Measuring Quality in Assisted Living: The Oregon Way

Background Assisted living (AL) is the largest residential long‐term care provider for dementia specific services in the United States (US), but no nationally accepted standards exist for measuring AL quality in the US Oregon's legislatively mandated Quality Measurement Program (QMP) requires all licensed AL communities to report five quality indicators annually: direct care staff retention, training compliance, falls with injury, prescriptions for nonstandard antipsychotic medication use, and resident satisfaction scores as measured by CoreQ. Methods We linked the 2024 QMP quality measure data to administrative data for all licensed Oregon AL communities with 25 or more beds ( n = 450). We examined whether these indicators differ by organizational characteristics. We used ordinary least squares regression models to examine the associations between these quality metrics and memory care (MC) endorsement, profit status, Medicaid contract, size, and rural location. Results MC communities reported a higher rate of falls with injury (11.6% vs. 7.3%) and double the rate of nonstandard antipsychotic prescriptions (18.6% vs. 9.1%) compared to standalone AL, with no statistically significant differences in staff retention, training compliance, or overall resident satisfaction scores. Nonprofit communities had higher staff retention (+11.5 percentage points) and higher resident satisfaction scores across all CoreQ domains compared to their for‐profit counterparts. Finally, communities with a Medicaid contract reported higher nonstandard antipsychotic prescription rates (+4.7 percentage points) and lower overall CoreQ scores compared to non‐Medicaid communities. Conclusions Performance on some QMP measures differed by MC endorsement, ownership structure, and Medicaid contract. Because QMP does not capture measures of resident acuity, these differences should not be read as differences in overall care quality. Future work should examine whether and how contextual information such as payer‐mix and case‐mix adjustment and ownership data could make public reporting more informative.