March 2013 e-journal club
March came in like a lion, with our biggest snowfall of the year, but fortunately most of the snow was gone in time for the arrival of our Weekend Warrior Program. We had a super group here for the March 16-17th Weekend Warrior program. We like meeting our peers from around the county, and when everyone contributes and asks questions, it makes it more enjoyable for everyone participating.
Mancl EE, Muzevich KM. Tolerability and Safety of Enteral Nutrition in Critically Ill Patients Receiving Intravenous Vasopressor Therapy. JPEN 2013 Jan 17 [epub ahead of print].
This was a retrospective, observational study using the electronic medical record for a 1 year period of 259 patients who simultaneously received enteral nutrition (EN) and an intravenous vasopressor medication. The primary outcome of this study was EN tolerance, as defined by the absence of: gastric residuals ≥300 mL, emesis, positive finding on KUB or abdominal CT, and bowel ischemia/perforation. The secondary outcomes included rates of adverse events, characterization of IV vasopressor dosages during overlap episodes, and mortality rates. The investigators looked for any potential relationship between administration of a given IV vasopressor as well as vasopressor dose and EN tolerability. Additional medical, demographic and nutrition data were collected as mentioned in the results section. Vasopressor dosages were converted to norepinephrine equivalents = [norepinephrine (mcg/min)] + [dopamine (mcg/kg/min) ÷ 2] + [epinephrine (mcg/min)] + [phenylephrine (mcg/min) ÷ 10] + [vasopressin (units/h) × 8.33].
The authors also provided details of the clinical course of the 3 patients that developed ischemic bowel during this observational study – I recommend reading the full text of this study for the valuable points made in the paper, and for full information on the cases.
Inclusion and Exclusion Criteria were:
Patients with a charted EN volume while receiving an IV vasopressor (dopamine, epinephrine, norepinephrine, phenylalanine, and/or vasopressin.)
Age ≤18 years old and were treated in a pediatric or neonatal ICU, overlap of EN and IV vasopressor ≤1 hour, or if data were identified as missing or incomplete during chart review.
The review identified 346 episodes of feeding/vasopressor overlap in 259 patients (some patients had reoccurrences of episodes of feeding with vasopressors). There were 259 cases of EN tolerance, and 87 episodes of patients meeting the study definition of feeding intolerance while receiving vasopressor medications. The average blood pressure during overlap was similar between groups.
Emesis occurred in 31 (9.0%), elevated gastric residuals (≥300 mL) in 50 (14.5%) and increased serum lactate in 106 (30.6%) overlap episodes. Serum lactate only rose to >2mmol/L in half of the patients with elevated lactate, and the maximum lactate during overlap was not significantly different between groups. Abdominal radiograph (KUB) was obtained in 11.9% and CT in 4% of overlap cases, with 36.6% of the abdominal KUBs and 0.9% of the abdominal CTs reported to have positive findings for intolerance (ileus, small bowel obstruction, or signs of bowel ischemia/perforation).
There were 3 cases of bowel ischemia/perforation identified (0.12% of patients)
A dose-response relationship was observed between maximum norepinephrine equivalent and likelihood of tolerating EN. Patients who tolerated EN received a lower maximum norepinephrine equivalent dose compared with those who did not tolerate EN (12.5 vs 19.4 mcg/min, p = 0.0009). Patients who were never prescribed vasopressin during the overlap episode were more likely to tolerate EN compared with those who received vasopressin (77.9% vs. 58.9%, p = 0.0027). Patients who never received dopamine tolerated EN more frequently than those who received dopamine (77.6% vs. 63.8%, p = 0.018), and patients who received phenylephrine were more likely to tolerate EN compared with those who did not receive phenylephrine (100.0% vs. 73.0%, p = 0.0023). None of these associations were dose related. Patients who tolerated EN were less likely to receive a promotility agent compared with those who did not tolerate EN (15.0% vs. 44.0%, p < 0.0001).
The median duration of feeding/vasopressor overlap was significantly shorter in the group that tolerated EN (26 vs. 65 hours, p = 0.0002) and the mean caloric intake was higher in the group that had objective tolerability (13.6 vs. 10.9 kcal/kg/d, p = 0.0034). The univariate analysis showed that patients who tolerated EN were less likely to experience a rise in serum lactate compared with those who did not tolerate EN (25.5% vs. 46.0%, p = 0.0003), however, serum lactate was not significantly associated with intolerance in multivariate analysis.
“EN is relatively well tolerated in patients receiving IV vasopressor support equivalent to 12.5 mcg/min of norepinephrine or less. Tolerability was less likely in patients receiving higher doses of IV vasopressors and in those receiving dopamine or vasopressin. These patients should be monitored more closely for signs of intolerance. In summary, critically ill patients receiving IV vasopressor support generally tolerate EN.”
This study is valuable because it adds to the very limited pool of data about the interaction between patients that receive vasopressor medications and EN. However, as the authors point out in the discussion section, it is a retrospective study design with inherent selection bias. Results of associations should not be used to imply causality and, as the authors state, “are meant to be hypothesis generating and ideally would lend support for a prospective study.” The significant association between use of prokinetic medications and feeding intolerance reported by this study is one example where it is intuitive that the medication is unlikely to have caused the intolerance. It is important to remember that any association between feeding intolerance and vasopressor medication may also not be cause-and-effect.
One issue that we discussed is the use of relatively physiologic amounts of gastric residuals (300 mL) as an indicator feeding intolerance, and the possibility that some patients that were classified as intolerant may actually have been fine. Certainly a 305 mL residual by itself would generally not be considered in the same category of feeding intolerance as bowel ischemia or perforation. It would be helpful to see potential feeding intolerance (elevated residuals) separated from negative clinical outcomes.
The case reports of the patients that had ischemic bowel are remarkable in that all of the patients were reportedly receiving intragastric feeding, in contrast to most other case reports. As mentioned above, these are worth reading in full.
Our Take Home Message (s)
Most patients that are hemodynamically stable and receiving modest doses of vasopressor medications tolerate EN.
The association between a particular dose of vasopressor medication and feeding tolerance needs to be tested in randomized studies before being worked into clinical guidelines, etc.
Other News on the UVAHS GI Nutrition Website: (www.ginutrition.virginia.edu):
--Tuesday, April 16th--Considerations for Nutrition Support in the Adult Cystic Fibrosis Patient--Christie Rogers, MS, RD, CNSC
--Tuesday, May 14th--The Role of Fiber in Nutrition Support--Joe Krenitsky, MS, RD
--Tuesday, June 18th--Nutrition Management of Food Allergies in the Pediatric Patient--Brandis Roman, RD
Check out What’s New:
Latest Practical Gastroenterology article:
--Carey, S. Bone Health After Major Upper Gastrointestinal Surgery. Practical Gastroenterology 2013;XXXVII(3):46.
Joe Krenitsky MS, RD
Carol Rees Parrish MS, RD
PS – Please feel free to forward on to friends and colleagues.