Clinical Investigations Low caloric intake is associated with nosocomial bloodstream infections in patients in the medical intensive care unit*

T here are many theoretical advantages to providing nutritional support to critically ill patients, such as improved respiratory muscle strength, ventilatory response to hypoxemia and hypercapnia, cardiac function, and wound healing (1– 7). These potential benefits have led a number of organizations to publish consensus statements, based on expert opinion and limited data from surgical, burn, and trauma patients, promoting aggressive nutritional supplementation (8–11). The paucity of data from rigorous studies evaluating the benefits and risks of nutritional supplementation in critically ill medical patients, however, continues to promote controversy regarding the importance of feeding during the relatively limited duration of an intensive care unit (ICU) stay (11, 12). The effect of caloric intake on infection has been a central issue in the debate about whether to feed critically ill patients. Malnutrition has been associated with poor immune function, which increases the risk of systemic infection (13). On the other hand, continuous enteric feeding has been associated with an increased risk of ventilator-associated pneumonia, presumably due to an increased risk of aspiration (14). Moreover, total parenteral nutrition (TPN) compared with enteral feeding has been associated with greater risk of infection (15–17). However, the overall impact of feeding on the risk of infection for critically ill medical patients remains uncertain (18). Nosocomial bloodstream infections (BSIs) are a major source of morbidity and mortality for hospitalized patients and disproportionately affect patients in ICUs (19–22). The increasing number of ICU beds in the United States (23), coupled with the increasing prevalence of Objective: To determine whether caloric intake is associated with risk of nosocomial bloodstream infection in critically ill medical patients. Design: Prospective cohort study. Setting: Urban, academic medical intensive care unit. Patients: Patients were 138 adult patients who did not take food by mouth for >96 hrs after medical intensive care unit admission. Measurements: Daily caloric intake was recorded for each patient. Participants subsequently were grouped into one of four categories of caloric intake: <25%, 25–49%, 50–74%, and >75% of average daily recommended calories based on the American College of Chest Physicians guidelines. Simplified Acute Physiology Score II and serum albumin were measured on medical intensive care unit admission. Serum glucose (average value and maximum value each day) and route of feeding (enteral, parenteral, or both) were collected daily. Nosocomial bloodstream infections were identified by infection control surveillance methods. Main Results: The overall mean ( SD) daily caloric intake for all study participants was 49.4 29.3% of American College of Chest Physicians guidelines. Nosocomial bloodstream infection occurred in 31 (22.4%) participants. Bivariate Cox analysis revealed that receiving >25% of recommended calories compared with <25% was associated with significantly lower risk of bloodstream infection (relative hazard, 0.24; 95% confidence interval, 0.10–0.60). Simplified Acute Physiology Score II also was associated with risk of nosocomial bloodstream infection (relative hazard, 1.27; 95% confidence interval, 1.01–1.60). Average daily serum glucose, admission serum albumin, time to initiating nutritional support, and route of nutrition did not affect risk of bloodstream infection. After adjustment for Simplified Acute Physiology Score II in a multivariable analysis, receiving >25% of recommended calories was associated with a significantly lower risk of bloodstream infection (relative hazard, 0.27; 95% confidence interval, 0.11–0.68). Conclusions: In the context of reducing risk of nosocomial bloodstream infections, failing to provide >25% of the recommended calories may be harmful. Higher caloric goals may be necessary to achieve other clinically important outcomes. (Crit Care Med 2004; 32:350–357)