Cognitively Impaired Nursing Home Residents with Feeding Tubes: Association with End-of-Life Care Process Measures

Evidence suggests that feeding tubes (PEGs) do not offer physical benefits to severely cognitively impaired nursing home (NH) residents, and may actually cause harm. Previous work demonstrated important associations between PEG use and either 1) structural or organizational NH characteristics, or 2) physician and patient factors. This study contributes by directly examining NH care process measures in New York State facilities to elucidate associations between care delivery and NH quality.

Data from several sources were used. The dependent variable- proportion of residents with severe cognitive impairment who used PEGs- came from the Minimum Data Set (MDS). NH end-of-life (EOL) care process measures were derived from a survey conducted in New York State facilities (n=313, response rate 52%). In a prior study, these EOL process measures were shown to be psychometrically reliable and valid. The EOL care processes were assessed using 49 items measured on a 5-point Likert scale. Facility characteristics previously shown to be associated with the dependent variable were obtained through OSCAR data. Bivariate analysis and multivariate logistic regression techniques were utilized.

Findings suggest several important associations between survey responses and the proportion of residents on PEGs. NHs that are more likely to re-evaluate EOL treatment options as residents' conditions change, and those that educate residents and families with regard to treatment options, are less likely to employ PEGs (p=0.01 and p=0.08, respectively).

Modifiable process measures associated with PEG use were identified. These findings could serve to focus future NH research and quality improvement efforts.