More than home health agencies' reputation may soon ride on their quality measures - cold hard cash could be on the line.
The Medicare Payment Advisory Commission wants to tie home health agency payments to quality measures in the near future, the influential advisory body to Congress indicated in its Sept. 9 meeting.
Despite industry reservations, MedPAC appears likely to recommend some sort of tie between payment and quality for HHAs. The main question is what quality indicators MedPAC will use.
Currently MedPAC is considering two data sets Medicare already uses - outcome-based quality improvement (OBQI) measures, 10 of which are already publicly displayed on Home Health Compare Web site, and outcome-based quality monitoring (OBQM) measures.
MedPAC also is looking at process-oriented measures suggested by Pfizer Inc. and Rand Corp.'s Assessing Care of Vulnerable Elders (ACOVE) project and as-yet-undetermined patient satisfaction surveys. You can't argue with rewarding providers for performance, says Bob Wardwell with the Visiting Nurse Associations of America. But he has advised MedPAC that the home health sector isn't "quite ready" for the financial tie to quality.
MedPAC most likely will end up recommending payment differences based on quality measures Medicare already uses. But there are lots of problems with those OASIS-based measures, notes consultant Pam Warmack with Clinic Connections in Ruston, LA.
Chief amongst those problems is the risk adjuster, MedPAC Commissioner Carol Raphael said in the meeting. The adjuster doesn't fully compensate for outcome differences in different types of patients, argued Raphael, CEO of the Visiting Nurse System of New York.
For example, HHAs treating psychiatric patients have a large number of patients who are confused, Warmack says. And by the nature of their condition, those psych patients "will have very little chance of experiencing a decrease in the degree of confusion or the frequency of the confusion," she complains.
The same goes for chronic care patients who have little chance of improving their functional outcomes, Raphael pointed out in the meeting.
Warmack fears HHAs would shun patient populations known to drag down outcomes if the Centers for Medicare & Medicaid Services ties reimbursement to quality measures.
Another problem is that clinicians continue to inaccurately fill out OASIS, which is the basis of the OBQI and OBQM measures, Warmack notes.
Studies have shown inter-rater reliability (agreement between different clinicians filling out an assessment) to be between 60 and 80 percent, and congruence (agreement with oneself) to be around 60 percent, a MedPAC staffer said. In the quality measure world, those are considered good figures, she claimed.
CMS' ongoing tweaking of the OASIS tool adds to the confusion, MedPAC acknowledged.
Often the variables affecting patient outcomes are beyond an agency's control, Raphael noted. In addition to patient characteristics, one example is rehospitalization. A study VNS-NY conducted showed one local hospital had a high rate of rehospitalizing patients while another had a low rate. "To what degree can we control rehospitalization, or does it have to do with patterns in the hospitals themselves?" Raphael asked.
Despite these concerns, MedPAC indicates it will forge ahead with some form of the recommendation. Even if OASIS and the related measures aren't perfect, they are probably good enough to start phasing in some sort of quality tie to payment, commissioners indicated in the meeting. A final recommendation on the matter could be included in MedPAC's report to Congress next spring.
If MedPAC makes the recommendation, "CMS will be hard-pressed not to adopt it," Wardwell predicts.
Agencies can get help: If payment does end up tied to OBQI or OBQM measures, at least agencies can obtain help in improving these measures from their state Quality Improvement Organizations. The new three-year contract for QIOs, which would start in August 2005, calls for the organizations to furnish even more help to HHAs seeking to improve their OBQI scores.