|
BACKGROUND/RATIONALE:
VA is committed to providing high quality care in its nursing home care units (NHCUs). A number of quality indicators are routinely used to measure NHCU performance. However, to answer the important question "how well is this nursing home performing," it is very useful to have a summary or composite measure that aggregates the individual measures. Research suggests that organizational performance is related to employee satisfaction and assessment of organizational functioning. The VHA periodically administers the All Employee Survey (AES) to assess job satisfaction, workgroup functioning and organizational culture.
OBJECTIVE(S):
Objective 1: To measure the performance of NHCUs using two composite measures of performance, one based on quality indicators and one also considering costs, using two approaches for determining the composite measures.
Objective 2: To analyze the extent to which NHCU performance is related to job satisfaction, workgroup functioning and organizational culture as assessed on the AES by employees in NHCUs.
METHODS:
Objective 1: The two approaches we will use for calculating the composite measures are Bayesian hierarchical latent variable models (BHMs) and Data Envelopment Analysis (DEA). "Good" composite measures require "good" individual measures. In our earlier work, risk-adjustment models using variables from the Minimum Data Set (MDS) were developed and validated for four quality indicators. As part of this grant, we will use MDS variables to develop and validate risk-adjustment models for four more indicators. We will use Bayesian hierarchical models to estimate risk-adjusted performance on the eight indicators, which will provide a useful standard for assessing the composite measures. Indicator-specific and composite measures of performance will be estimated using 2003 to 2007 MDS data.
Objective 2: The AES was administered in 2004 and 2006. We will use a Bayesian hierarchical model to1) analyze the relationship between AES responses in each year and performance in that year; and 2) analyze the extent to which changes in AES scales from 2004 to 2006 are related to changes in NHCU performance over this period.
FINDINGS/RESULTS:
While awaiting funding approval, we wrote several papers in which we developed aspects of the methodology we will be using in the grant. One paper (published in Medical Care) evaluates two alternative Bayesian latent variable models; one paper (under review) illustrates how DEA can be used to calculate composite measures of quality; and an essay (forthcoming) discusses reflective versus formative scales in the context of composite measures of quality.
IMPACT:
The development and validation of MDS-based risk-adjustment models for more quality indicators and a better understanding of how to most usefully aggregate risk-adjusted quality indicators will allow the VA to better measure nursing home performance, a cornerstone of performance improvement. An increased understanding of organizational and managerial characteristics associated with high performing nursing homes will help identify managerial practices and interventions that could lead to improvements in quality of care.
PUBLICATIONS:
None at this time.
DRA:
Aging and Age-Related Changes, Health Services and Systems
DRE:
Quality of Care
Keywords:
Management, Organizational issues, Quality assessment
MeSH Terms:
none
|
