Goal: To identify factors influencing survival after percutaneous endoscopic gastrostomy (PEG) tube insertion in elderly patients with severe dementia.
Background: Use of PEG in patients with dementia raises medical and ethical questions.
Study: The cohort included 189 patients aged ≥64 years with severe dementia who underwent PEG tube insertion in 2002 to 2011 at a tertiary hospital and were followed through 2014. Data were collected on background diseases, laboratory tests conducted 14 (±10) days before and 90 (±10) days after PEG, and date of death. Multivariate Cox regression analysis was performed and cumulative survival curves plotted.
Results: Mean baseline serum albumin level was significantly lower in patients who died ≤30 days after PEG (n=32, 16.9%) than patients who survived longer (2.9±0.5 vs. 3.3±0.5 g/dL, P<0.001), and significantly higher in patients who survived 1 year after PEG (n=96, 50.8%) than patients who died earlier (3.4±0.5 vs. 3.1±0.5 g/dL, P=0.002). No significant differences were found for the other laboratory parameters. After adjustment for background diseases, serum albumin level above the median (3.3 g/dL) was the only predictor of survival (hazard ratio 0.51; 95% confidence interval, 0.37-0.72; P<0.001). Patients with stable/increased serum albumin levels at 90 days after PEG (n=80, 42.3%) survived longer than patients with decreased levels. The only predictor of survival after adjustment for age and background diseases was a stable/increased albumin level at 90 days (hazard ratio 0.59; 95% confidence interval, 0.42-0.85; P=0.004).
Conclusions: Survival after PEG tube insertion is associated with a higher serum albumin level at baseline and a stable/increased serum albumin level during follow-up.