Survival after percutaneous endoscopic gastrostomy placement in older persons

Background. The prolongation of life is an important consideration in the d ecision to initiate long-term tube feeding. This report critically synthesi zes the evidence regarding the impact of percutaneous endoscopic gastrostom y (PEG) tube placement on survival in older persons. Methods. A systematic search was conducted using MEDLINE from January 1980 until January 1999. Articles reporting survival data in older persons (mean or median age >65 years) after PEG tube placement were identified. The num ber and age of subjects, length of follow-up, setting, and survival data we re extracted from all eligible studies. Mortality data at 1, 2, 6, and 12 m onths after PEG placement were quantitatively synthesized. Clinical charact eristics associated with decreased survival among subjects with PEG tubes w ere identified. Results. Five cohort studies compared survival in patients with and without feeding tubes in nursing homes, but none demonstrated a survival benefit. Another cohort study reported increased survival for tube fed patients with amyotrophic lateral sclerosis. The pooled proportion of all subjects survi ving after PEG placement was as follows: 1 month = 0.81 (95% confidence int erval [CI], 0.74-0.88), 2 months = 0.70 (95% CI, 0.65-0.74), 6 months = 0.5 6 (95% Ct, 0.200.92), and 12 months = 0.38 (95% CI, 0.26-0.49). Advanced ag e and malignancy were the factors most often reported to be associated with poorer survival among subjects with PEG tubes. Conclusions. The impact of PEG placement on survival is not known because t he level of evidence is limited. PEG tubes may prolong life in selected pop ulations. However, the majority of older patients selected for PEG placemen t will not survive 1 year after the procedure. Certain factors may identify those patients more likely to derive a survival benefit From long-term tub e feeding. This information may offer some guidance to decision makers for whom prolongation of life is an important factor in the tube-feeding decisi on.