Software-Guided insulin therapy

From Clinfowiki
Jump to: navigation, search

Intensive control of glycemic control has been shown to improve clinical outcomes. As part of clinical decision support system (CDSS), various models of software-guided systems have been developed. Studies show that Software-guided algorithms have been shown to be effective in achieving tight glycemic control in the critical illness.

Background

Large randomized control studies showed that regulation of blood glucose levels to <110mg/dl using intravenous insulin protocol improved clinical outcomes. [2] Other studies suggest target glucose levels less than 140-180mg/dl. [5] The clinical outcomes include improved mortality rates, decrease in sepsis, need for dialysis, duration of ventilation support, and overall ICU stay. Other studies also demonstrated strict glycemic control is beneficial in other settings including diabetic ketoacidosis, non-ketotic hyperosmolar state, myocardial infarction, cardiogenic shock, postoperative period complications. [3]

Achieving the goals of strict glycemic control requires extensive nursing efforts such as frequent glucose monitoring, implementing complex protocols. Many efforts were sought to streamline the process using a protocol. The NICE-SUGAR trial was a paper-based protocol which was six pages long and involved 56 “action codes.”[4] The trial was not successful in achieving a consensus of target glucose range or suggesting a valid and effective model for protocol based insulin therapy. There were many limitations such as adherence rate(less than 50%), errors, insulin dosing protocol violations.

There has been a growing effort in developing a software algorithms and computerized protocols to improve the current protocols.

Software-guided insulin therapy is intended to evaluate the current and cumulative patient blood glucose values, drive the blood glucose level towards a predetermined target range. The system will then titrate rate of insulin infusion, subcutaneous injection for the purpose of maintaining the patient's blood glucose level in a target range.

Goals of glycemic control using software-guided insulin therapy

Ease of use, minimal burden of end-users, automated data entry, improved adherence rate, proven efficacy. Targeted effects would include a quick improvement in blood glucose level, maintain glucose levels in a target range with minimal variations while avoiding hypoglycemia, integrated into existing hospital systems.

Types of Software algorithms

Heuristic model


Converting paper-based protocols into a software program. Improved adherence of the protocol, reduced errors compared to paper-based protocols. There are limitations due to the simplistic approach of once a day calculation without taking into daily variable glucose levels which also is a factor in the care of patients. [6]

Proportional-integral-derivative (PID) model


Uses previous blood glucose values to titrate insulin administration using a dynamic multiplier reflecting the changes afterwards to insulin sensitivity. These algorithms allow real-time adjustments but may require numerous iterations to achieve target glucose levels (18 or higher). The PID controls continuously make small adjustments and becomes more accurate as data accumulates. Many commercial software programs use this model. Glucommander (Glytec), GlucoCare (Pronia Medical systems), EndoTool (Hospira/Monarch medical technologies), GlucoStabilizer(Alere Informatics Solutions) are used as protocols with studies showing efficacy. [7-10]

Glucommander

Glucommander is currently one of the most utilized and studied software. Originally started by a simple control system from an article published by White et al., the idea was a collection of an order set relying a formula:

(blood glucose − 60) × multiplier = insulin dose/h.

Multipliers represent a wide variety of patients with insulin resistance. Multipliers would change throughout the course of therapy.

It was later programmed into computerized system after 2000 with an insulin pump where the glucose levels were entered and the rate of insulin was determined by the rate of change of glucose levels. The computer notifies nurse when the next glucose value is needed, which may be between 20-120 minutes.

Many studies showed that glucommander effectively improved hyperglycemia. [11, 12]

GlucoCare

GlucoCare is a computerized calculator based on the “Yale Insulin Infusion Protocol” [13]. The Yale insulin infusion protocol is a complex insulin infusion determining protocol that was determined by the rate of blood glucose change. The original manual protocol was complex and needed to be followed meticulously to not be driven off of protocol. GlucoCare was studied with multiple targets of blood glucose levels and was successful in eliminating hypoglycemic episodes that was presented in the original Yale Insulin Infusion Protocol. [14]

EndoTool

The exact mechanism that this software is based on is not published and the data is currently not sufficient in proving its efficacy. Benefit of using EndoTool would be the ability to incorporate into existing EHR.

GlucoStabilizer

GlucoStabilizer also uses similar calculation and uses multiplier. GlucoStabilizer was studied on both insulin infusion and subcutaneous insulin. [15]

Model predictive controls (MPCs)


Incorporate patient specific parameters such as age, sensitivity, diagnosis of diabetes to predict patient’s response to insulin therapy and changes in blood glucose levels. Increase in parameter measurements can lead to improvement in accuracy and decrease sampling rate by up to 50%. Current models are Stochastic Targeted glycemic control and Space GlucoseControl (B. Braun)

Stochastic targeted (STAR) glycemic control

The Stochastic targeted glycemic control uses a simulated fundamental metabolic dynamic model called Insulin-Nutrition-Glucose(ICING) metabolic model.[16] It takes account intra- and inter- patient variability by encompassing many physiologic factors such as endogenous insulin production, insulin clearance, insulin sensitivity etc. Studies show that STAR glycemic control protocol has provided tight blood glucose control with a maximum 5% expected risk of light hypoglycemia. Clinical workload was reduced up to 30%. [17]

Multiple pilot clinical trials has been proposed and validated but no commercial software has been developed at this point.

Space GlucoseControl

The Space GlucoseControl system is a clinical decision support system that is integrated in infusion pump platforms used in critical care environment. The Space GlucoseControl module attached to the infusion pump contains a self-learning eMPC algorithm. The eMPC algorithm determines insulin rate by the blood glucose monitoring and actual rate of enteral, parenteral nutrition.

The Space GlucoseControl system has been well researched in multiple trials and available commercially in Europe. [18]


References


  1. White NH, Skor D, Santiago JV: Practical closed-loop insulin delivery: a system for the maintenance of overnight euglycemia and the calculation of basal insulin requirements in insulin-dependent diabetics. Ann Intern Med 97:210–213, 1982
  2. Van den Berghe G, Wouters P, Weekers F, Intensive insulin therapy in the critically ill patients. N Engl J Med 345: 1359–1367, 2001
  3. Malmberg K, the DIGAMI Study Group: Prospective randomized study of intensive insulin treatment on long term survival after acute myocardial infarction in patients with diabetes mellitus. BMJ 314:1512–1515, 1997
  4. NICE-SUGAR Study Investigators, Finfer S, Chittock DR, Su SY. Intensive versus conventional glucose control in critically ill patients. N Engl J Med. 2009;360(13):1283–97
  5. Moghissi ES, Korytkowski MT, DiNardo M, American Association of Clinical Endocrinologists; American Diabetes Association. American Association of Clinical Endocrinologists and American Diabetes Association consensus statement on inpatient glycemic control. Endocr Pract. 2009;15(4):353–69.
  6. Thomas AN, Marchant AE, Ogden MC, Implementation of a tight glycaemic control protocol using a web-based insulin dose calculator. Anaesthesia. 2005;60(11):1093–10
  7. Goldberg PA, Siegel MD, Sherwin RS, Implementation of a safe and effective insulin infusion protocol in a medical intensive care unit. Diabetes Care. 2004;27(2):461–7.
  8. Shetty S, Inzucchi SE, Goldberg PA. Adapting to the new consensus guidelines for managing hyperglycemia during critical illness: the updated Yale insulin infusion protocol. Endocr Pract. 2012;18(3):363–70.
  9. Saager L, Collins GL, Burnside B. A randomized study in diabetic patients undergoing cardiac surgery comparing computer-guided glucose management with a standard sliding scale protocol. J Cardiothorac Vasc Anesth. 2008;22(3):377– 82.
  10. Juneja R, Roudebush C, Kumar N. Utilization of a computerized intravenous insulin infusion program to control blood glucose in the intensive care unit. Diabetes Technol Ther. 2007
  11. Joseph Aloi, Bruce W. Bode, Jagdeesh Ullal. Comparison of an Electronic Glycemic Management System Versus Provider-Managed Subcutaneous Basal Bolus Insulin Therapy in the Hospital Setting, J Diabetes Sci Technol. 2017 Jan; 11(1): 12–16. Published online 2016 Sep 25
  12. Bode B, Clarke JG, Johnson J. Use of Decision Support Software to Titrate Multiple Daily Injections Yielded Sustained A1c Reductions After 1 Year. J Diabetes Sci Technol. 2018 Jan;12(1):124-128. doi: 10.1177/1932296817747886. Epub 2017 Dec 17.
  13. inpatient.aace.com/sites/all/files/Yale_IIP_MICU120-160_2011.pdf
  14. Michael R. Marvin, Minimization of Hypoglycemia as an Adverse Event During Insulin Infusion: Further Refinement of the Yale Protocol Diabetes Technol Ther. 2016 Aug 1; 18(8): 480–486.Published online 2016 Aug 1.
  15. Rattan Juneja, Corbin P Roudebush, Stanley A Nasraway. Computerized intensive insulin dosing can mitigate hypoglycemia and achieve tight glycemic control when glucose measurement is performed frequently and on time Crit Care. 2009; 13(5): R163. Published online 2009 Oct 12.
  16. Lin J, Razak NN, Pretty CG A physiological Intensive Control Insulin-Nutrition-Glucose (ICING) model validated in critically ill patients. Comput Methods Programs Biomed. 2011 May;102(2):192-205
  17. Evans A, Le Compte A, Tan CS. J Diabetes Sci Technol. Stochastic targeted (STAR) glycemic control: design, safety, and performance. 2012 Jan 1;6(1):102-15.
  18. Jan Blaha, Barbara Barteczko-Grajek, Pawel Berezowicz. Space GlucoseControl system for blood glucose control in intensive care patients - a European multicentre observational study. BMC Anesthesiol. 2016; 16: 8.

Submitted by Jinsoo Chang