Tight Glycemic Control: How Sweet It Is! : Tight Glycemic Control: How Sweet It Is! Virginia Point of Care Coordinators April 22, 2005


Disclosures: Disclosures State faculty for the Surgical Infection Prevention Initiative No financial or other conflicts of interest Claudette Dalton, M.D.


“What, me worry?”: “What, me worry?” Surgical Infection Prevention Initiative (SIP) /Surgical Care Improvement Project (SCIP) Literature JCAHO/ CMS/ ACS/ ASA/ ICU/ POC/ NQF/ IOM/ QI/ PI/ VHQC ICU standard of care ? General standard of care


Core Knowledge Needed: Core Knowledge Needed Impact on outcomes Target BGs/Protocols Difference between insulins and how they are given How/when to test consistently Treatment and prevention of hypoglycemia Documentation pathways Terms and definitions


Definitions: Definitions Hyperglycemia is a blood sugar over 110 in a fasting patient and over 125 in a patient who has eaten. Hypoglycemia—40-70 mg/Dl Point of care testing Immediate results that alter management Diabetes mellitus Types I and II, gestational Hyperglycemia Steroids Stress Other meds


Conditions that Predispose to Hypoglycemia: Conditions that Predispose to Hypoglycemia Advanced age Decreased oral intake Chronic renal failure Liver disease Use of Beta blockers Mistiming of meals in relationship to insulin dosing Infrequent or missed monitoring Lack of coordination with transportation and floor Knowledge deficits by providers Unreadable, unusual or convoluted orders Difficult to follow protocols Physician insisting on different protocol


What is the evidence? : What is the evidence? Risk of microvascular complications Renal and retinal disease Diabetes Control and Complications Trial Risk of macrovascular complications CAD and stroke Capes SE. Stroke 2001; 32:2427 DIGAMI and Malmberg K. Circulation. 1999. 99:2626-2632 Risk of mortality Risk of infections Cost of care ACE and AACE recommendations


Unanswered Questions: Unanswered Questions What is “optimal control”? How long does the patient need to be in good control? Can we take “tight control” too far? What is the role of lipids in glucose control? Do we need to aggressively treat other medical conditions at the same time?


The Role of Blood Sugar in Infections: The Role of Blood Sugar in Infections Poor wound healing in general/many already colonized Deoxyglucose inhibits glycolytic metabolism which generates energy for superoxide production No absolute Km identified but glucose level proportional to neutrophil activity Granulocyte functions—improve when glucose control is good—i.e. <200mg/dL Adherence Delays chemotaxis Impairs phagocytosis Decreased bacteriocidal activity


Other DM Complications in Surgery: Other DM Complications in Surgery Cellular immunity Decreased complement fixation Collagen—increased collagenase activity Role of microvascular damage 34% of insulin dependent diabetics are colonized with s. aureus Cardiac cellular function


Vanderbilt StudyLatham R. et.al. Infact Control Hosp Epidemiol. 2001; 22:607-612: Vanderbilt Study Latham R. et.al. Infact Control Hosp Epidemiol. 2001; 22:607-612 Prospective, 1044 CABG and valve ops 6% had undiagnosed diabetes SSI pts.—62% of known diabetics had hyperglycemia/37% of non-DM patients Dx of DM associated with 2.7X risk for SSI Rate of SSIs correlated with degree of hyperglycemia Hyperglycemia during periop is independent risk factor


Vanderbilt, con’t: Vanderbilt, con’t Similar to other studies, 6% were undiagnosed diabetics 19% in this study had abnormal HgbA1c and another 11% had glucose >200 But Hgb A1c did not correlate with SSIs Still, suggest that screening with HgbA1c for diagnosis of DM is cost effective if therapy is initiated


Perioperative Glucose Control: Perioperative Glucose Control 1,000 cardiothoracic surgery patients Diabetics and non-diabetics with hyperglycemia Patients with a blood sugar > 300 mg/dL during or within 48 hours of surgery had more than 3X the likelihood of a wound infection! Latham R, et al. Infect Control Hosp Epidemiol. 2001.


What factor makes the difference?: What factor makes the difference? Patients may be undiagnosed (4.2%--or higher!) Most infections when glucose level is >200 mg/dL Risk same if glucose high anytime in first 48 hours Hyperglycemia doubles risk—2—2.7X 20mg/Dl increase = 30% increase risk of death May directly affect cardiac cellular function Can be stress or medication induced Capes SE. Lancet. 2000. 773-778 and Clement S. Diabetes Care. 2004. 27:553-591


Ain’t No Mountain High Enough…: Ain’t No Mountain High Enough… Enormous percentage of our patients are diabetic Another percentage are undiagnosed or hyperglycemic from other causes Adding nutrition and crisis management Source of blood Timing of testing Tests used


Ain’t no Valley Low Enough…: Ain’t no Valley Low Enough… Hypoglycemia is a blood glucose of 40-70 Institution dependent Cause seizures, brain death if too low Anesthesia and sedation block usual symptoms Blood source Timing of testing Tests used


Is There a River Wide Enough?: Is There a River Wide Enough? Who is the crew and who is the coxswain? Untraditional looking crew Nutritionist, Pharmacist, QI/PI, managers Lab, nurses, doctors, educators Constant educational needs Policies Which protocol? Point of care testing and decision making Patients go through multiple units while in the hospital—transitions are trouble points Costs/equipment


Protocol, protocol, who has a protocol? : Protocol, protocol, who has a protocol? Portland, van den Berghe, Yale, home-grown? Elements to look for While NPO, when feeding, when crisis Timing of doses/testing Subcu vs IV—continuous (CII) vs. bolus Different protocol for night shift, for sicker patients, for iconoclastic docs? Start higher to avoid going lower—how low is too low? How many get hypoglycemic on each protocol? KISS –or not? Education, re-education and more education Requires an IV for most of the protocols


UVA Protocol: UVA Protocol ICU generated 95 is ICU target, <175 is SIP target, 125-175 is general floor target No subcu Tests q 1h till stable X 2, then q 2h Hypoglycemia at 80 mg/Dl--!!! This is very unusual Capillary unless needs checking, then venous—not sure why we do not use arterial in ICU


UVA SIP Glucose Compliance: UVA SIP Glucose Compliance


PortlandFurnary. J Thorac CardiovascSurg, 2003; 125: 1007-21: Portland Furnary. J Thorac CardiovascSurg, 2003; 125: 1007-21 http://www.starwood.com Endocr Pract 2004; 10: 21-33 Tests q .5-2 hrs Continuous IV only Avg. 3 day Blood glucose 13,649 patients since 1987—prospective interventional 1.5% hypoglycemia rate (60 mg/Dl)


van den BergheNEJM 2001. 345:1359-1367: van den Berghe NEJM 2001. 345:1359-1367 1548 SICU patients. Randomized, prospective, controlled. IV insulin to maintain between 80 and 110 mg/Dl Relatively short Measures q 2h until goal, then q 4h Hypoglycemia at 60 No additional protocols for adding nutrition, crises, weaning


Yale: Yale Goldberg PA, et al. Diabetes Care. 2004; 27:461-467 Current BG leads you to table. Hourly rate of change is guide. Nomogram. Complicated. Target is 100-139 mg/Dl. Very little hypoglycemia Mean time to target is 4.6 hrs. Median is 9 hrs. Protocol rated “easy”, no additions for nutrition, weaning, crises.


Other protocols: Other protocols Markovitz—Endocr Pract. 2002; 8:10-16 Has default algorithm Testing frequency lowers as stabilizes Hourly rate=hourly maintenance rate +(blood glucose-150)/ISF Cut off is 100 UNC—not published yet Target of 80-110. Has no hypoglycemia cut-off. Florida Hospital—not published, looks like blend of Markovitz format and Portland amounts Glucommander


Free Form Protocols-Basic Concepts: Free Form Protocols-Basic Concepts Usual start dose is 0.15 u/kg Continuous IV weaned to bolus weaned to usual Think Basal/Nutritional/Correction (Crisis) as three distinct levels with different needs Basal needs long acting agent like glargine Nutritional needs medium acting at 1 unit/10 gms of CHO Crisis/correctional needs short acting like Lispro or Aspartine


More Basics to Keep in Mind: More Basics to Keep in Mind Use regular insulin or NPH in drips Regular insulin at doses of 0.5-1 unit/ml Infuse at 0.1 unit dose increments Use IV fluids with glucose—usually D5 Monitor potassium Have D50 available and oral CHO also.


But is CII cheaper than SQ? : But is CII cheaper than SQ? Direct and indirect costs for 3 days of q 4h SQ=$32/pt Costs of 3 days of Cont IV infusion with q 1-2 h test =$170/pt ($138 difference) Cost of DSWI =$2613/pt + $2081 for 1.8 additional days $4694-138 =$4556/pt or $4,556,000 per 1000 CABGs US Hospital savings = 103K CABGs =$469 million/yr


Point of Care Testing: Point of Care Testing Essential Timing is crucial Which blood source? Urine? Same over time? Sensitivity vs. specificity What interferes with the test you use? No way to get trends at this time What would you want in a testor that you do not have now…? Who needs to be involved? What skills do they need or bring to share?


More on POC Testing: More on POC Testing Bedside monitoring vs. central lab Does the person doing the test matter? Self-monitoring? Cost, accuracy, accountability Will we live long enough to see a non-invasive bedside monitor? Wireless? Ketones, albumin, HbA1c, glycated serum proteins-better than blood glucose?


The Pieces You Need: The Pieces You Need Know the literature and other rationales Have a credible champion The right protocol Forms, policies and order sets The right team Enough equipment Strong Quality Improvement department


What to Do with the Pieces: What to Do with the Pieces Start with one unit Keep all data in one place Solid communication system Accurate test administered by trained professionals Timely changes in treatment Start high, move lower Never stop educating Have a safety plan Consider special circumstances


Data that may help you…: Data that may help you… Knowing what percentage of patients are diabetic—and guesstimating percent of unrecognized hyperglycemic patients Literature Knowing what surgical infection rate is Estimating cost to your institution in terms of: Mortality LOS Financial Costs


Thanks…: Thanks… Questions, comments, suggestions? ced2t@virginia.edu