integration ability.

"You need a system that will integrate OASIS into your own comprehensive assessment tool," says Vienna. "You don’t want a system that is a stand-alone document; otherwise it will be too duplicative." It may also work for you if the vendor has integrated it into its own assessment tool.

6. Look for basic edit checks.

Crisler says the Center for Health Services and Policy Research has encouraged all vendors to include edit checks. These are checks that help prevent errors. For example, they may ensure that all necessary fields are filled in and would flag logical inconsistencies. This would include assessments that say a patient is "comatose," but that note somewhere else that the patient is "fully independent in ambulation." The edits may also prevent "out of range" answers, such as entering a three when the choices are 1. Male, and 2. Female.

7. Don’t take the lazy way out.

Don’t pick a system that allows staff to carry over information from one assessment to another, only changing the fields for which they have new information, both Crisler and Vienna say. Such practices may save staff time, but they will wreak havoc on your outcomes and increase the potential for inaccuracy because old information is allowed to accidentally pass through again unchanged. As a result, your outcomes will appear flat. Agencies with this kind of system in the demonstration project "didn’t look like they were doing as good a job as they actually were because they would carry over all this information, and it looked like their patients weren’t improving," Vienna says.

8. Look for skip patterns.

"If a patient doesn’t have a wound, it should skip over that whole piece," Crisler says.

9. Ensure appropriate items are on appropriate assessments.

For example, date of birth and gender need to be on the initial assessment only. Likewise, you will need to note certain items at discharge that will not need to be on the previous assessments of the patients, Crisler says.

[Editor’s note: These guidelines and more are in the OASIS Basics ($68) educational package from the Center for Health Services and Policy Research. Call (303) 756-8350, or fax to (303) 759-8196.]