Tube feeding in nursing home residents with severe and irreversible cognitive impairment.

OBJECTIVES: This study was undertaken to determine the degree to which selected demographic and clinical variables are associated with the use of feeding tubes in older nursing home residents with very severe and irreversible cognitive impairment.

DESIGN: Descriptive, cross-sectional, population-based study.

SETTING: Kansas nursing homes from January 1, 1994, through June 30, 1998.

PARTICIPANTS: A total of 4,997 nursing home residents with very severe and irreversible cognitive impairment comprised the study population. Subjects were identified using data in Minimum Data Set (MDS) reports. Those who were included in the study population were over the age of 65, had two consecutive Cognitive Performance Scale scores of 6, were without evidence of significant subsequent improvement, and were not comatose.

MEASUREMENTS: The MDS data on each resident were examined to determine subjects' age, ethnicity, gender, Medicaid eligibility, disease diagnoses, physical functioning, and oral/nutritional status, including feeding tube status. The MDS was also used to determine urban versus rural location of the nursing home. The association between the use of feeding tubes and selected demographic and clinical characteristics was then examined using bivariate and logistic regression tests.

RESULTS: Of the 4,997 residents in the study population, 577 (11.6%) had feeding tubes. In multivariate analysis, feeding tube use was found to be associated with swallowing problems (odds ratio (OR) 5.4, 95% confidence interval (CI) 4.3-6.8); urban location of nursing home (OR 2.9, 95% CI 2.3-3.5); non-white race (OR 2.7, 95% CI 1.9-3.6); stroke (OR 2.5, 95% CI 2.0-3.1); and absence of dementia (OR 2.5, 95% CI 2.0-3.1). Feeding tubes were also more weakly associated with age <86 years, male gender, dependency for all activities of daily living, and absence of a living will. Feeding tube use was not found to be associated with chewing problems, Medicaid status, or resuscitation status.

CONCLUSIONS: Clinical variables including swallowing problems, stroke, and absence of dementia were strongly associated with the use of feeding tubes in this cross-sectional, population-based study. In addition, urban location of nursing home and non-white race were significantly associated with feeding tubes. These findings suggest that feeding tube decisions are strongly influenced by nonclinical factors and invite further investigation.