Correspondence between OASIS home health assessment data and administrative claims data

Objective: Correspondence between the home health Outcome and Assessment Information Set (OASIS) and administrative claims is not well understood.

Methods: The 2004 Chronic Care Warehouse, both the Medicare 5% standard analytic file and OASIS. The study sample included 95,674 fee-for-service Medicare beneficiaries who survived the calendar year, entered and were discharged from home health care between 1/1/2004 and 11/20/2004, and did not incur hospice services. OASIS and linked administrative claims were compared in terms of depression ascertainment and readmission to acute or post-acute care through the first 30 days after discharge from home health. Results: Prevalence of depressed mood was lower by nurse assessment (21.3%) than a standard claims-based algorithm for major depression (23.4%). Observed differences varied by patients' living arrangement, and were most similar for patients living with family (20.6% by nurse assessment versus 21.6% via claims) and least similar for patients living alone (23.7% by nurses versus 28.2% in claims) or with non-relatives (23.9% by nurses versus 33.4% in claims). Among 66,075 patients discharged from home health to the community without claims-based evidence of readmission only 2.5% were indicated as having being readmitted in OASIS. The degree of correspondence was lower for 29,599 home health patients with claims-based evidence of subsequent readmission, where in OASIS 26.3% were indicated as having been discharged to the community.

Conclusions: The chronic care warehouse provides valuable insight on individual factors and organizational processes that may be used to improve quality of care for chronically ill Medicare beneficiaries.