Economics and Quality Care: The Case for Patient Blood Management Why We Should Avoid Transfusions Irwin Gross, M.D., Medical Director Transfusion Services, Eastern Maine Medical Center Disclosures • Medical Advisory Board – Strategic Healthcare Group A little about Eastern Maine Medical Center • 370 bed community and tertiary care hospital in rural Maine • Large hospitalist service • High risk obstetrics • Trauma center, Level 2 • Dialysis center • Family Practice residency • Cardiac surgery program – Approx. 450 cases/yr. • Active heme/onc service – 10 oncologists • No transplant surgery Patient Blood Management (PBM) The timely application of evidence-based medical and surgical concepts designed to maintain hemoglobin concentration, optimize hemostasis and minimize blood loss in an effort to improve patient outcome. What is Patient Blood Management: The Three Pillars Optimize erythropoieis Minimize bleeding & blood loss © Axel Hofmann/Shannon Farmer – SHEF Meeting Perth August 2010 Harness & optimize physiologic tolerance of anemia Can Patient Blood Management Reduce the Need for Transfusions? Red Cell Units Transfused FY 1994 – FY 2010 9470 Patients Transfused: FY 1994 - 2011 Patients Transfused 3000 2684 2422 2393 2468 2500 2773 2843 2474 2263 2157 1998 2000 1846 1815 1686 1550 1737 1608 1534 1606 1500 1000 500 FY 94 FY 95 FY 96 FY 97 FY 98 FY 99 FY 00 FY 01 FY 02 FY 03 FY 04 FY 05 FY 06 FY 07 FY 08 FY 09 FY 10 FY 11 0 Transfusion Rates All Cases: CABG, Valve, CABG/Valve April 2008 – March 2011 Transfusion rate in 2006: 48% 23% 14% Cardiac Surgery and Transfusions • With reduction in transfusion rate from 48% to approximately 20% , there was a reduction in: – – – – – – Perioperative AMI New onset renal failure Perioperative infection Stroke Length of stay No change in mortality Transfusion Rate: Orthopedics Blood Transfusion: A Growth Industry Copyrights 2006 – 2011 Strategic Healthcare Group LLC. All Rights Reserved Our Demographic Challenge • Rapidly increasing percentage of population over the age of 65 and decreasing percentage of population between 17 and 65 – Decrease in eligible donors – Increase in population that are transfusion recipients – Demand may exceed supply in the near future Transfusion is expensive! Between $832 and $1284 per unit in the EMHS System Blood and Blood Components: Cost Savings – All Components (EMMC) • Total blood acquisition costs in FY ‘06 were $3,200,000 • Cost savings compared to base year, FY ’06* – – – – – – FY ’07 FY ’08 FY ’09 FY ’10 Total (Acquisition) Total (ABC) $ 850,000 $ 1,400,000 $ 1,600,000 $ 1,550,000 $ 5,400,000 $ 17,280,000 ** * No change in per unit cost from blood supplier from 2007 – 2010 * Using ABC 3.2 times acquisition cost The Macroeconomic Burden of Transfusion • Retrospective cohort study by American College of Medical Quality • Evaluated all 38.7 million hospitalizations in U.S. in 2004 • Adjusted for age, gender, comorbidities, admission type and DRG • Charges per transfused patient were $17,194 more than charges for nontransfused, matched patient • Total excess financial burden: $40 billion Am J Med Qual 2010;25:289-296 Is Blood Utilization in the U.S. Optimal? Copyright 2006 – 2011 Strategic Healthcare Group LLC. All Rights Reserved Observed Variation in Hospital-Specific Transfusion Rates for Primary Isolated CABG Surgery With Cardiopulmonary Bypass During 2008 (N = 798 Sites) Bennett-Guerrero, E. et al. JAMA 2010;304:1568-1575. Are Transfusions Safe: Risk vs. Benefit • The risks of transfusion are increasingly well defined and extend far beyond concerns about disease transmission (e.g. HIV, hepatitis) • The benefits of red cell transfusion, except in severe hemorrhage, have never been proven • While anemia is associated with poorer outcomes, red cell transfusion to treat anemia does not appear to improve outcomes Impact of Transfusion on Patient Outcomes Author (year) Population Leal-Novel et al (2001) Cardiac surgery Wu et al (2001) Elderly with MI Engoren et al (2002) Cardiac surgery Vincent et al (2002) Critically ill Malone et al (2003) Trauma Corwin et al (2004) Critically ill Dunne et al (2004) Trauma Innerhofer et al (2005) Orthopedic surgery Weber et al (2005) Orthopedic surgery Koch et al (2006) Cardiac surgery n 738 Impact of Blood Transfusion - Higher mortality rate - Longer ICU stay 78,974 - Lower 30-day mortality (if admission Hct<33*) - Higher 30-day mortality (if admission Hct>36) 1,915 - Higher 5-year mortality rate 3,534 - Higher 14- and 28-day mortality rate 15,534 - Higher mortality rate - More ICU admission - Longer ICU/hospital stay 4,892 - Higher mortality rate - Longer length of stay - More number of complications 9,539 - Higher mortality rate - More ICU admission - More SIRS 308 - More infections 695 - Longer time to ambulation - Longer length of stay 5,841 - More AF Impact of Transfusion on Patient Outcomes Author (year) Koch et al (2005) Population Cardiac surgery Murphy et al (2007) Cardiac surgery Surgenor et al (2009) Cardiac surgery Pedersen et al (2009) Total hip replacement Nikolsky et al (2009) PCI after MI van Straten et al (2010) Cardiac surgery D’Ayala et al (2010) Lower extremity amputation O’Keeffe et al (2010) Lower extremity revascularization Veenith et al (2010) Elderly undergoing cardiac surgery n Impact of Blood Transfusion 11,963 - Higher postop mortality rate - Higher postop morbidity rate (RF, prolonged ventilatory support, serious infection, cardiac complications, & neurologic events) 8,598 - Higher mortality rate - More ischemic complications - More infectious complications 3,254 - Decreased survival after cardiac surgery 4,508 - Higher mortality rate - More pneumonia 2,060 - Higher 30-day and 1-year mortality rate 10,435 - Worse early survival 300 - More postop adverse events - Longer ICU/hospital stay 8,799 - Higher mortality rate - More pulmonary complications - More infectious complications 874 - Higher mortality rate - Longer ICU/hospital stay Transfusion and Outcomes • Following transfusion: – In all of the studies, Hb increased – Is this a useful metric of clinical efficacy? – In 79% of the studies, DO2 increased – In 16% of the studies, VO2 increased – In none of the studies, did ischemia (as measured by lactate) improve – In all of the studies, transfused patients had poorer outcomes Efficacy of RBC Transfusion in the Critically Ill • Meta-analysis - 45 observational studies of 272,596 patients • Included surgical (trauma, general, ortho, neuro, and cardiac) and general ICU patients • 42 of 45 studies: risks outweighed benefits of transfusion; risk neutral in 2 studies • Transfusion an independent risk factor for increased: • • • • Mortality Infection Multi-organ dysfunction ARDS Marik, et al CCM 2008;36:2667-2675 Adverse Effects of Allogeneic Blood • Storage Lesion • TRIM • TRALI: 1 in 2,000 transfusions; primarily plasma rich products; up to 20% mortality • Hemolytic transfusion rxns: (1:4,000) • Bacterial contamination (Plts.-1:75,000) • HIV, HCV, HBV • Febrile and allergic reactions: 1-2% • Other: ARDS, TA-GVH, SIRS, TACO, TAHA, unknown viruses Adverse effects of RBC transfusion contrasted with other risks.Risk is depicted on a logarithmic scale. Relative risk of: Storage lesion, TRIM, SIRS, etc? Carson J L et al. Ann Intern Med doi:10.1059/0003-4819156-12-201206190-00429 ©2012 by American College of Physicians Storage Lesion: Tissue Oxygenation and Red Cell Transfusions Copyright 2006 – 2011 Strategic Healthcare Group LLC. All Rights Reserved Phase Contrast Video – Single file RBCs Click to play Cytoscan – Pre & Post Transfusion Click to play RCTs to Evaluate Red Cell Storage • Age of Blood Evaluation (ABLE): comparing 8 day old blood vs. standard issue (2-42 days) in ICU patients – Outcome: 90 day all-cause mortality • Age of Red Blood Cells in Premature Infants (ARIPI): 8 day old blood vs. standard issue – Outcome: 90 day all-cause mortality and organ dysfunction • Red Cell Storage and Duration and Outcomes in Cardiac Surgery: < 14 day vs. > 20 day old blood – Outcome: Post-op mortality • Red Cell Storage Duration Study (RECESS): Cardiac surgery patients randomized to < 10 days vs. > 21 day old blood – Outcome: Change in Composite Multi-organ Dysfunction Score (MODS) Adverse Effects of Allogeneic Blood • “Stored RBCs resulted in significantly malperfused and underoxygenated microvasculature “ – Decreased functional capillary density • May contribute to multi-organ failure • Applies to stored autologous blood Tsai, A.G. et al. Transfusion 2004;44:1626-34 Transfusion Related Immunomodulation (TRIM) Transfusion and Cadaver Renal Allograft Survival Transfusion Related Immunomodulation (TRIM) • Improved renal allograft survival • Increase in nosocomial infection rates leading to increased LOS, resource consumption, total hospital costs • Increased cancer recurrence rates in transfused patients • Increased long-term mortality in CABG • A linear dose-response curve of adverse clinical effects is well documented in large studies using multivariate analysis Nosocomial Infections And Transfusion Nosocomial Infection Rates in Critically Ill Patients P < .0001 16 Percent of Patients • Prospective cohort study • Patients stratified by probability of survival into quartiles • Lowest quartile for survival: transfusion had no impact on infections • Highest quartile for survival: transfusion resulted in significant increase on infections 14.3 All Patients Transfused Patients Non-transfused Patients 12 8 7.5 5.8 4 0 157/2,085 61/428 96/1,657 Taylor RW, et al. Crit Care Med. 2006;34:2302-08. Small Amounts of Intraoperative Blood Loss Heralds Worse Postoperative Outcome • Database study of NSQIP data with propensity–score matching • 8728 nonvascular thoracic operations • 6.6 % received one or two units during surgery • After propensity adjustment – Increased pulmonary complications and LOS – Increased systemic sepsis and wound infections – Composite morbidity 43% vs.32% Ferraris, et al. Ann Thorac Surg 2011 S. aureus and Iron Surface Determinant (IsdB) . Lowy FD. N Engl J Med 2011;364:1987-1990. Transfusion and Mortality Association Between Blood Transfusion and Risk of Death Marik, et al CCM 2008;36:2667-2675 Transfusion and Mortality in CABG • Retrospective study of long-term outcome in 1,915 patients after primary CABG • Excluded patients who died within 30 days of surgery • Transfused patients (546) were matched by propensity score (age, gender, size, LOS, perfusion time and STS risk) with pts. not transfused and 5-year mortality compared • Adjusted 5-year mortality 70% higher in transfused group (p<0.001) Engoren et al, Ann Thorac Surg 2002;74:1180-6 Mortality and Transfusion-Cardiac Surgery Fig 3. Kaplan-Meier estimates of survival based on equal propensity scores of any transfusion (XFN) versus no transfusion (No XFN). (CABG = coronary artery bypass grafting Engoren et al, Ann Thorac Surg 74:1180-6, 2002 Transfusion and ARDS Association Between ARDS and Transfusion Marik, et al CCM 2008;36:2667-2675 Transfusions in Acute Coronary Syndrome Blood Transfusion and Clinical Outcomes in Patients with ACS • Retrospective analysis of 24,112 patients from GUSTO IIb, PURSUIT, and PARAGON B trials • Main outcome: 30 day mortality • Data adjusted for baseline characteristics: bleeding, transfusion propensity, nadir hematocrit • Transfusion was associated with a hazard ratio for death of 3.94 (3.26 – 4.25) Rao, et al. JAMA, Oct. 6, 2004, Vol 292, No. 13. pp. 1555 -1562 Transfusion and Cancer Survival Influence of Transfusion on Outcome in Pancreatic Cancer • 67 patients underwent pancreaticoduodenectomy for cure • 25 patients received > 3 units • No difference in tumor size, stage, histology compared to group receiving < 2 units • Median and cumulative 3 year and 5 year survival worse with > 3 units – 5 year survival 68.9% vs. 30.2% World J Surg, 2008 Influence of Transfusion on Colorectal Cancer Recurrence • Cochrane meta-analysis involving 12,127 patients • Evaluated role of transfusion in colorectal cancer recurrence • Overall OR for recurrence was 1.41 (95% CI 1.20-1.67) in transfused patients Amato, A, et al. Cochrane Database System Rev 2006;(1): CD005033 Lymphoma and Transfusion • Meta-analysis of 14 studies – Case control and cohort studies – Outcome reported as relative risk (RR) • Previous RBC transfusion associated with later development of nonHodgkin lymphoma, RR 1.2 (95%CI 1.07-1.35, P< .01) • In subset analysis, risk greatest for development of chronic lymphocytic leukemia/small lymphocytic lymphoma Castillo, JJ, et al. Blood. 2010;116(16):2897-2907) Radiation Therapy, Transfusion and Outcomes: DAHANCA 7 Study • Evaluated effect of transfusion on diseasespecific and overall survival in head/neck SCCa • Low hemoglobin group had poorer survival than high hemoglobin group • Low hgb group with transfusion showed trend toward poorer survival than low hgb group without transfusion Acta Oncologica, 2011; 50:1006-1014 Risk vs. Benefit “The issue (no longer) is whether or not blood transfusion is harmful, but the inflection point at which it is associated with more harm than benefit.” -Rao et al, JAMA 2005;292(13) Blood Product Transfusions The Traditional Concept • Blood products are an effective therapeutic intervention The New Concept • Transfusion of blood products is an undesirable outcome to be avoided Goodnough L.T. “Primum non nocere” (First, Do No Harm) The accumulating evidence suggests that, when we fail to prevent an avoidable transfusion, we are harming patients. When to Transfused Red Cells: The Transfusion Threshold Copyright 2006 – 2011 Strategic Healthcare Group LLC. All Rights Reserved Survival in Patients with Profound Anemia in the Absence of Transfusion Ford, et al, Blood 2005 106: Abstract 949 Hb Level (gm/dL) Survival 2.5-3.0 6/7 (85.7%) 3.1-4.0 9/11 (81.8%) 4.1-5.0 29/32 (90.6%) 5.1-6.0 30/30 (100%) 6.1-7.0 48/48 (100%) Overall 122/128 (95.3%) Transfusion in Critical Care: TRICC Trial A prospective, randomized trial comparing a restrictive red cell transfusion strategy vs. a liberal transfusion strategy Copyright 2006 – 2011 Strategic Healthcare Group LLC. All Rights Reserved Transfusion in Critical Care • 838 critically ill patients randomized to a restrictive (7-9 g/dl) or liberal (10-12 g/dl) transfusion strategy • Overall 30 day mortality similar (no transfusion benefit in liberal transfusion group) • Mortality rates significantly lower in restrictive transfusion group for those with APACHE score < 20 and those < age 55 • No difference in patients with clinically significant heart disease Hebert. NEJM 1999;340:409-17 Liberal or Restrictive Transfusion in High-Risk Patients after Hip Surgery • Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair (FOCUS) – RCT of 2016 patients with hip fracture requiring surgery – > age 50 (mean age 81.6 years) with cardiovascular disease (62.9%) or risk factors for CVD (37.1%) – Randomized to liberal (single unit if Hgb < 10 g/dL) vs. restrictive (single unit for Hgb < 8 g/dl or symptomatic from anemia Carson , JL et al. NEJM. December, 2011 Liberal or Restrictive Transfusion in High-Risk Patients after Hip Surgery • Results – Primary outcome: death or inability to walk 10 feet without assistance at 60 days – No difference between liberal and restrictive transfusion group – Low rate of cardiovascular adverse events in both groups – 65% fewer units transfused in restrictive group and half received no transfusion Carson , JL et al. NEJM. December, 2011 The “10/30 Rule” Should be Banished (also the 9/27 rule and 8/24 rule) No single hemoglobin threshold should be used to make a transfusion decision No rigid “transfusion trigger” based on hemoglobin alone But… for most patients, consideration of red cell transfusion should begin at hemoglobin of 7 gm/dl (hematocrit of 21%) % Transfusions Distribution of Pretransfusion Hematocrits 2006 Mean 24.3 % SD 4.28 2010 Mean 20.9 % SD 3.27 Red Cells Should Not be Ordered in “Pairs” For Red Cells: • In the absence of acute hemorrhage RBC’s should be given as single units • Each unit should be an independent clinical decision • Each unit represents an incremental increase in adverse consequences • Remember: transfusion of red cells may NOT increase oxygen delivery at a tissue level Napolitano - Crit Care Med 2009 Vol. 37, No. 12 Reduce Phlebotomy and Procedure-related Blood Losses • Meticulous surgical hemostasis • Meticulous hemostasis with “bedside” procedures • Minimize phlebotomy blood loss – Only necessary labs – Smallest sample practical Role of Intravenous Iron in Inpatients • Most inpatients have inflammation and functional iron deficiency (FID) – – – – Chronic inflammatory diseases Malignancy Perioperative Trauma • Oral iron is poorly tolerated, poorly absorbed, and not bioavailable • Intravenous iron may play a role in improving erythropoiesis – There may also be a limited role for ESAs in inpatients Conclusions • Red cell transfusions in patients who are not bleeding and have hemoglobin levels greater than 6-7 g/dL are generally associated with worse outcomes • The best way to optimize oxygen delivery to the tissues is not by transfusing stored RBCs; instead: – Optimize oxygenation – Optimize hemodynamics Conclusions • Even mild anemia contributes to all cause morbidity and mortality and should be evaluated and treated, when possible – Anemia management before elective surgery helps decrease transfusion rates • Intravenous iron replacement may be effective at increasing hemoglobin in patients with anemia of chronic inflammation and in patients with an acute inflammatory process • Limited role for ESAs in inpatients Anemia independently associated with increased: • morbidity • hospital length of stay • likelihood of transfusion • mortality Anemia Spahn DR. Anesthesiology 2010; 113(2) 1-14 Beattie WS, et al Anesthesiology 2009; 110(3) 574-81 Dunne JR, et al J Surg Res 2002; 102: 237-44 Shander A. Am J Med 2004; 116(7A) 58S-69S Bleeding associated with increased Triad of • Morbidity • ICU and hospital length of stay Independent RBC transfusion independently associated in a • Mortality dose-dependent relationship with increased: Risk Factors • Elective : ~0.1% •Morbidity for Blood loss • Subgroups: •ICU and hospital length of stay Adverse•Mortality • Vascular 5–8%& Transfusion • Up to 20% with severe bleeding Outcomes bleeding Beattie WA, et al Anesthesiology 2009 • Major organ damage 30–40% Murphy GJ, et al Circulation 2007 Causes Salim A, et al J Am Coll Surg 2008 • On average 75 – 90% local surgical interruption or Bernard AC, et al J Am Coll Surg 2008 vessel interruption • 10–25% acquired or congenital coagulopathy Shander A. Surgery 2007 © Axel Hofmann/Shannon Farmer – Hearnshaw SA, et al Aliment Pharmacol Ther 2010 SHEF Meeting Perth August 2010 Blair SD, et al Br J Surg 1986 So Why Practice Patient Blood Management? • Higher quality and safety: better patient care • Lower costs • “Stewardship” of the blood supply – Optimal use of blood – Efficient use of resources Patient Blood Management: The Three Pillars • • • • • • Pre-op anemia screening Refer for further evaluation if necessary ESAs Intravenous Iron Note: anemia is a contraindication for elective surgery Optimize erythropoieis • • • • • • • • Identify and manage bleeding risk and anticoagulants ANH Cell Salvage DDAVP TXA, Amicar Topical hemostatics Meticulous surgical hemostasis Avoid secondary hemorrhage Minimize phlebotomy Minimize bleeding & blood loss © Axel Hofmann/Shannon Farmer – SHEF Meeting Perth August 2010 • • Optimize hemodynamics Optimize ventilation and oxygenation Low hemoglobin threshold for transfusion Minimize oxygen consumption Avoid/treat infections promptly Harness & optimize • physiologic • tolerance of • anemia Before I go, a topic for another time • There is a substantial LACK of evidence that an increased PT/INR (at least < 2.0) can or should be used to make clinical decisions re: benefit of preprocedure frozen plasma • Avoid prophylactic plasma transfusions – No proven clinical benefit – Substantial risk of acute pulmonary reactions