Grabowski D, Cawley J, Hirth R; AcademyHealth. Meeting (2003 : Nashville, Tenn.).
Abstr AcademyHealth Meet. 2003; 20: abstract no. 789.
University of Alabama-Birmingham, Health Care Organization & Policy, RPHB 330, 1665 University Boulevard, Birmingham, AL 35294 Tel. (205) 975-8967 Fax (205) 934-3347
RESEARCH OBJECTIVE: Previous economic studies have been unable to conclusively identify the causes of variation in the type and quality of care in nursing homes. We offer a novel hypothesis: that variation in the methods of care and quality of care across nursing homes and over time is related to local variation in the cost of nursing home labor. Based on a microeconomic model of factor substitution, we predict that a rise in the wages of nursing home workers will be associated with 1) less labor used by nursing homes; 2) care increasingly provided in materials-intensive (i.e. labor-saving) ways; and 3) a rise in indicators of low-quality care. STUDY DESIGN: We examined three outcomes of interest. 1) The quantity of labor used by the nursing home, which is expressed as the number of Nurse Aide hours used per resident day, and the number of professional nurse (RN and LPN) hours used per resident day. We estimate these models using ordinary least squares regression. 2) The extent of materials-intensive methods of care. Outcomes of interest in this second category include: percent of residents on a feeding tube, percent of residents on a catheter, percent of residents on psychoactive drugs, and percent of residents under physical restraints. 3) Quality of care, as measured by the percent of residents with bedsores, the number of health-related deficiencies cited by state inspectors, and the total number of deficiencies cited by state inspectors. For outcomes that are expressed as a percentage of residents, we use the logit transformation of the dependent variable and ordinary least squares regression. For outcomes that are counts of deficiencies, we estimate a negative binomial regression model. POPULATION STUDIED: We test our predictions using longitudinal data on U.S. nursing homes, specifically, the linked annual data of the federal On-line Survey Certification and Reporting (OSCAR) system from 1991 through 1998 for the 48 contiguous states. Between 1991 and 1998, there were 125,791 surveys of nursing homes within the 48 contiguous states. PRINCIPAL FINDINGS: The results of our analyses confirm our prediction that higher wages result in nursing homes using fewer hours of labor (both by Nurse Aides and by professional nurses). We also find that a one-standard-deviation change in wages is associated with the average nursing home raising its percent of residents on psychoactive drugs by 7%. We find little effect of wages on the extent to which other forms of care are provided in materials-intensive ways. We also find little evidence that higher nursing home wages result in worse quality of care, as measured by pressure sores or deficiencies. CONCLUSIONS: If policymakers wish to address the problem of low quality nursing home care, they must first identify its causes. We find evidence that nursing home staffing and the use of psychoactive drugs in nursing homes are related to local labor market conditions. IMPLICATIONS FOR POLICY, DELIVERY OR PRACTICE: Policymakers may need to give further consideration to minimum staffing requirements within nursing facilities. Such restrictions could help to insulate the quality of patient care from local labor market fluctuations. In conjunction with minimum staffing requirements, state policymakers could also index Medicaid payments to nursing facilities based on area differences in wages for nursing home workers, in a manner similar to physician payment under the Medicare program for acute-care services.
Publication Types:
Keywords:
- Humans
- Medicare
- Nurses' Aides
- Nursing Homes
- Nursing Staff
- Salaries and Fringe Benefits
- United States
- economics
- hsrmtgs
Other ID:
UI: 102275757
From Meeting Abstracts