Hospital characteristics associated with feeding tube placement in nursing home residents with advanced cognitive impairment.

Abstract

Tube-feeding is of questionable benefit for nursing home residents with advanced dementia. Approximately two-thirds of US nursing home residents who are tube fed had their feeding tube inserted during an acute care hospitalization. To identify US hospital characteristics associated with higher rates of feeding tube insertion in nursing home residents with advanced cognitive impairment. The sample included nursing home residents aged 66 years or older with advanced cognitive impairment admitted to acute care hospitals between 2000 and 2007. Rate of feeding tube placement was based on a 20% sample of all Medicare Claims files and was assessed in hospitals with at least 30 such admissions during the 8-year period. A multivariable model with the unit of the analysis being the hospital admission identified hospital-level factors independently associated with feeding tube insertion rates, including bed size, ownership, urban location, and medical school affiliation. Measures of each hospital's care practices for all patients with serious chronic illnesses were evaluated, including intensive care unit (ICU) use in the last 6 months of life, the use of hospice services, and the ratio of specialist to primary care physicians. Patient-level characteristics were also considered. Endoscopic or surgical insertion of a gastrostomy tube during a hospitalization. In 2797 acute care hospitals with 280,869 admissions among 163,022 nursing home residents with advanced cognitive impairment, the rate of feeding tube insertion varied from 0 to 38.9 per 100 hospitalizations (mean [SD], 6.5 [5.3]; median [interquartile range], 5.3 [2.6-9.3]). The mean rate of feeding tube insertions per 100 admissions was 7.9 in 2000, decreasing to 6.2 in 2007. Higher insertion rates were associated with the following hospital features: for-profit ownership vs government owned (8.5 vs 5.5 insertions per 100 hospitalizations; adjusted odds ratio [AOR], 1.33; 95% confidence interval [CI], 1.21-1.46), larger size (>310 beds vs <101 beds: 8.0 vs 4.3 insertions per 100 hospitalizations; AOR, 1.48; 95% CI, 1.35-1.63), and greater ICU use in the last 6 months of life (highest vs lowest decile: 10.1 vs 2.9 insertions per 100 hospitalizations; AOR, 2.60; 95% CI, 2.20-3.06). These differences persisted after controlling for patient characteristics. Specialist to primary care ratio and hospice use were weakly or not associated with feeding tube placement. Among nursing home residents with advanced cognitive impairment admitted to acute care hospitals, for-profit ownership, larger hospital size, and greater ICU use was associated with increased rates of feeding tube insertion, even after adjusting for patient-level characteristics.