Should Percutaneous Endoscopic Gastrostomy (PEG) Be Used in Intensive Care Unit Patients?

Abstract

Purpose: With rising health care costs, a current trend is to use PEG to shorten hospital length of stay (LOS) and to expedite discharge to the nursing home (NH). Because of this trend, gastroenterologists increasingly are being asked to perform PEGs in acutely ill patients with multiple comorbidities. To determine mortality in patients from medical intensive care units (MICU) versus the medical floors on whom PEG consults had been requested? Methods: A sample of 191 consecutive patients [mean age of 81 (range 49–105 yrs), 58% female and 81% from NH] referred for PEG placement were included. Patient age, gender, BMI, race, community vs. NH, new PEG vs. replacement, reason for PEG, MICU vs. floor, LOS, clinical lab values, comorbidity, nutritional status classification score and outcome (discharge vs. death) were collected. The data was examined using Student t-tests, chi-square analysis and logistic regression analysis. Results: MICU and medical floor patients had similar age (P = .0736) and gender distribution (P = .278). Significantly more MICU patients came from the community vs. NH (41% vs. 15%, P = .002). Also, a significantly higher proportion of MICU patients died compared to medical floor patients (59% vs. 18%, P < .0005). Deaths were similar in new PEG and replacement PEG patients (P = .859). MICU patients had significantly longer LOS (mean of 56 vs. 18 days, P = .0240) but similar mean BMI (P = .5123) and nutritional risk assessment scores (P = .457) when compared with medical floor patients. Logistic regression analysis determined that deaths were 93% more likely to occur in MICU patients (P < .0005), independent of the 10% decreased risk of death for every 1-unit increase in BMI (>11), and independent of the >15 million-fold likelihood of death in patients with moderately compromised (P < .0005) or severely compromised (P < .0005) nutritional status classification. Further, age, gender, race, place of residence (community vs. NH) and LOS were not risk factors for death in this sample of PEG consult patients. Conclusions: MICU patients are more severely ill than patients on the floors and may not survive long enough to benefit from PEG. It may be more reasonable to use nasogastric tubes for feeding and allow a “cool off” period between getting a consult for PEG and actually inserting a PEG.Future studies should work to identify both positive and negative factors associated with death in MICU patients to aid gastroenterologists in their decisions regarding PEG insertions.