PREDICTIVE FACTORS FOR EARLY MORTALITY AFTER PERCUTANEOUS ENDOSCOPIC GASTROSTOMY

Background: Percutaneous endoscopic gastrostomy (PEG) is a safe access procedure for enteral nutrition. The purpose of this investigation is to identify predictive factors for early mortality after PEG. Methods : A retrospective review of the hospital records of 416 patients under going PEG from June 1, 1989, through December 31, 1991, was conducted. Patient demographics, admitting diagnosis, indication for PEG, risk f actors for early mortality, and cause and date of death were reviewed. Logistic regression analysis was used to develop a model to predict e arly mortality after PEG. The follow-up period ranged from 1 to 30 mon ths. Results: The overall mortality rate in this review was 227 of 416 patients (54.6%). The 7- and 30-day case fatality rates were 39 of 41 6 (9.4%) and 97 of 416 (23.3%), respectively. Logistic regression anal ysis showed that urinary tract infection (odds ratio (OR) = 3.05; 95% confidence interval (CI) = 1.45 - 6.43) and previous aspiration (OR = 6.86; 95% CI = 3.27 - 14.4) were predictive factors for death at 1 wee k after PEG. Patients who had both risk factors had a 48.4% probabilit y of dying within 7 days after PEG insertion, whereas those who had no risk factors had a 4.3% probability of death. Urinary tract infection (OR = 2.00; 95% CI = 1.17 - 3.41), previous aspiration (OR = 3.62; 95 % CI = 2.00 - 6.55), and age greater than 75 years (OR = 2.49; 95% CI = 1.47 - 4.21) were predictive factors for death at 1 month after PEG. Patients who had all three risk factors had a 67.1% probability of de ath at 1 month while those who had no risk factors had a 10% probabili ty of death. Conclusions: A subgroup of patients exists that has a ver y high mortality rate after PEG. Less invasive ways of nutritionally s upporting these high-risk patients should be evaluated.