Critically ill patients are consistently underfed. Feeding protocols are standardized system tools used to guide nutrition practices, but to date have failed to improve delivery of nutrition. The PEP uP Protocol is a new enhanced feeding protocol. Twenty North American Intensive Care Units (ICUs) will assess baseline nutrition practices. Ten ICUs will be randomized to implement the PEP uP Protocol and educational intervention, and ten will be randomized to continue usual care. Nutrition practices will be reevaluated 6 months after baseline. The investigators hypothesize that the PEP uP Protocol will increase delivery of nutrition, and may ultimately lead to improved survival of critical illness.

There is a well known and well described relationship between malnutrition, immune dysfunction, and infection. Critically ill patients are often hypermetabolic and can rapidly become nutritionally compromised. Repeated efforts over the past few years have not significantly improved the amount of calories delivered via the enteral route. Historically, feeding protocols have been used to guide the delivery of enteral nutrition (EN) but they frequently utilize conservative, reactionary approaches to optimizing nutrition.

We propose a new, innovative approach that protocolizes an aggressive set of strategies to providing EN and shifts the paradigm from reactionary to proactive followed by de-escalation if nutrition therapy is not needed. The key components of this new protocol are the following: 1) Starting feeds at the target rate based on increasing evidence that some patients tolerate starting nutrition at a higher rate of delivery and that slow start ups are not necessary. For patients who are hemodynamically stable, we propose to shift from an hourly rate target goal to a 24 hour volume goal and give nurses guidance on how to make up this volume if there was an interruption for non-gastrointestinal reasons. This 'volume-based' goal represents a significant shift in practice from traditional fixed hourly rate goals to a new protocol in which nurses can increase the hourly rate depending on how many hours they have left in the day to ensure that the patient receives the 24 hour volume within the day. 2) For patients who are deemed unsuitable for high volume intragastric feeds, we provide an option to initiate 'trophic feeds.' Trophic feeds represent an idea to provide a low volume of a concentrated feeding solution for 24 hours or longer, designed to maintain gastrointestinal structure and function rather than meet their protein and caloric goals. This option should reduce the numbers of patients ordered to be kept nil per os (NPO). Thus, PEP uP patients may gain some of the benefit of early EN. 3) To optimize tolerance in the early phase of critical illness, we propose to use a semi elemental feeding solution instead of a standard polymeric solution. There is some evidence that these semi elemental solutions are better assimilated than polymeric solutions in the critical care setting. These solutions can be changed to a more traditional polymeric solution once the patient is tolerating adequate amounts of nutrition. 4) Rather than wait for a protein debt to accumulate because of inadequate delivery of EN, protein supplements are prescribed at initiation of EN and can be discontinued if EN is well tolerated and they are receiving all their protein requirements through their standard EN. This strategy guarantees that the patient will most likely receive all their protein requirements in the early phase of their critical illness. 5) Rather than wait for a problem with gastrointestinal tolerance to develop, we propose to start motility agents at the same time EN is started with a re-evaluation in the days following to see if it is necessary. By preventing delayed gastric emptying, which frequently occurs in this patient population, we can improve nutritional adequacy. 6) Based on emerging evidence that a higher gastric residual volume (the volume of feeds remaining in the stomach when the bedside nurse aspirates the feeding tube) is safe and perhaps results in greater nutritional adequacy, we will include a higher gastric residual volume of 300ml in our protocol. It has been shown in one randomized trial that a feeding protocol that starts a motility agent empirically at the time of initiation of feeds and uses a higher threshold for a critical gastric residual volume (250 ml) improves nutritional adequacy.

Since the bedside nurses initiate and utilize feeding protocols to achieve target goals, we will couple this newer generational feeding protocol with a comprehensive nurse-directed nutritional educational intervention that will focus on its safe and effective implementation. This focus on nursing nutrition education represents a major shift away from traditional education which has focused on dietitians and physicians.

Our hypothesis is that this aggressive feeding protocol combined with a nurse-directed nutrition educational intervention will be safe, acceptable, and effectively increase protein and energy delivery to critically ill patients. We postulate that this increased provision of calories and protein may translate into improved clinical outcomes, particularly for the patients at the extremes of weight, but the current study is not powered to demonstrate such a difference.

Inclusion Criteria:

• Critically ill adults (i.e. >18 years of age)
• Mechanically ventilated before or within the first 6 hours of admission to ICU

Exclusion Criteria:

• Nutrition (either EN or Parenteral Nutrition (PN)) started before admission to ICU
• Not intubated within 6 hours of admission to ICU
• Receiving non-invasive ventilation (i.e. mask ventilation) during the first 6 hours of ICU stay
• Moribund (as evidenced by death within 48 hours of admission to ICU)