Medical care for nursing home residents: Differences by dementia status

OBJECTIVE: To understand the use of medical services by nursing home reside nts. DESIGN: Descriptive, longitudinal study comparing medical service use of re sidents by dementia status and describing the use of medical services follo wing detection of fever or infection. SETTING: Fifty-nine randomly selected nursing homes in Maryland from 1992 t o 1995. PARTICIPANTS: 2153 residents admitted to one of 59 randomly selected nursin g homes. MEASUREMENT: A panel of psychiatrists and neurologists ascertained dementia based on review of medical records, interview data with significant others and nursing staff, and results of a cognitive exam. Medical service use wa s abstracted from medical records. BACKGROUND: Understanding the use of medical services by nursing home resid ents as distinct from services provided by the nursing home is important, p articularly as new medical care models are tested. This study compares the medical service use of residents by dementia status and describes the use o f medical services following detection of fever or infection. RESULTS: Residents with dementia compared with those without dementia had l ower annual rates of physician visits (10.2 vs 12.7, P < .001) and hospital izations (0.9 vs 1.2, P < .001), virtually the same rate of emergency depar tment visits, and similar lengths of stay in the hospital. Subsequent to in fection, a lower proportion of residents with dementia had either a physici an visit, an emergency department visit, or a hospital admission compared w ith residents without dementia (27.2% vs 32.2%, P < .001). In 87% of infect ions, an antibiotic was used, implying meaningful contact with a physician. Residents with dementia compared with those without dementia had fewer phy sician visits subsequent to fevers (20.6% vs 29.9%, P < .001) and infection s (21.8% vs 27.5%, P < .001). CONCLUSIONS: The association of less medical service use by individuals wit h dementia compared with those without dementia may reflect differences in health status or implicit end-of-life decision-making and a proclivity towa rd less-aggressive treatment for these individuals.