OASIS Alert

Published on Tue Sep 30, 2003 PDF

Your data will be up for comparison any day now.

No matter how misleading you think the government's home health outcome results are, the world will see your agency's scores soon.

The national rollout of the public patient outcome comparison will look much the same as the eight-state pilot program that launched in May. The home health quality initiative will post the same 11 patient outcomes on the Centers for Medicare & Medicaid Services Home Health Compare Web site (see "OBQI:  Your Care is Reduced to 11 Outcomes For Public Comparision").

And three of those outcomes will be highlighted in newspaper advertisements across the country when the program launches. CMS plans to run at least one ad per state featuring 40 to 50 HHAs' outcomes, reports the National Association for Home Care & Hospice.

CMS says it will send a preview of the HHQI data to agencies' QIES mailboxes between Oct. 1 and Oct. 21, when the initiative is scheduled to begin.

The government had planned to provide information about the redesign and changes made to Home Health Compare at a Sept. 18 Open Door Forum, but the meeting was cancelled because of the government shutdown for Hurricane Isabel. It has been rescheduled for Oct. 7 at 2 p.m. ET, a CMS spokesperson tells Eli.

CMS also has scheduled an explanation of the details of the HHQI launch in an Oct. 3 satellite broadcast from 1 p.m. to 3:30 p.m. ET. HHAs can sign up for the broadcast over the Web at http://cms.internetstreaming.com, or view the session at the same site for a one-year period after Oct. 3.

New Search by ZIP Code

The Home Health Compare Web site, available at www.medicare.gov, will include one significant Phase II change, NAHC notes. Visitors will be able to search by the number of beneficiaries served by an agency in a certain zip code. CMS added this new search function after Phase I participants complained that numerous agencies that didn't actually serve the searcher's geographic area came up in the search.

The new search option will display agencies with the highest number of beneficiaries served in the zip code at the top and the lowest at the bottom, explains NAHC's Mary St. Pierre.

"It won't eliminate the problem" of wrong agencies showing up in a search, but it will help identify those that really do cover the area by showing whether they have a substantial number of beneficiaries served in the zip code, St. Pierre expects.

Once your patient outcomes go out in newspaper ads and on the Web, patients, referral sources, the press and many other groups will be perusing them. And what they see doesn't always correlate with the care they get, experts worry.

"Risk-adjustment plays an important part in how your agency fares in the outcome measure," says Brian Ellsworth with the Connecticut Home Care Association. And many agencies have questions about the risk-adjustment calculations.

In Connecticut, for example, when looking at emergent care and hospitalization, agencies don't think the model adequately accounts for the proportion of dually eligible long-stay Medicaid clients with chronic conditions, he tells Eli, leaving some agencies penalized for the types of patients they serve.

New York has similar concerns, agrees Carol Rodat with the Home Care Association of New York State. The Empire State has a popular program that keeps beneficiaries out of nursing homes with home care. Those patients have chronic, long-term conditions and generally aren't expected to improve, Rodat points out.

Under the HHQI program, New York HHAs that serve those types of patients have lower outcomes than other agencies, and thus look like they are providing a lower quality of care, Rodat tells Eli. The risk-adjustment CMS uses in the calculations doesn't compensate for this population, she maintains.

About 40 percent of New York agencies listed on Home Health Compare serve only this long-term, chronic population, and their outcomes look poor compared to the agencies that serve a mix of traditional Medicare patients and long-term patients, Rodat reports. "It skews the outcomes," she says.

Rodat expects other states with similar programs for long-term care will see similar results.

Inconsistent Data Collection

Another problem with HHQI is that the data simply isn't reliable enough to be used for public reporting, Rodat argues. Clinicians are still very inconsistent in how they collect OASIS data, and interrater reliability between assessments is low, she claims. That's the case even in New York, where the industry has been using OASIS for seven years (since it was part of the OASIS demonstration project).

Finally, because the patient is in the home and not in an institutional setting, HHAs have very little control over some of the outcomes measured by HHQI, Rodat says. For example, the management of oral medications outcome can be influenced by how physicians prescribe medications, whether the patient takes old meds without permission, if the patient takes someone else's meds, etc., she notes - all conditions largely out of an HHA's hands.

More examples are the acute care hospitalization and emergent care outcomes, points out Bob Wardwell with the Visiting Nurse Associations of America. "They pick up lots of coding errors and situations that are not reflective of agency care problems," Wardwell says.

"This data should never have been used for public reporting," Rodat declares. "It is not appropriate for this purpose."

"OBQI was really designed to be a prelude to more careful and specific internal and external review, rather than a de facto scoring system," Wardwell agrees. "Items like this are a good fit for the former, but not for the latter."

Despite its flaws, Home Health Compare is set to debut this month. Now the burden is on agencies to ensure that patients understand what the chosen outcomes reveal about the care they provide, Wardwell says.