Nursing Home Facility Risk Factors for Infection and Hospitalization: Importance of Registered Nurse Turnover, Administration, and Social Factors


Address correspondence to Sheryl Zimmerman, PhD, Cecil G. Sheps Center for Health Services Research, The University of North Carolina at Chapel Hill, 725 Airport Road, Campus Box 7590, Chapel Hill, NC 27599. E-mail:


OBJECTIVES: Determine the relationship between a broad array of structure and process elements of nursing home care and (a) resident infection and (b) hospitalization for infection.

DESIGN: Baseline data were collected from September 1992 through March 1995, and residents were followed for 2 years; facility data were collected at the midpoint of follow-up.

SETTING: A stratified random sample of 59 nursing homes across Maryland.

PARTICIPANTS: Two thousand fifteen new admissions aged 65 and older.

MEASUREMENTS: Facility-level data were collected from interviews with facility administrators, directors of nursing, and activity directors; record abstraction; and direct observation. Main outcome measures included infection (written diagnosis, a course of antibiotic therapy, or radiographic confirmation of pneumonia) and hospitalization for infection (indicated on medical records).

RESULTS: The 2-year rate of infection was 1.20 episodes per 100 resident days, and the hospitalization rate for infection was 0.17 admissions per 100 resident days. Except for registered nurse (RN) turnover, which related to both infection and hospitalization, different variables related to each outcome. High rates of incident infection were associated with more Medicare recipients, high levels of physical/occupational therapist staffing, high licensed practical nurse staffing, low nurses' aide staffing, high intensity of medical and therapeutic services, dementia training, staff privacy, and low levels of psychotropic medication use. High rates of hospitalization for infection were associated with for-profit ownership, chain affiliation, poor environmental quality, lack of resident privacy, lack of administrative emphasis on staff satisfaction, and low family/friend visitation rates. Adjustment for resident sex, age, race, education, marital status, number of morbid diagnoses, functional status, and Resource Utilization Group, Version III score did not alter the relationship between the structure and process of care and outcomes.

CONCLUSIONS: The association between RN turnover and both outcomes underscores the relationship between nursing leadership and quality of care in these settings. The relationship between hospitalization for infection and for-profit ownership and chain affiliation could reflect policies not to treat acute illnesses in house. The link between social factors of care (environmental quality, prioritizing staff satisfaction, resident privacy, and facility visitation) and hospitalization indicates that a nonmedical model of care may not jeopardize, and may in fact benefit, health-related outcomes. All of these facility characteristics may be modifiable, may affect healthcare costs, and may hold promise for other, less-medical, forms of residential long-term care.