Clinical Research. Keywords enteral nutrition; clinical protocols; critical illness; trauma; nutritional support; critical care

Transcription

1 539705NCPXXX / Nutrition in Clinical PracticeTaylor et al research-article2014 Clinical Research Improving Enteral Delivery Through the Adoption of the Feed Early Enteral Diet Adequately for Maximum Effect (FEED ME) Protocol in a Surgical Trauma ICU: A Quality Improvement Review Nutrition in Clinical Practice Volume XX Number X Month 201X American Society for Parenteral and Enteral Nutrition DOI: / ncp.sagepub.com hosted at online.sagepub.com Beth Taylor, DCN, RD 1,2 ; Rebecca Brody, PhD, RD 2 ; Robert Denmark, PhD 3 ; Robert Southard, MD 4 ; and Laura Byham-Gray, PhD, RD 2 Abstract Background: Despite the research supporting adequate enteral nutrition (EN) in intensive care unit (ICU) patients, underfeeding is still common. This quality improvement (QI) project was done to determine the effect of volume-based feeding on adequacy of EN delivery and provision of calories and protein in a surgical/trauma ICU (STICU). Materials and Methods: Mechanically ventilated STICU patients (n = 111) fed at least 72 hours after achieving their target goal of EN during their first week of admission were reviewed retrospectively in a QI project. Data were obtained before (n = 54) and after (n = 56) initiation of a volume-based feeding protocol (FEED ME Feed Early Enteral Diet adequately for Maximum Effect). Results: The proportion of EN volume and calories delivered increased significantly (rate based, 63% ± 20%; FEED ME, 89% ± 9%; P <.0001), as did grams of protein/kg actual body weight (1.13 ±.29 to 1.26 ±.37; P =.036) using the FEED ME protocol. Groups were similar in patient demographics, clinical characteristics, and nutrition practices. Only slightly more diarrhea (rate based, 0; FEED ME, 6; P =.046) in gastric-fed patients was noted. The incidence of gastric residual volume >350 ml (rate-based, 20 episodes; FEED ME, 11 episodes; P =.34) and emesis (5 vs 2 episodes; P =.22) was similar. Conclusion: A change in standard of practice to an EN volume-based feeding approach in a STICU led to a significant improvement in adequacy of calories and protein delivered, with only a slight increase in diarrhea. (Nutr Clin Pract. XXXX;xx:xx-xx) Keywords enteral nutrition; clinical protocols; critical illness; trauma; nutritional support; critical care According to the Society of Critical Care Medicine (SCCM) and the American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.) clinical practice guidelines (CPGs) for the provision of nutrition therapy in the critically ill patient, enteral nutrition (EN) has the potential to aid in attenuating the metabolic response to stress, preventing oxidative cellular injury, and positively modulating the immune response. 1 In contrast to the medical intensive care unit (MICU), patients managed postoperatively in a surgical/trauma intensive care unit (STICU) are less likely to get EN (77.8% vs 54.6%, P <.05) and more likely to get parenteral nutrition (PN) (13.9% vs 4.4%, P <.05) due to concern for postoperative ileus, aspiration, and anastomotic dehiscence. 2 In addition, a delay in initiation of EN is common, and when started, STICU patients are likely to get fewer goal calories delivered (45.8% vs 56.1%, P <.05) than MICU patients due to holding of feeds for diagnostic testing and multiple operative procedures. 3 Delays and inadequacies of EN delivery have been shown to increase complications and mortality. 2 In the ICU setting, enteral feeding protocols help promote compliance with CPG-directed nutrition practices. 4 A largescale, cluster randomized trial of a novel feeding protocol entitled Protein-Energy Provision via the Enteral Route in Critically Ill Patients (the PEP up protocol) took this a step further with a design to make up for lost EN infusion time due to diagnostic testing and procedures. 5 With this protocol, Heyland et al 6 changed the focus of EN delivery from the traditional rate-based approach (EN prescribed in ml/h) to a system-changing volume-based approach (EN prescribed in ml/d) to meet CPG best practice recommendations. For example, instead of a prescribed rate per hour (60 ml/h), a patient would have a prescribed volume per day (1440 ml/24 hours). This would provide the bedside nurse some freedom in From 1 Department of Food and Nutrition, Barnes-Jewish Hospital, St Louis, Missouri; 2 School of Health Related Professions, Nutritional Sciences, Rutgers University (formerly University of Medicine and Dentistry), Newark, New Jersey; 3 School of Health Related Professions, Interdisciplinary Studies, Rutgers University (formerly University of Medicine and Dentistry), Newark, New Jersey; and 4 Department of Surgery, Washington University School of Medicine, St Louis, Missouri. Financial disclosure: The Dietitians in Nutrition Support (practice group of the Academy of Nutrition and Dietetics) research grant was used to support a data collector and the education of the nursing staff. Corresponding Author: Beth Taylor, DCN, RD, Barnes-Jewish Hospital, Washington University School of Medicine Mailstop , One Barnes-Jewish Hospital Plaza, St Louis, MO 63110, USA. bet1217@bjc.org

2 2 Nutrition in Clinical Practice XX(X) Table 1. Differences Between FEED ME and PEP up Protocols. Characteristic FEED ME PEP up Enteral product Variety (range of 1.0- to 2.0-kcal/mL products) Single semi-elemental, 1.5-kcal/mL product Volume based Yes Yes Start of protocol After patient achieved goal rate of EN As soon as EN started Initiation and advancement SB: Initiate 20 ml/h, increase in 10-mL/h increments every 4 h to goal Gastric: Initiate 100 ml every 4 h, increase SB or gastric: Initiate at goal ml/h Option to order trophic feeds only: 20 ml/h in 50-mL increments every 4 h to goal Time clock 24-h clock 7 am to 7 am 24-h clock 7 am to 7 am Makeup rate calculation Based on EN prescribed goal rate, hours EN held, and hours remaining Based on EN volume prescribed, volume missed, and hours remaining Example: SB feeds goal 70 ml/h held for 6 h (10 am to 4 pm) = new rate 80 ml/h from 4 pm to 7 am Example: SB feeds goal 1680 ml/24 h held for 6 h; 1470 ml remaining to infuse before 7 am 1470 ml/15 h New rate 98 ml/h from 4 pm to 7 am GRV threshold 350 ml 250 ml Promotility agents routinely used No Yes Protein supplement routinely used No Yes Maximum hourly infusion rate small 120 ml/h 150 ml/h bowel Gastric feeding maximum 400 ml every 4 h (given as intermittent feeding) 600 ml (given as continuous 150 ml/h) EN, enteral nutrition; FEED ME, Feed Early Enteral Diet adequately for Maximum Effect; GRV, gastric residual volume; PEP up, Protein-Energy Provision via the Enteral Route in Critically Ill Patients; SB, small bowel. adjusting the hourly rate as needed to reach the prescribed daily volume (ie, to make up for hours when the EN was held for procedures or diagnostic testing). The PEP up trial, completed after a positive feasibility trial, 6 included 18 mixed medical/surgical ICUs. However, similar to their initial feasibility trial, the combined population contained 80% 85% of patients from a MICU rather than a surgical ICU. 5,6 The investigators determined in both trials that implementation of this protocol, combined with a nurse-directed educational intervention, led to increased EN protein and calories delivered. 5,6 A quality improvement (QI) audit in May 2012 was performed at Barnes-Jewish Hospital, a bed teaching hospital in St Louis, Missouri. This audit demonstrated that the average daily EN volume intake for STICU patients was approximately 37%, below the desired literature-based goal of 80%. 4 Therefore, the unit-based multidisciplinary group decided to implement a volume-based approach in August The FEED ME (Feed Early Enteral Diet adequately for Maximum Effect) protocol, a modified PEP up volume-based feeding approach, was implemented following a nursing and physician educational campaign in August September The FEED ME protocol was modified to address local feeding barriers (eg, comfort of surgeons regarding maximum infusion rates, holding of EN for procedures, diagnostic testing, initial or return trips to the operating room, and gastric residual volumes >350 ml) and to maximize the likelihood of successful EN delivery (Table 1). This QI retrospective medical record review was completed to determine if the FEED ME protocol was more effective in the delivery of EN volume, calories, and protein without an increase in gastrointestinal (GI) complications (emesis, gastric residuals, or diarrhea) in STICU patients compared with the previous rate-based standard. Materials and Methods Development and Implementation of the FEED ME Protocol A STICU multidisciplinary team led by the full-time registered dietitian (RD) and including intensivists, the pharmacist, and STICU nurses (manager and bedside nurses) reviewed the PEP up protocol. A group decision was made to develop a variation targeting the specific barriers in the STICU and addressing the concerns of the physician leadership regarding rapid EN advancement in surgical patients. The FEED ME protocol was also designed to make up for lost EN infusion time due to diagnostic testing, radiographic studies, and multiple operations. The STICU physician leadership made the decision to enroll patients in the FEED ME protocol after they had achieved their target goal of EN, excluding those who were only fed a trophic rate, defined as 20 ml/h or a 50-mL bolus infused every 4 hours. The same eligibility criteria were used for both groups.

3 Taylor et al 3 Figure 1. FEED ME (Feed Early Enteral Diet adequately for Maximum Effect) protocol. BSSR, bedside shift report; Compass, electronic health record; D/C, discontinue; OR, operating room; Q, every; TF, tube feeding. The FEED ME protocol clearly illustrated how to calculate the new feeding rate or bolus based on feedings missed and when to revert back to the original goal (Figure 1). It further defined how to proactively feed when a nil per os (NPO) aftermidnight order was received. In the event of a GI complication or test that required withholding of EN, the protocol clearly stated to start a dextrose solution to avoid the risk of hypoglycemia (Figure 1). Following development of the protocol, an initial educational campaign was undertaken from August 1, 2012, to October 15, Several didactic sessions combined with individual instruction, bedside tools (Figure 1), and frequent reminders (eg, signs at the nurses station, reminders in the unit newsletter) were used to instruct the entire staff of 102 nurses and 8 nurse practitioners. Physician education was undertaken at the monthly quality improvement meeting, individually and via . As it pertained to the QI project, nursing compliance with the FEED ME protocol was intermittently monitored and feedback provided to individual nurses and management. Compliance with the FEED ME protocol during the data collection period was near 90% when the dietitian performed intermittent checks. The primary area of noncompliance was the direction for the nurse to increase the volume of EN delivered when NPO at midnight. Daily compliance for the entire review period was not collected. Sample Description and Project Design This pre- and postimplementation design took place in a 36-bed intensivist-driven STICU at Barnes-Jewish Hospital (BJH). During the preimplementation (rate-based) phase (January July 2012), the STICU contained 24 beds, which was expanded to 36 beds prior to the postimplementation (FEED ME, volume-based) period (October 2012 April 2013). The same fulltime RD supervised the nutrition care of the patients and was available to troubleshoot any FEED ME questions or concerns by the STICU staff for both phases. The electronic health records (EHRs) of adult (aged 18 years) patients admitted to the STICU over a 14-month period who received rate-based feeding (January 2012 July 2012) or FEED ME/volume-based feeding (October 2012 April 2013) were screened for possible inclusion. Patients who were mechanically ventilated upon arrival to the STICU or within 6 hours, had an STICU length of stay of at least 7 days, and received at least 72 hours of EN after obtainment of their target EN goal were included in the evaluation. The objective of this

4 4 Nutrition in Clinical Practice XX(X) QI review was to determine the efficacy and safety of the FEED ME protocol in patients who initially demonstrated EN tolerance. For example, nontrauma postsurgical patients admitted to the STICU for extubation (removal of the breathing tube) and 24-hour observation would have been excluded from the QI review, as would severely ill patients receiving highdose vasopressor support only allowed EN at a trophic rate. The same criteria were used when collecting data for the preimplementation group. All consecutive patients admitted to the STICU meeting inclusion criteria during the study periods were included in the data collection. All data were collected or calculated from information in the patients EHR (Sunrise Enterprise 5.5, an Eclipsys Corporation product, Atlanta, Georgia, USA), including age, race, ethnicity, sex, body weight (either reported or measured), body mass index (BMI), 7 calculated ideal body weight (IBW) 8 based on recorded height and sex, and clinical characteristics that comprised type of admission (trauma vs nontrauma surgical), Acute Physiology and Chronic Health Evaluation II (APACHE II) 9 criteria, diabetes history, gastroesophageal reflux disease (GERD) history, average daily blood glucose value, hours of mechanical ventilation, and length of ICU stay. Also collected from the EHR were data on nutrition care practices, including hours until EN initiation post-sticu admittance, hours until RD consult completed, type of EN product, gastric (prepyloric) vs small bowel feeding (postpyloric), 10 calories prescribed per kilogram based on their BMI, and grams of protein prescribed per kilogram of IBW. Nutrition outcome data extracted from the EHR included reasons EN was held, mean percentage of calories and protein delivered, mean calories/kg delivered, mean grams of protein/kg delivered, and incidence of GI complications including diarrhea, emesis, and gastric residual volume (GRV). The STICU QI committee approved the QI project proposal. The institutional review board (IRB) of the Washington University School of Medicine (the IRB used for patient research at BJH) evaluated the proposal and determined the QI project was exempt from IRB review. Approval from the IRB of Rutgers, the State University of New Jersey (formerly the University of Medicine and Dentistry of New Jersey), Newark campus was obtained prior to beginning the retrospective review. In accordance with the recommendations of the IRBs, patients <18 years of age, pregnant women, and prisoners were excluded from the review. This QI review project was supported by an unrestricted grant from the Dietitians in Nutrition Support, a practice group of the Academy of Nutrition and Dietetics. Definition of Variables Hours until EN initiation and RD consult documented in the EHR were calculated from STICU admission time. The EN formula and volume of EN delivered (in milliliters) were extracted from the EHR flow sheets kept by the nurses. Following approved standards in nutrition assessment in the SICU, 1,7,11 Table 2. Estimation of Caloric and Protein Needs. BMI (kg/m 2 ) Category Energy Requirements, a kcal/kg/d 24 < > Clinical Condition Recommended Daily Protein Intake 24 Protein Requirements, b g/kg IBW/d Normal 0.75 Metabolic stress (trauma/surgery) Hemodialysis Peritoneal dialysis Continuous dialysis, obesity BMI, body mass index; IBW, ideal body weight. a These values are recommended for critically ill patients and all obese patients; add 20% of total calories in estimating energy requirements in non critically ill patients. The lower range within each BMI category should be considered in insulin-resistant or critically ill patients to decrease the risk of hyperglycemia and infection associated with overfeeding. 24 b Additional protein requirements are needed to compensate for excess protein loss in specific patient populations, such as patients with burn injuries, large open wounds, external drains, and protein-losing enteropathy or nephropathy. Lower protein intake may be necessary in patients with chronic renal insufficiency not treated by dialysis and certain patients with liver disease and hepatic encephalopathy. 11 target daily calories and grams of protein per day were calculated and found in the RD consult (Table 2). Total calories and protein delivered were calculated from the milliliters of EN delivered multiplied by the calories per milliliter and grams of protein per milliliter for the EN formula used. Grams of protein delivered were divided by the patients actual and ideal body weights to determine grams per kilogram (g/kg) for each weight category. To calculate the percentage of volume, calories, and grams of protein delivered for each patient receiving some EN for at least 72 hours after achieving their target EN goal until STICU day 7, discharge, or death (whichever occurred first), we divided the total amount delivered (milliliters, calories, or grams) by the total amount prescribed. The average daily blood glucose (BG) level was calculated from totaling the BG levels reported on the daily blood metabolic panel drawn at 23:00 hours and dividing this total by the number of days. All values (total amounts delivered and BG) were averaged for each patient, which was then used to determine a group mean. Clinical outcomes were defined as the presence of GI complications that may be associated with EN, including emesis, diarrhea, and high GRV. Incidence of emesis was defined as the occurrence of gastric contents detected in the oropharynx or outside the mouth due to spontaneous regurgitation of EN. 12 Diarrhea was defined as the subjective interpretation by the RN of experiencing 3 loose or liquid stools/d. 13 According to STICU procedures, a GRV >350 ml was considered high; this

5 Taylor et al 5 Figure 2. Inclusion flowchart. EN, enteral nutrition; FEED ME, Feed Early Enteral Diet adequately for Maximum Effect; STICU, surgical/trauma intensive care unit. is below the recommended maximum amount in the literature yet agreed upon by the STICU nursing staff and leadership. 14,15 EN was held for 4 hours in the presence of GRV >350 ml. The nursing staff charted the occurrence of any these complications in the EHR input/output flow sheet. Incidences per patient were added to obtain a total incidence count for the patient group. Statistical Analysis The power calculation was based on data from the PEP up trial 5 and a smaller single ICU trial, 16 both of which demonstrated improvement in the delivery of EN calories and protein. In the PEP up study, the patients in the intervention group received 12% more calories (P <.01) and 14% more protein (P =.004) than did the control group, 5 whereas in the smaller study, the patients received 34% more calories (P <.0001) and 26% more protein (P =.007). Accordingly, a medium to high effect size of 0.6 for calories and medium to low effect size of 0.4 for protein were chosen. A priori analysis with G*Power version (Universitat Dusseldorf, Germany) 17 for a 1-tailed t test of the difference between 2 independent means (rate based vs FEED ME), using these effect sizes, and at an α error level of.05 with 80% power resulted in a sample size range of patients per group. The primary end points evaluated were the mean proportions of EN volume, calories, and protein received over those prescribed during the first 7 days in the STICU. The safety of the FEED ME protocol was compared with the rate-based approach by documenting the prevalence of GRV >350 ml, emesis, and diarrhea in each group. Continuous variables are expressed as means and standard deviations (SD), and categorical variables are presented as counts and percentages. For comparisons between the preimplementation (rate-based) and postimplementation (FEED ME) patient groups, Mann- Whitney U tests were used for continuous variables given the sampling distributions were nonnormal, and the χ 2 test was used for the categorical data. Statistical significance was considered at a 5% level (P.05). The data were analyzed using the Statistical Program for Social Sciences (SPSS) version 20.0 database (SPSS, Inc, an IBM Company, Chicago, IL). Results Descriptive Results In total, 1750 patients were admitted to the STICU during the 7-month preimplementation (rate-based) period, and 1998 were admitted in the postimplementation (FEED ME) period (Figure 2). Of these, 797 (46%) and 929 (46%) were mechanically ventilated within 6 hours of admission, respectively. Fifty-nine patients in the rate-based and 62 patients in the FEED ME group achieved target EN goal and received additional EN for at least 72 hours. An additional 5 patients in the rate-based group and 6 in the FEED ME group were excluded because they were pregnant, were younger than 18 years, or designated as a prisoner. Thus, there were 54 patients in the rate-based group and 56 in the FEED ME group in the final sample for the QI review. There were no significant differences between the groups according to demographic or clinical characteristics (Table 3). The STICU admits trauma patients and those from various surgical services, including vascular, colorectal, hepatobiliary, liver transplant, kidney/pancreatic transplant, orthopedic, gynecology oncology, and general surgery. Because of the overall small sample size and the small numbers in each surgical category except trauma (44% of the patients in the

6 6 Nutrition in Clinical Practice XX(X) Table 3. Clinical Characteristics and Demographics. Characteristic Rate Based (n = 54) FEED ME (n = 56) P Value a Men, No. (%) 29 (53.7) 31 (55.4).862 Race, No. (%) White 35 (64.8) 36 (64.3) Nonwhite 19 (35.2) 20 (35.7).94 Admission diagnosis, No. (%) Trauma 24 (44.4) 21 (37.5) Nontrauma surgical 30 (55.6) 35 (62.5).459 Diabetes history, No. (%) No 39 (72.2) 44 (78.6) Yes 15 (27.8) 12 (21.4).439 GERD history, No. (%) No 46 (85.2) 49 (87.5) Yes 8 (14.8) 7 (12.5).724 Age, mean (SD), y 55.3 (18.3) 53.7 (19.1).804 Body mass index, mean (SD), kg/m (10.2) 29.6 (8.6).938 APACHE II, mean (SD) 21.1 (6.7) 20.2 (5.2).51 Morning blood glucose, mean (SD) (20.7) (25.5).228 Mechanical ventilation, mean (SD), d 11.2 (7.6) 13.5 (9.0).146 Length of ICU stay, mean (SD), d 12.2 (7.7) 15.0 (8.6).053 Mortality, No. (%) 11 (20.3) 7 (12.5).264 APACHE II, Acute Physiology and Chronic Health Evaluation II; FEED ME, Feed Early Enteral Diet adequately for Maximum Effect; GERD, gastroesophageal reflux disease; ICU, intensive care unit; SD, standard deviation. a There were no statistically significant differences between the clinical characteristics and demographics in the 2 groups. rate-based group and 38% in the FEED ME group), the results are presented as trauma vs nontrauma (all other surgical services). The severity of illness was similar (APACHE II score: 21.1 [rate based] vs 20.2 [FEED ME]; P =.51) between the groups. There were 11 deaths in the rate-based group and 7 deaths in the FEED ME group (P =.264). There was a trend toward a longer length of stay in the FEED ME group (rate based, 12.2 days; FEED ME, 15.0 days; P =.053), but when the patients who died were removed, the strength of this difference lessened (P =.09). Time on mechanical ventilation (11.2 vs 13.5 days; P =.146) was similar. Nutrition Care Practices and Outcomes The nutrition care practices and outcomes are presented in Table 4. There were no significant differences between the groups with regard to hours until EN initiation (P =.36) and RD consult completed (P =.62), location of feeding tip (P = 1.0), or grams of protein prescribed per kilogram of IBW (P =.07). The only significant difference found between the groups was the calories/kg prescribed for the BMI range of , with an average of 20 calories/kg in the rate-based group and 25 calories/kg in the FEED ME group (P =.0001). The average amount of protein prescribed was similar between the groups (rate based, 1.4 g/kg; FEED ME, 1.6 g/ kg; P =.07). All levels prescribed fell within the current CPG recommendations. 1 A Kolmogorov-Smirnov 2-sample test demonstrated the 2 groups (rate-based vs FEED ME) had statistically significant different distributions with regard to percent volume (P =.018), protein (P =.017), and calories (P =.018) delivered, although the distribution pattern within each group for all 3 variables was similar (Figure 3). The rate-based group presented with bimodal distributions, while the FEED ME group distributions were slightly positively skewed (Figure 3). According to nursing staff documentation on the input/output flow sheet, the delivery of EN was held 122 times in the ratebased group and 96 times in the FEED ME group. EN was held 54 times for operations and 48 times for diagnostic studies or radiographic procedures in the rate-based group, which decreased to 48 times (P =.49) and 37 times (P =.142) in the FEED ME group. The incidence of high GRV (>350 ml; rate based, 20 times; FEED ME, 11 times; P =.34) was similar between the groups and led to short periods of EN cessation. When EN feeds were held for a high GRV, they were resumed after 4 hours if the rechecked GRV dropped below 350 ml. The incidence of emesis (rate based, 5 times; FEED ME, 2 times; P =.22) was comparable between the groups and again led to only short periods of EN cessation if at all. There was a significant difference in diarrhea (rate based, 0 times; FEED ME, 6 times; P =.046). More patients fed via the gastric route in the FEED ME group (n = 6) had diarrhea than did those fed via the small bowel (n = 1), but there was no documentation of EN being held due to diarrhea.

7 Taylor et al 7 Table 4. Nutrition Care Practices. Characteristic Rate Based (n = 54) FEED ME (n = 56) P Value a Route of EN, No. (%) Gastric 27 (50) 28 (50) Small bowel 27 (50) 28 (50) 1.0 Time to initiation of EN, mean (SD), h (17.8) 42.8 (25.1).362 Time to EN goal rate, mean (SD), h 63.6 (20.8) 60.9 (27.4).62 Time to documented RD consult, mean (SD), h 40.0 (25.4) 41.0 (25.8).979 Protein grams per kilogram IBW prescribed, mean (SD) 1.4 (0.3) 1.6 (0.4).073 Calories per kilogram prescribed based on BMI, mean (SD) BMI <15 NA NA BMI (4.76) 29.1 (3.52).836 BMI (4.02) 25.5 (4.39).946 BMI (3.84) 25.3 (3.14).0001 BMI (4.33) 16.9 (3.75).475 BMI, body mass index; EN, enteral nutrition; FEED ME, Feed Early Enteral Diet adequately for Maximum Effect; IBW, ideal body weight; NA, not applicable; RD, registered dietitian; SD, standard deviation. a There were no statistically significant differences in the nutrition care practices between the groups except for calories prescribed for 1 BMI category. Figure 3. Percent of volume delivered by group. FEED ME, Feed Early Enteral Diet adequately for Maximum Effect. The mean percent of calories delivered (calories delivered/calories prescribed) significantly increased when the FEED ME protocol was used (rate based, 63% ± 20%; FEED ME, 89% ± 9%; P <.0001) (Figure 4). The level of significance of protein delivered was dependent on the reference weight (actual weight vs IBW) used (Table 5). The grams of protein per kilogram of actual body weight increased from 1.13 ± 0.29 (rate based) to 1.26 ± 0.37 (FEED ME) (P =.036); this difference dropped in significance when the IBW was used, but the levels provided were higher (rate based, 1.41 ± 0.30; FEED ME, 1.56 ± 0.36; P =.07). Discussion This QI project demonstrates that a change in feeding practice to a volume-based approach increases the delivery of EN

8 8 Nutrition in Clinical Practice XX(X) Figure 4. Effect of feeding regimen on adequacy of Calories Delivered. FEED ME, Feed Early Enteral Diet adequately for Maximum Effect. P < Table 5. Effect of Feeding Regimen on Protein Delivered. Characteristic Rate Based (n = 54), Mean ± SD FEED ME (n = 56), Mean ± SD P Value Protein g/kg actual weight 1.13 ± ± Protein g/kg IBW 1.41 ± ± FEED ME, Feed Early Enteral Diet adequately for Maximum Effect; IBW, ideal body weight; SD, standard deviation. volume, calories, and protein to STICU patients, without increasing GRV or emesis and only slightly increasing diarrhea. The multidisciplinary STICU team chose to undertake this project given the priority of timely and adequate EN in the ICU as outlined by key opinion leaders. 18 This review was limited to STICU patients who had achieved their target goal of EN, which clearly biased the findings to patients with EN tolerance. Regardless, the clinical importance of establishing a protocol that allows the practitioner to take steps to overcome barriers and optimize EN delivery in STICU patients was illustrated. Interestingly, a bimodal distribution for volume, calories, and protein delivered was noted in the rate-based group (Figure 3). This can be explained by the lack of effort to replace volume of EN held for any reason. Therefore, most patients fell within 2 groups: those who experienced frequent interruptions of their EN (receiving low volumes) and those who did not (receiving consistent volumes). In a STICU, sicker patients are often the ones requiring multiple operations and frequent diagnostic testing. In the current QI review, nearly half of the rate-based patients (46%) received less than 50% of their prescribed calories mainly due to these reasons. A large retrospective study demonstrated that sicker patients benefit most from early and consistent EN. 19 Therefore, a rate-based protocol that does not address EN downtime may negatively affect the outcome of the sickest patients by affecting the delivery of EN calories and protein. In addition, work by Alberda et al 20 suggests the greatest benefit of EN delivery is realized when at least 80% of estimated calorie needs are being delivered. Conversely, the volume-based FEED ME distribution was positively skewed (Figure 3), suggesting regardless of the number of interruptions, a higher percentage of EN was delivered. Any interruption (whether it happened often or infrequently) to patient feeding was addressed, leading to an average of 89% of prescribed goal delivered for calories and protein. The success of the protocol was dependent on nursing compliance with the protocol. In the STICU, protocols are routinely used for all aspects of patient care, including sedation, ventilator weaning, glycemic control, electrolyte repletion, and feeding. Random audits performed during the QI review demonstrated a near 90% nursing compliance with the FEED ME protocol. However, during the period for which data were collected, weekly reminders were provided via signs throughout the STICU, and rewards (meal tickets) were given to nurses who correctly instituted the protocol in their patients. A followup audit completed 11 months after the end of the initial data collection period demonstrated 82% nursing compliance. Reminders continue to occur via the unit newsletter and sign postings. Follow-up audits will take place after any protocol changes and on a semi-annual basis to ensure sustained nursing compliance. Compared with indirect calorimetry, predictive equations of caloric requirement are accurate 50% 70% of the time with even less accuracy in the obese population. 21 Most practitioners use some type of predictive equation to help guide their EN prescription. The STICU RD used equations based on BMI, supporting the concept of permissive underfeeding in obese critically ill patients (Table 2). 11 Protein requirements were calculated based on the patients IBW, taking into account metabolic stress, renal function, and BMI (Table 2). 11 During the QI review, the same approaches to prescribe calories and protein were used in both patient groups. The amount of calories prescribed for patients in all BMI categories was comparable between the groups, except for a slight difference in the BMI group. However, in this BMI category, the average calories prescribed (20 kcal/kg in the rate-based and 25 kcal/kg in the FEED ME groups) fell within the standard of practice for the STICU (Table 2). The mean time to initiation of EN in both the rate-based and FEED ME groups (45.6 vs 42.9 hours) supported the practice of early EN ( 48 hours). 19,22,23 Recently, there has been academic discussion that early EN should be defined on CPGs as commencing within 24 hours of insult or injury as opposed to hours. Two meta-analyses provide insight on this debate. 24,25 The pooled data of 6 randomized controlled trials regarding EN in ICU patients demonstrated patients fed within 24 hours of injury or admittance to the ICU had significantly lower mortality (odds ratio [OR], 0.34; 95% confidence

9 Taylor et al 9 interval [CI], ; 234 patients) and significantly lower pneumonia rates (OR, 0.31; 95% CI, ; 80 patients) than did patients who were not fed within 24 hours. 24 In the current QI project, there was limited emphasis placed on improving time to EN initiation, as opposed to making up for lost time of EN delivery. If future CPGs suggest initiation of EN within 24 hours of insult or injury, the FEED ME protocol will be modified to focus on improvement of timely EN initiation. Prior to introduction of the new volume-based standard of care, some members of the QI committee voiced concern for a potentially higher risk of emesis, increased GRVs, and diarrhea with the use of a volume-based approach. For this reason, the initial maximum levels were limited to 120 ml/h for small bowel feeding and 400 ml for an intragastric bolus (infused over minutes via gravity drip), which was more conservative than the PEP up protocol, which allowed for a maximum of 150 ml/h via either small bowel or gastric feeding. A significant increase in diarrhea was noted in FEED ME patients, primarily in those fed into the stomach. Theoretically, higher bolus rates may have led to this intolerance. To counteract this finding, the QI committee decided to adjust a future iteration of the FEED ME protocol to continuous feeds only, regardless of location of feeding. However, it is difficult to interpret the absence of diarrhea in the ratebased group. This may have been due to the small sample size or data collection only for the first week of admission. STICU patients often receive large doses of narcotics and sedatives during their first week of stay that may decrease bowel motility. The FEED ME protocol had a GRV threshold of 350 ml compared with 250 ml in the PEP up protocol. The premise around holding EN for a high GRV is the expectation that a certain GRV is associated with an increased risk of aspiration and subsequently aspiration pneumonia. 14 However, research suggests high GRV does not correlate with aspiration. 14,26 In a recent randomized trial of 449 patients, the authors suggested GRV had no value in predicting risk of ventilator-associated pneumonia. 12 The control group had EN held for a GRV threshold of 250 ml, while GRV was not checked (and therefore EN not held) in the study group. They found, despite an increase of emesis in the intervention group (90 vs 60 events; P <.002), there was no significant difference in pneumonia rates (16.7% intervention vs 15.8% control) between the groups. 12 A higher proportion of patients in the intervention group did reach 80% of their calorie target (OR, 4.13; 90% CI, ; P <.001). 12 This study was presented to the QI committee for consideration. The committee decided to continue the 350-mL GRV threshold to use as a reportable measure of EN tolerance. In the STICU, if the GRV is notable for EN product, as opposed to bilious output, a small bowel feeding tube is placed to allow for simultaneous small bowel feeding and gastric decompression via the nasogastric tube. The number 1 barrier to EN delivery noted in the current QI review was not GI intolerance but rather NPO status for diagnostic testing and return trips to the operating room. This is similar to other studies of STICU patients that found the main barriers to EN were resuscitation, diagnostic imaging, and operative procedures. 27,28 To challenge the NPO after midnight philosophy, 1 group performed a feasibility study in a trauma ICU in which they implemented a decreased fasting protocol for patients scheduled for a tracheostomy, permanent feeding tube, orthopedic procedure, inferior venacaval filter placement, eye surgery, or ear, nose, and throat surgery. 29 The patients were allowed to receive small bowel feedings until the time of the procedure or gastric feeds until 45 minutes before the procedure. The authors found patients in the intervention group had a trend toward more EN provided and no significant difference in complications (including regurgitation and aspiration). 29 Approval for a similar approach to be incorporated into the FEED ME protocol is being sought from the anesthesiologist and surgical groups at the STICU institution. There are several limitations and strengths to this QI endeavor. Although not as robust as a prospective randomized control trial, the QI project does illustrate how bedside clinicians can translate research findings into improved patient care in their own institution. Data for analysis were obtained from the EHRs that are entered by nursing personnel, making data easy to read and interpret. However, nursing-entered data are subject to human error. Although the nursing staff was highly compliant with the protocol during the QI review, this level of compliance may decrease over time with new hires and fewer reminders and rewards. In the historical controls, an EN protocol was available, but a campaign with awards, frequent reminders, and bedside algorithms did not occur. The increased focus on the delivery of EN itself may have prompted the nursing staff to be more diligent in providing EN, accounting for some of the improvement in EN delivery. The PEP up sites also struggled to implement the protocol, with the absence of its use noted when the protocol champion was absent, such as on the weekends. 5 Even thus, the most efficient way to affect successful feeding is through the use of protocols that allow for the standardization of care and overcome impediments to EN. 30 Protocols may consist of preprinted orders, bedside algorithms, computerized order sets, or other forms of instructional material. 31,32 EN protocols enable bedside nurses to initiate, monitor, and advance EN safely and in a timely fashion to meet each individual patient s prescribed goal. 4,33 For the current QI effort, bedside algorithms (Figure 1) were placed in each room for nursing reference and to optimize protocol adherence. Last, this QI effort was restricted to STICU patients with demonstrated EN tolerance per the discretion of the physician leadership. This directive restricted data collection only to patients who initially tolerated advancement of their EN. However, the FEED ME protocol did address the most common barrier to adequate EN provision in the STICU: the holding of EN for tests and procedures. The goals of the initial steps in instituting the FEED ME protocol were to determine the safety and efficacy in patients tolerant of EN. Future steps will address patients restricted to trophic feeds and timely rate of advancement to goal EN delivery.

10 10 Nutrition in Clinical Practice XX(X) Conclusion A change in standard of practice to the FEED ME protocol, an EN volume-based feeding approach, was successfully implemented in an STICU. The FEED ME protocol offered a solution to overcome inadequate delivery of EN calories and protein due to feeding interruptions with only a slight increase in diarrhea. Continuous audits should be performed to encourage protocol adherence and determine effectiveness and safety, leading to planned improvements in the protocol as needed. Acknowledgments This work would not have been possible without the care and attention of the nursing and physician staff of the surgical trauma intensive care unit at Barnes-Jewish Hospital, St Louis, Missouri. References 1. McClave SA, Martindale RG, Vanek VW, et al. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (A.S.P.E.N.). JPEN J Parenter Enteral Nutr. 2009;33(3): Lewis SJ, Andersen HK, Thomas S. Early enteral nutrition within 24 h of intestinal surgery versus later commencement of feeding: a systematic review and meta-analysis. J Gastrointest Surg. 2009;13(3): Drover JW, Cahill NE, Kutsogiannis J, et al. Nutrition therapy for the critically ill surgical patient: we need to do better! JPEN J Parenter Enteral Nutr. 2010;34(6): Heyland DK, Cahill NE, Dhaliwal R, Sun X, Day AG, McClave SA. Impact of enteral feeding protocols on enteral nutrition delivery: results of a multicenter observational study. JPEN J Parenter Enteral Nutr. 2010;34(6): Heyland DK, Murch L, Cahill N, et al. Enhanced protein-energy provision via the enteral route feeding protocol in critically ill patients: results of a cluster randomized trial. Crit Care Med. 2013;41: Heyland DK, Cahill NE, Dhaliwal R, et al. Enhanced protein-energy provision via the enteral route in critically ill patients: a single center feasibility trial of the PEP up protocol. Crit Care. 2010;14(2):R Pesce-Hammond K, Wessel J. Nutrition assessment and decision making. In: Merritt R, ed. The A.S.P.E.N. Nutrition Support Practice Manual. 2nd ed. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 2005: Holcombe B. Fundamentals of nutrition support practice and management. In: Merritt R, ed. The A.S.P.E.N. Nutrition Support Practice Manual. 2nd ed. Silver Spring, MD: American Society for Parenteral and Enteral Nutrition; 2005: Knaus WA, Draper EA, Wagner DP, Zimmerman JE. APACHE II: a severity of disease classification system. Crit Care Med. 1985;13(10): Jabbar A, McClave SA. Pre-pyloric versus post-pyloric feeding. Clin Nutr. 2005;24(5): Klein S, Taylor B, Rubin D. Nutrition support in patients with gastrointestinal disease. In: Wolfe M, ed. Digestive Diseases. 2nd ed. St Louis, MO: Elsevier; Reignier J, Mercier E, Le Gouge A, et al. Effect of not monitoring residual gastric volume on risk of ventilator-associated pneumonia in adults receiving mechanical ventilation and early enteral feeding: a randomized controlled trial. JAMA. 2013;309(3): Lebak KJ, Bliss DZ, Savik K, Patten-Marsh KM. What s new on defining diarrhea in tube-feeding studies? Clin Nurs Res. 2003;12(2): Metheny NA, Schallom L, Oliver DA, Clouse RE. Gastric residual volume and aspiration in critically ill patients receiving gastric feedings. Am J Crit Care. 2008;17(6): McClave SA, Lukan JK, Stefater JA, et al. Poor validity of residual volumes as a marker for risk of aspiration in critically ill patients. Crit Care Med. 2005;33(2): Sheean PM, Peterson SJ, Zhao W, Gurka DP, Braunschweig CA. Intensive medical nutrition therapy: methods to improve nutrition provision in the critical care setting. J Acad Nutr Diet. 2012;112(7): Faul F, Erdfelder E, Buchner A, Lang AG. Statistical power analyses using G*Power 3.1: tests for correlation and regression analyses. Behav Res Methods. 2009;41(4): Weenen TC, Jentink A, Pronker ES, et al. Patient needs and research priorities in the enteral nutrition market: a quantitative prioritization analysis [published online December 13, 2013]. Clin Nutr. 19. Khalid I, Doshi P, DiGiovine B. Early enteral nutrition and outcomes of critically ill patients treated with vasopressors and mechanical ventilation. Am J Crit Care. 2010;19(3): Alberda C, Gramlich L, Jones N, et al. The relationship between nutritional intake and clinical outcomes in critically ill patients: results of an international multicenter observational study. Intensive Care Med. 2009;35(10): Cooney RN, Frankenfield DC. Determining energy needs in critically ill patients: equations or indirect calorimeters. Curr Opin Crit Care. 2012;18(2): Moore FA, Moore EE. The evolving rationale for early enteral nutrition based on paradigms of multiple organ failure: a personal journey. Nutr Clin Pract. 2009;24(3): Heighes PT, Doig GS, Sweetman EA, Simpson F. An overview of evidence from systematic reviews evaluating early enteral nutrition in critically ill patients: more convincing evidence is needed. Anaesth Intensive Care. 2010;38(1): Doig GS, Heighes, PT, Simpson F, et al. Early enteral nutrition, provided within 24 h of injury or intensive care unit admission, significantly reduces mortality in critically ill patients: a meta-analysis of randomised controlled trials. Intensive Care Med. 2009;35: Doig GS, Heighes PT, Simpson F, Sweetman EA. Early enteral nutrition reduces mortality in trauma patients requiring intensive care: a meta-analysis of randomised controlled trials. Injury. 2011;42: Metheny NA, Stewart BJ, McClave SA. Relationship between feeding tube site and respiratory outcomes. JPEN J Parenter Enteral Nutr. 2011;35(3): Passier RH, Davies AR, Ridley E, McClure J, Murphy D, Scheinkestel CD. Periprocedural cessation of nutrition in the intensive care unit: opportunities for improvement. Intensive Care Med. 2013;39(7): Chung CK, Whitney R, Thompson CM, Pham TN, Maier RV, O Keefe GE. Experience with an enteral-based nutritional support regimen in critically ill trauma patients. J Am Coll Surg. 2013;217: Pousman RM, Pepper C, Pandharipande P, et al. Feasibility of implementing a reduced fasting protocol for critically ill trauma patients undergoing operative and nonoperative procedures. JPEN J Parenter Enteral Nutr. 2009;33(2): Cahill NE, Narasimhan S, Dhaliwal R, Heyland DK. Attitudes and beliefs related to the Canadian critical care nutrition practice guidelines: an international survey of critical care physicians and dietitians. JPEN J Parenter Enteral Nutr. 2010;34(6): Mackenzie SL, Zygun DA, Whitmore BL, Doig CJ, Hameed SM. Implementation of a nutrition support protocol increases the proportion of mechanically ventilated patients reaching enteral nutrition targets in the adult intensive care unit. JPEN J Parenter Enteral Nutr. 2005;29(2): Arabi Y, Haddad S, Sakkijha M, Al Shimemeri A. The impact of implementing an enteral tube feeding protocol on caloric and protein delivery in intensive care unit patients. Nutr Clin Pract. 2004;19(5): Woien H, Bjork IT. Nutrition of the critically ill patient and effects of implementing a nutritional support algorithm in ICU. J Clin Nurs. 2006;15(2):