TMCnet News
Hypocaloric Enteral Nutrition Protects Against Hypoglycemia Associated with Intensive Insulin Therapy Better Than Intravenous Dextrose [American Surgeon, The](American Surgeon, The Via Acquire Media NewsEdge) Intensive insulin therapy treats hyperglycemia but increases the risk of hypoglycemia. Typically, intravenous dextrose is given to prevent hypoglycemia; however, enteral nutrition is preferred. We hypothesized that the provision of hypocaloric enteral nutrition would protect against hypoglycemia. A retrospective analysis was performed evaluating patients treated with intensive insulin therapy comparing the use of enteral nutrition versus a dextrose-only intravenous solution. Nutrition in the 2 hours before each blood glucose test was assessed, and the association with hypoglycemia (50 mg/dL or less) evaluated. Risk of hypoglycemia as a function of nutrition type and rate was estimated by multivariable regression. A total of 26,140 blood glucose tests were collected on 1289 patients. Hypoglycemia occurred in 6.4 per cent of patients. In regression models, enteral nutrition was the strongest protective factor against hypoglycemia (P < 0.001) with the largest risk reduction (steepest portion of the curve) occurring at 60 per cent goal. Hypocaloric enteral nutrition showed a greater risk reduction than a peripheral dextrose-only intravenous solution alone. In the setting of intensive insulin therapy, the provision of enteral nutrition, even if hypocaloric, is sufficient to protect against hypoglycemia. Future prospective studies should evaluate the efficacy of enteral nutrition in reducing the risk of hypoglycemia and whether lower rates of hypoglycemia correspond to improved outcomes. HYPERGLYCEMIA IN CRITICALLY ill patients has been shown to increase infectious complications and mortality.1-3 As a result, intravenous insulin therapy has been widely adopted to control hyperglycemia and im- prove outcomes.4, 5 However, there is ongoing concern regarding the rates of hypoglycemia in patients treated withintensiveinsulintherapy(IIT)tomaintaintightblood glucose (BG) control (80 to 110 mg/dL).6-9 Furthermore, recenttrialshavefoundanincreaseinmortalityinpatients treated with IIT.10, 11 ThelandmarkstudyadvocatingIITbyVandenBerghe et al. was unique in that a dose of 200 to 300 g (680 to 1020 kcal) of intravenous dextrose was provided in the first 24 hours after intensive care unit admission followed by the initiation of either total parenteral nutrition (TPN) or en- teral nutrition (EN) within the first 48 hours after admis- sion.4 Since that study, however, little emphasis has been placed on the potential role of early nutritional provision in improving outcomes in patients treated with IIT. There is ongoing controversy regarding the timing of initiation and type of nutrition that is optimal for critically ill patients. It is well known that fasting worsens insulin resistance, and both early feeding and preoperative car- bohydrate administration are associated with decreased inflammation during critical illness or injury.12, 13 How- ever, Casear and colleagues14 demonstrated fewer com- plications in patients started on parenteral nutrition on Day 8 compared with patients initiated on parenteral nutrition on Day 2. A number of studies have shown improved outcomes with hypocaloric feeds (to provide 33 to 70% of daily carbohydrate needs and full protein needs) in obese patients.12, 15-21 Such feeding regimens provide better metabolic equilibrium and nitrogen bal- ance while preserving lean muscle mass without altering BG control.12, 16 We have previously shown that the provision of balanced nutrition, defined as nutrition that provides both carbohydrate and protein calories, protects against hypoglycemia in the critically ill surgical patient.22 However, the volume of balanced nutrition required to protect against hypoglycemia has not been previously studied. This analysis builds on the previous study by the researchers,22 which aims to determine the dose- response to EN. Although EN is the first choice,16 this dose relationship holds true whether balanced nutrition is EN or TPN.22 We sought to determine the required volume of EN required to minimize a patient's risk of subsequent hypoglycemia (50 mg/dL or less). Materials and Methods A retrospective analysis of a prospectively collected data set was performed on a cohort of critically ill surgical patients who were admitted to the surgical intensive care unit (SICU) of an academic medical center from June 2006 to November 2010 and received IIT. This study was approved by the Institutional Re- view Board at the institution. Insulin Protocol and Blood Glucose Measurements The protocol for insulin and BG measurements at Vanderbilt University Medical Center is described in detail elsewhere.22 Briefly the glucose target range of all critically ill, mechanically ventilated patients is between 80 and 110 mg/dL. If a patient has serum BG values above 110 mg/dL, the patient is placed on intravenous computerized insulin protocol to manage the BG levels. BG measurements are performed every 2 hours by trained nurses using the SureStep^ Pro (OneTouch^; Lifescan, Inc., Milpitas, CA) Professional Blood Glu- cose Monitoring System. The Computerized Physician Order Entry (CPOE) algorithm uses a modification to a protocol described by White et al.23 and Bode et al.24 with doses computed using the following formula: Insulin doseðunits=hrÞ4ðMultiplierÞ: The Multiplier is a variable that is initially set at 0.03 and adapts according to a set protocol. It can never fall below zero. BG values exceeding the high target threshold (110 mg/dL) on two consecutive BG mea- surements, or exceeding 200 mg/dL on one reading, trigger a multiplier increase of 0.01. BG values below the low target threshold (80 mg/dL) decrease the multiplier by 0.01, and BG values below 60 mg/dL decrease the multiplier by 0.02. When BG values fall below 60 mg/dL, the protocol orders a calculated dose of intravenous 50 per cent dextrose to correct or pre- vent hypoglycemia. The intravenous insulin infusion is simultaneously withheld for 2 hours. The protocol requires infusion of D10W at 30 mL/hr (10 kcal/hr) if the patient is not receiving either EN or TPN to protect against hypoglycemia. Nutrition Protocol The protocol for nutrition at Vanderbilt University Medical Center is described elsewhere in more de- tail.22 Briefly, all patients admitted to the SICU and placed on IIT receive minimum nutritional provision of D10W (10% dextrose in water) at 30 mL/hr (72 g [245 kcal] of dextrose/24 hours). Enteral nutrition at half goal rate (as determined by ideal body weight, degree of critical illness, and caloric need) is begun within 48 hours of admission if the patient is able to tolerate EN. Enteral nutrition was provided through both intragastric and postpyloric tubes, depending on which was present. Intragastric tubes are used in our SICU unless the patient is not tolerating intragastric tube feeds (i.e., have large gastric residuals), in which case a postpyloric tube would be placed. The tube feed regimen consists of an immune-enhancing, peptide- based formulation with 1.5 cal/mL. This formulation provides 45 per cent of total calories in the form of carbohydrate. If full EN is not tolerated, TPN is begun by the seventh day after SICU admission. Clinical nutritionists provide daily bedside management of in- dividual nutrition needs for all patients in the SICU. Inclusion criteria consisted of treatment with a com- puterized IIT protocol to maintain euglycemia (80 to 110 mg/dL), remaining on protocol for at least five BG measurements and 12 hours, and receiving EN (but never TPN) during the course of the SICU stay. TPN was excluded because it is a known confounder and associ- ated with both the risk of hypoglycemia and the receipt of EN. Patients who died within 24 hours of admission to the intensive care unit (ICU) were excluded. Data Collection Patient factors including gender, age, weight, date of hospital and ICU admission, mortality, Acute Physi- ology and Chronic Health Evaluation II (APACHE II) score at ICU admission, and diagnosed diabetes were obtained from the electronic health record and the SICU registry. The SICU registry is an Institutional Review Board-approved repository of clinical data that is prospectively collected and maintained on every patient admitted to the SICU. The SICU registry has been previously used for research purposes to dem- onstrate whether the provision of balanced nutrition may alter the risk for hypoglcemia.22 Patient weight was missing for 99 patients (7.7% of the study group) and was imputed to the population median (85 kg) and APACHE II score was missing for 45 patients (3.5% of study group) and imputed to the population median.22 All components of the insulin protocol (BG values, test times, insulin dose, multiplier, adherence to IIT pro- tocol, and treatment with a dose of 50% dextrose) are recorded prospectively, and each eligible patient had multiple recorded BG measurements. BG measure- ments during the first 12 hours (induction phase) on the IIT protocol were excluded as were protocol in- terruptions greater than 6 hours. Because BG levels are not normally distributed, BG values were transformed for analysis.25 Time variable data for each patient in- cluded data from the IIT protocol and current nutri- tional provision. The provision of dextrose and EN was recorded every hour and administrations in the 2 hours before the BG test were summed. Oral intake, which was very rare, was excluded. Statistical Analysis Continuous variables that were normally distributed were summarized by reporting the mean and standard deviation and compared using a two-sample t test. Continuous variables that were not normally distrib- uted were summarized by reporting the median and interquartile range and compared using the Wilcoxon rank-sum test. Differences in proportions were com- pared using the x2 test. Multivariable logistic re- gression models were fit to predict hypoglycemia at the next BG test based on volume of EN or dextrose- containing intravenous fluid (IVF) provided over the past 2 hours, weight, time since previous BG measurement, and count of previous ''near-miss'' hy- poglycemic events. Other variables considered in re- gression models included nonlinear fits of tube feeding and dextrose administration, patient age, gender, APACHE II score, time on IIT protocol, and recent use of vasopressors. These other predictors variables were not included in the final model because they were not associated with the outcome and the small number of outcomes constrained robust model-building. A hy- poglycemic event was defined as BG 50 mg/dL or less to maximize power; BG levels 40 mg/dL or less did not provided an adequate number of samples to allow for analysis. Measurements of BG and time between successive BG measurements were calculated using data beginning 12 hours after insulin therapy initiation to allow for stabilization on the IIT protocol. BG values are repeated measures within patients and this correlation must be accounted for within the model to prevent overly optimistic standard errors. Therefore, robust covariance estimates using the Huber-White method to adjust the variance-covariance matrix for clustered data are presented.26, 27 Model terms in- cluded diagnosed diabetes, weight (three-knot spline), time elapsed from the previous BG test, the cumulative count of previous ''near-miss'' hypoglycemic events (BG 60 mg/dL or less), and provision of EN or dextrose- containing IVF (mL) in the 2 hours preceding the index BG measurement. Patient weight, rather than body mass index, was included in the model as a result of missing height values for patients early in the study period. Variables were fit as restricted cubic splines where relationships were nonlinear.28 The model was validated with 100 bootstrap iterations using the R RMS package.28 All confidence intervals are at the 95 per cent level, and a two-sided P value of < 0.05 in- dicated statistical significance. Analysis was per- formed using R Version 2.11.0 (www.r-project.org). Results During the study period, 26,140 BG measurements were obtained in 1,289 patients meeting inclusion criteria. Hypoglycemia (50 mg/dL or less) occurred in 83 of 1289 patients (6.4%) and 104 (0.4%) of BG measurements. Demographics and clinical character- istics of those who experienced one or more hypo- glycemic events at some point during their SICU stay compared with those who did not are displayed in Table 1. Patients who experienced a hypoglycemic episode weighed less, were sicker (higher APACHE II score), and had an increased length of stay (LOS) with higher maximum glucose and higher mortality. A multivariable logistic regression model was built to determine the association with hypoglycemia at the subsequent BG test. Predictors independently associ- ated with hypoglycemia at the next BG measurement are displayed in Table 2. The risk of hypoglycemia by volume of EN or dextrose-containing IVF infused in 2-hour increments was assessed and is shown in Figure 1. The risk of hypoglycemia at the next BG measurement decreased linearly as volume of EN or dextrose-containing IVF increased. In the case of EN, the slope of the re- lationship was steeper; the risk of hypoglycemia de- creases faster with each milliliter of EN than with each milliliter of dextrose. Assuming a 70-kg patient who requires 30 kcal/kg/day, 1400 mL of a 1.5-kcal/mL tube feed would be required to provide full nutritional support. A goal tube feed rate of 60 mL/hour would be required to provide this volume of tube feeds. Figure 1 shows that the provision of approximately 60 per cent of goal tube feeds (35 mL/hour, to provide 20 kcal dextrose/hr) is associated with the largest decrease in risk of hypoglycemia (steepest portion of the curve) at the next BG measurement. In the case of dextrose- containing IVF, the risk of hypoglycemia steadily de- creases to a risk of just below 0.05; however, large volumes of IVF (up to 150 mL per hour, to provide 25 kcal dextrose/hr) are required to achieve even this modest decrease in risk of hypoglycemia. The odds ratio for hypoglycemia observed with hypocaloric enteral nutrition provided at a rate of 30 mL/hr was 0.56 (standard error interval 0.53 to 0.59) compared with an odds ratio of 0.78 (standard error interval 0.71 to 0.85) for D5 infused at 30 mL/hr. Discussion The use of IIT to prevent hyperglycemia is an im- portant aspect of the care of the critically ill surgical patient. A number of factors have been shown to be associated with increased risk of hypoglycemia, in- cluding time on IIT,29 need for dialysis, previous di- abetes, sepsis, the need for vasopressors, and body mass index.30-32 We have previously shown that the provision of balanced nutrition, to include either EN or TPN, is protective against hypoglycemia.22 We are unable to compare TPN dose versus EN dose and the associated risk of hypoglycemia in this study because TPN was an exclusionary factor. However, we expand on the dose-response of EN. In this study, we have shown that the provision of EN, even if hypocaloric, is sufficient to protect against hypoglycemia. Although additional benefit in guarding against hypoglycemia is observed with the provision of full nutritional support, halving the baseline risk of hypoglycemia (from 0.8 to 0.4) is more important than halving the attenuated risk of hypoglycemia (from 0.4 to 0.2). In theory, halving the baseline risk prevents two episodes of hypoglycemia in 1000 BG tests and the second reduction prevents only a single episode of hypoglycemia in 1000 BG tests. Furthermore, the infusion of dextrose-contain- ing IVF alone, although protective against hypogly- cemia, requires much higher volumes to halve the baseline risk of hypoglycemia. Although the D10 used in our study can be given through peripheral intravenous line, higher concentrations require cen- tral access. The provision of hypocaloric EN is able to reduce the risk of hypoglycemia at the next BG measurement to less than 0.4, whereas the provision of even 150 mL/hr of five per cent dextrose-contain- ing IVF decreases the risk of hypoglycemia only to approximately 0.5. Whether the protective effect of EN is the result of energy concentration, caloric content, or route of administration (intravenous vs enteral) requires additional study. Figure 1 demonstrates that the provision of hypo- caloric nutrition at 35 mL/hr (60% goal rate to provide 20 kcal dextrose/hr) reduces the risk of hypoglycemia from 0.8 to 0.4. The provision of D5 at more than 150 mL/hr (25.5 kcal dextrose/hr) is required to achieve a similar decrease in risk of hypoglycemia. This sug- gests that EN is more effective in reducing the risk of hypoglycemia than a dextrose-containing IVF alone, perhaps as a result of the mixed macronutrients and additives (antioxidants, immune modulators, and vi- tamins) present in EN formulas or the route of ad- ministration itself. Furthermore, much lower volumes of EN than five per cent dextrose IVF are needed to reduce the risk of hypoglycemia. Critically ill patients frequently experience volume overload secondary to multiple intravenous medications, the use of vaso- pressors, and the need for intravenous fluid re- suscitation after surgery or traumatic injury. This state of volume overload can lead to a multitude of com- plications, including abdominal compartment syn- drome, pleural effusions, and pulmonary edema. Many critical care practitioners thus encourage the judicious use of fluid in the critically ill patient population to prevent such complications. Providing D5 at 150 mL/hr to prevent hypoglycemia would supply an additional 3.6 L of fluid daily, which is prohibitive in the ICU setting. In contrast, the provision of hypocaloric EN achieves the same reduction in risk of hypoglycemia while supplying far less fluid (840 mL/day). This study expands on a previous study investigating balanced nutrition and hypoglycemia in the critically ill surgical patient.22 Along with the previous study, this study cannot confirm causality as a result of the retrospective review and regression analysis. Addi- tionally, APACHE II scores and weight were imputed for missing data, thus introducing potential bias if true values differed greatly from imputed values, similarly performed in the previous study.22 The definition of a hypoglycemic event at 50 mg/dL limited the number of potential events for analysis, which limited statisti- cal power; however, this definition of hypoglycemia has been previously used.22 Associations in these data, which exclude all patients who were exposed to TPN, are similar to our more comprehensive ICU database but remove the confounding effects of TPN. The in- sulin protocol used by our SICU uses point-of-care testing to monitor BG. These can be less reliable at low BG values; therefore, this introduces some variability into our data at low BG levels. Additionally, it was not possible from our data to specify the type of EN pro- vided to each patient; however, protocol dictates the use of an immune modulation formula in our SICU. Finally, the route of enteric feeding (whether intra- gastric or postpyloric) and tolerance of said feeds could not be ascertained from our data. In our unit, intra- gastric feeds are generally provided, unless high gastric residuals are observed. In that case, a postpyloric tube would be placed. Given the high adherence to our SICU nutritional protocol, we expect that the majority of patients in our study were fed through intragastric tubes and that all were tolerant of feeds, whether provided intragastrically or postpylorically. However, this re- mains a limitation, because we cannot specify whether the protective effect of hypocaloric nutrition is ob- served in patients being fed intragastrically versus postpylorically nor can we comment on the relation- ship between tolerance of feeds and hypoglycemia. Conclusion This study supports the conclusion that the provision of hypocaloric EN to critically ill patients decreases the risk of subsequent hypoglycemia. There is little additional decrease in risk when full nutritional sup- port is provided. Future studies are needed to ascertain the role played by individual nutritional components (including carbohydrate, fat, and protein) in decreasing a patient's risk of hypoglycemia. Additionally, whether decreases in rates of hypoglycemia in patients treated with IIT correspond to improved outcomes remains to be more thoroughly evaluated. Acknowledgments Financial support was provided in part by a National Institutes of Health T32 training grant in Diabetes and Endocrinology 5T32DK007061-35 (R.M.K.). REFERENCES 1. Dungan KM, Braithwaite SS, Preiser JC. Stress hyper- glycaemia. Lancet 2009;373:1798-807. 2. Furnary AP, Wu Y. Clinical effects of hyperglycemia in the cardiac surgery population: the Portland Diabetic Project. Endocr Pract 2006;12:22-6. 3. Furnary AP, Wu Y, Bookin SO. Effect of hyperglycemia and continuous intravenous insulin infusions on outcomes of cardiac surgical procedures: the Portland Diabetic Project. Endocr Pract 2004;10:21-33. 4. van den Berghe G, Wouters P, Weekers F, et al. Intensive insulin therapy in the critically ill patients. N Engl J Med 2001;345:1359-67. 5. Van den Berghe G, Wilmer A, Hermans G, et al. Intensive insulin therapy in the medical ICU. N Engl J Med 2006;354: 449-61. 6. Preiser JC, Devos P. Clinical experience with tight glu- cose control by intensive insulin therapy. Crit Care Med 2007; 35:S503-7. 7. Devos P, Presier J. Current controversies around tight glucose control in critically ill patients. Curr Opin Clin Nutr Metab Care 2007;10:206-9. 8. Schultz MJ, Harmsen RE, Spronk PE. Clinical review: strict or loose glycemic control in critically ill patients-implementing best available evidence from randomized controlled trials. Crit Care 2010;14:223. 9. Finfer S. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283-97. 10. Brunkhorst F M, Engels C, Bloos F, et al. Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 2008;358:125-39. 11. Investigators-NICE SUGAR Initiative. Intensive versus conventional glucose control in critically ill patients. N Engl J Med 2009;360:1283-97. 12. Reeds DN. Nutrition support in the obese, diabetic patient: the role of hypocaloric feeding. Curr Opin Gastroenterol 2009;25:151-4. 13. Burke PA, Young LS, Bistrian BR. Metabolic vs. nutrition support: a hypothesis. J Parenter Enteral Nutr. 2010;34:546-8. 14. Casaer MP, Mesotten D, Hermans G, et al. Early vs. late parenteral nutrition in critically ill adults. N Engl J Med 2011;365: 506-17. 15. Port AM, Apovian C. Metabolic support of the obese in- tensive care unit patient: a current prospective. Curr Opin Clin Nutr Metab Care 2010;13:184-91. 16. Raza N, Benotti P, Still CD. Nutrition in critically ill obese patients. Crit Care Clin 2010;26:671-8. 17. Boitano M. Hypocaloric feeding in the critically ill. Nutr Clin Pract 2006;21:617-22. 18. Martindale RG, Maerz LL. Management of perioperative nutrition support. Curr Opin Crit Care 2006;12:290-4. 19. Dickerson RN, Boschert K, Kudsk KA, et al. Hypocaloric enteral tube feeding in critically ill obese patients. Nutrition 2002; 18:241-6. 20. Jimenez FJJ, Leyba C, Jimenez LMJ, et al. Study of hypo- caloric peripheral parenteral nutrition in postoperative patients (European Project). Clin Nutr 1995;14:88-96. 21. Stapleton RD, Jones N, Heyland DK. Feeding critically ill patients: what is the optimal amount of energy? Crit Care Med 2007;35:S535-40. 22. Kauffmann RM, Hayes R, Jenkins JM, et al. Provision of balanced nutrition protects against hypoglycemia in the critically ill surgical patient. J Parenter Enteral Nutr 2011;45:686-94. 23. White NH, Skor D, Santiago JV. Practical closed-loop in- sulin delivery. A system for maintenance of overnight euglycemia and the calculation of basal insulin requirements in insulin- dependent diabetics. Ann Intern Med 1982;97:210-3. 24. Bode BW, Braithwaite S, Steed RD, et al. Intravenous in- sulin infusion therapy: indications, methods, and transition to subcutaneous insulin therapy. Endocr Pract 2004;10:71-80. 25. Kovatchev BP, Cox D, Gonder-Frederick LA, et al. Sym- metrization of the blood glucose measurement scale and its ap- plications. Diabetes Care 1997;20:1655-8. 26. PJ, H. Proc Fifth Berkeley Symposium Math Stat, 1967;1: 221-33. 27. White H. Maximum likelihood estimation of misspecified models. Econometrica 1982;50:1-25. 28. Harrell F E. Regression Modeling Strategies with Applica- tions to Linear Models, Logistic Regression, and Survival Analy- sis. 1st ed. New York, NY: Springer-Verlag; 2001. 29. Mowery NT, Dossett L, Gunter OL, et al. Duration on protocol predicts severe hypoglycemia while on intensive insulin therapy in the surgically critically ill population. World J Surg 2012;26:270-7. 30. Vriesendorp TM, van Santen S, DeVries H, et al. Predis- posing factors for hypoglycemia in the intensive care unit. Crit Care Med 2006;34:96-101. 31. Arabi YM, Tamim H, Rishu AH. Hypoglycemia with in- tensive insulin therapy in critically ill patients: predisposing factors and association with mortality. Crit Care Med 2009;37:2536-42. 32. Dickerson RN, Hamilton L, Connor K, et al. Increased hypoglycemia associated with renal failure during continuous in- travenous insulin infusion and specialized nutritional support. Nutrition 2011;27:766-72. RONDI M. KAUFFMANN, M.D.,* RACHEL M. HAYES, PH.D.,[dagger] AMANDA H. VanLAEKEN, M.S.,[double dagger] PATRICK R. NORRIS, PH.D.,* JOSE J. DIAZ, M.D.,§ ADDISON K. MAY, M.D.,* BRYAN R. COLLIER, D.O.[double dagger] From the *Division of Trauma and Surgical Critical Care, Department of Surgery, and the [dagger]Informatics Center, Information Technology Integration, Vanderbilt University Medical Center, Nashville, Tennessee; [double dagger]Carilion Roanoake Memorial Hospital, Roanoke, Virginia; and the §Department of Trauma, SHOCK Trauma Center, University of Maryland, Baltimore, Maryland Presented in poster format at the 33rd Congress of Clinical Nu- trition and Metabolism, Gothenburg, Sweden, September 3-6, 2011. Address correspondence and reprint requests to Bryan R. Collier, D.O., Carilion Roanoke Memorial Hospital/Virginia Tech Carilion School of Medicine, 1906 Belleview Avenue, Roanoke, VA 24014. E-mail: [email protected]. (c) 2014 Southeastern Surgical Congress |