ANEMIA MANAGEMENT PROTOCOLS IN THE CARE OF HEMODIALYSIS PATIENTS- EXAMINING PATIENT OUTCOMES by Sushila Saunders B S N , University of British Columbia, 1991 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN NURSING UNIVERSITY OF NORTHERN BRITISH COLUMBIA July 2010 ©Sushila Saunders, 2010 1*1 Library and Archives Canada Bibhotheque et Archives Canada Published Heritage Branch Direction du Patnmoine de I'edition 395 Wellington Street Ottawa ON K1A 0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre reference ISBN 978-0-494-75113-8 Our file Notre reference ISBN 978-0-494-75113-8 NOTICE AVIS The author has granted a nonexclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or noncommercial purposes, in microform, paper, electronic and/or any other formats L'auteur a accorde une licence non exclusive permettant a la Bibhotheque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par telecommunication ou par I'lntemet, preter, distnbuer et vendre des theses partout dans le monde, a des fins commerciales ou autres, sur support microforme, papier, electronique et/ou autres formats The author retains copyright ownership and moral rights in this thesis Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission L'auteur conserve la propnete du droit d'auteur et des droits moraux qui protege cette these Ni la these ni des extra its substantiels de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autonsation In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis Conformement a la lot canadienne sur la protection de la vie privee, quelques formulaires secondares ont ete enleves de cette these While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant l+l Canada ABSTRACT The role of hemodialysis nurses in renal anemia management has evolved through the implementation of nurse-driven protocols Using the Nursing Role Effectiveness Model, this case control study examined a nurse-driven renal anemia management protocol approach in contrast to a physician-driven approach to assess patient outcomes and costs of renal anemia management from two comparable hemodialysis centres Both protocol and control groups achieved target hemoglobin levels between 110-120 g/L In the protocol group, 75% of patients reached TSAT >20% versus 25% in the control group Iron costs were $17,000 higher in the control group over the study period The protocol group used more epoetin alfa, trending upwards to approximately 10,000 units per person by the end of the study with over $8 000 per person higher costs compared to the control group Evaluation of standardized dosing in renal anemia treatment is suggested for patients with diabetes or cardiac comorbidities n TABLE OF CONTENTS Abstract n Table of Contents in List of Tables v List of Figures vi Glossary vn Acknowledgement IX Chapter 1 Introduction Background and Need Incidence and Prevalence of Chronic Kidney Disease Etiology of Chronic Kidney Disease Hemodialysis Treatment for ESRD Regional Settings Current Trends in the Management of Renal Anemia in ESRD Significance of Study Research Context Research Questions 1 2 2 3 4 6 9 9 13 15 Chapter 2 Review of Literature Renal Anemia Erythropoietin and ESA Therapy Iron Target Hemoglobin Clinical Decision-Making Decision-Making Tools Renal Anemia Management Protocol (RAMP) Evaluating Patient Outcomes Nursing Role Effectiveness Model 16 16 17 21 25 31 32 33 36 39 Chapter 3 Methods Study Design Sources of Data Sample Study Variables Data Analysis Cost effectiveness Ethics 42 42 42 43 44 45 46 46 111 47 Procedures Chapter 4 Results Comparison of Control and Protocol Groups Data Analysis Outcome Data Hemoglobin Levels Transferrin Saturation Levels Ferritin Levels Analysis of Renal Anemia Medication Use Intravenous Iron Erythropoeitin Stimulating Agents Confounding Variables Economic Analysis Intravenous Iron Epoetin Alfa Chapter 5 Discussion Research Question One Research Question Two 49 49 54 54 54 57 59 60 60 61 64 65 65 66 68 68 68 Chapter 6 Conclusion 76 References 80 Appendix A 98 Appendix B 104 Appendix C 105 Appendix D 106 Appendix E 107 Appendix F 108 Appendix G 110 Appendix H 117 Appendix I 124 IV LIST OF TABLES Table 1 Stages of Chronic Kidney Disease Table 2 Percent Distribution of Incident ESRD Patients in Canada by Table 3 2 Primary Cause of Renal Failure 1997-2006 3 Differences of Anemia Management Practice in Physician- 12 driven and Nurse-driven Protocol Approaches Table 4 Distribution of Patients in the Control and Protocol Groups 50 Table 5 Gender, Race, and Age of Patients in Control and Protocol Groups 51 Table 6 Age Distribution of Control and Protocol Groups Table 7 Hospitalization, Hemodialysis Days, Diagnosis and Co-morbidities 51 of the Control and Protocol Groups 53 Table 8 Control Group Mean Transferrin Saturation Levels and z Scores 58 Table 9 Protocol Group Mean Transferrin Saturation Levels and z Scores 59 Table 10 Effect Size of the Difference Between the Means of Control and Protocol Groups in Darbepoetin Alfa Dosages Table 11 The t tests for the Equality of Means in Epoetin Alfa Dosages Between the Control and Protocol Groups Table 12 63 64 Average Per Patient Cost for Epoetin Alfa in the Control and Protocol Groups 68 v LIST OF FIGURES Figure 1 Comparison of Age Distribution Northern Health and Okanagan Health Service Delivery Area Figure 2 7 Comparison of Population Ethnic Origins in Northern Health and the Okanagan Health Service Delivery Area Figure 3 7 A Comparison of Aboriginal Populations in British Columbia Figure 4 8 A Comparison of Income Distribution Between the Northern Health and the Okanagan Health Service Delivery Area Figure 5 8 The Nursing Role Effectiveness Model Highlighting the Relationship Between Structure, Process and Outcomes 40 Figure 6 Selection Process for Control and Protocol Groups 46 Figure 7 Comparison of Gender Split in Control and Protocol Group 52 Figure 8 Comparison of Race in Control and Protocol Group 52 Figure 9 A Comparison of Incidence of Co-morbidities in the Control and Protocol Groups Figure 10 55 A Comparison of Hemoglobin Level Means Between the Control and Protocol Groups Figure 11 56 The Percentage of Patients Reaching Target Transferrin Saturation Levels in the Control and Protocol Groups 60 Figure 12 Monthly Mean Ferritin Levels in the Control and Protocol Groups 61 Figure 13 Total Monthly Intravenous Iron Use in Control and Protocol Groups Figure 14 62 The Average Dose of Epoetin Alfa Per Person in the Control and Protocol Groups Figure 15 65 Total Monthly Cost of Intravenous Iron in Control and Protocol Groups 67 VI GLOSSARY Algorithm A set of clinical steps diagramming a decision tree which directs patient care BCPRA British Columbia Provincial Renal Agency, plans and monitors the delivery of renal care in British Columbia CKD Chronic kidney disease EPO Erythropoietin, amino acid hematopoietic growth factor ESA Erythropoietic stimulating agents ESRD End stage renal disease Hgb Hemoglobin, the oxygen carrying protein molecule in red blood cells KDOQI Kidney Disease Outcomes Quality Initiative, recognized as the international standard for evidenced-based clinical practice guidelines for all stages of chronic kidney disease, developed by multidisciphnary nephrology experts KRP Kelowna Renal Program serves the Okanagan Health Services Delivery Area with hemodialysis sites in Kelowna, Rutland and Vernon, British Columbia NH Northern Health NREM Nursing Role Effectiveness Model, A structure-process-outcome nursing model of care developed to examine the contribution that nursing roles have on patient outcomes (Irvine, Sidani & McGilhs Hall, 1998) The process component identifies three roles of the nurse and categorizes these roles based on the functions or activities of the nurse, independent, dependent, and interdependent roles vn • The structure component of the NREM consists of patients, nurses, and organizational variables that impact processes and outcomes of care • The outcome component of the NREM is the patients' health status, and the direct and indirect costs associated with nursing care • Nurse-sensitive outcomes are patient outcomes that can be attributed to nursing practice/actions NRP Northern Renal Program serves Northern British Columbia, has hemodialysis sites located in Prince George, Terrace and Ft St John, British Columbia Nurse-driven That which is initiated and directed by nursing practice OHDSA Okanagan Health Delivery Service Area Physician-driven That which is initiated and directed by physician practice PROMIS Patient Records, Outcomes and Management Information System, database system adopted by British Columbia Provincial Renal Agency Protocol Algorithm using evidence-based clinical practice guidelines RAMP Renal Anemia Management Protocol based on BCPRA standards and used by the Northern Renal Program TS AT transferrin saturation, an indicator of iron status percentage of iron saturating iron binding sites on transferrin vin ACKNOWLEDGMENT It is my pleasure to thank those who supported me through this thesis process I am grateful to my supervisor, Dr Martha MacLeod for her invaluable nursing and research expertise, to Dr Peter MacMillan for opening the world of statistics a little wider to a student with endless questions, and to Dr Vince Salyers for his encouragement, direction, and advice Special thanks are due Dr Malcolm Ogborn for his nephrology and research expertise, encouragement, and listening ear There are many people who have encouraged me to pursue this degree and too many to mention on this page Of these people, I would like to gratefully acknowledge Dr Robin Lowry for patiently teaching me about nephrology for the past five years, always having the time in his busy schedule to answer a clinical question and sparking an interest to pursue further education The following people have been instrumental in bringing me through this process by practical means My parents instilled in me the value of higher education and consistently prayed for me My husband Marc supported me by listening to my complaints, proofreading papers, doing housework and keeping life in order I couldn't have done it without you' Finally, I would like to thank my children, Jonathan and Amy, for their optimism ("you'll get it done, Mom") and their ability to live in the moment They have taught me what is really important in life IX CHAPTER 1 Introduction Nephrology nurses in the hemodialysis setting are facing new challenges in the health care system Greater incidence of chronic kidney disease worldwide, coupled with limited financial and human resources in the health care system, have placed significant demands on the hemodialysis nurse In response to these demands, hemodialysis nurses are adopting new approaches to practice to provide safe, economical and effective care Current trends in the hemodialysis setting are placing greater emphasis on enhancing the decision-making role of the nurse to provide clinically sound practice and cost effective care in renal anemia management Increasingly, protocols based on best practice guidelines are followed in the care of the hemodialysis patient The clinical effectiveness of these protocols is commonly measured by patient physiological outcomes such as achieving target hemoglobin levels, transferrin saturation levels and ferritin levels The cost of protocol care can be examined by reviewing the average use of treatments such as erythropoeitin stimulating agents and intravenous iron There is little evidence to show whether using or not using a protocol affects patient and cost outcomes, or whether these differences in outcomes can be attributed to nursing care Using the concepts of structure, process, and outcomes discussed in the Nursing Role Effectiveness Model, this study evaluated two approaches to renal anemia management Outcomes and costs associated with anemia management for hemodialysis patients were examined in two comparable regions within British Columbia (BC) in this case control study 1 Background and Need Incidence and Prevalence of Chronic Kidney Disease Chronic kidney disease (CKD) is defined as the presence of kidney damage for greater than three months with a glomerular filtration rate (GFR) of less than 60 ml/min/1 73 m2, with or without evidence of abnormalities in urinalysis, diagnostic imaging or renal biopsy (Eknoyan & Levin, 2002, National Institutes of Health, 2008) Table 1 classifies CKD into five stages based on GFR level (National Kidney Foundation, 2002) Table 1 Stages of Chronic Kidney Disease Stage Description GFR (ml/min/m2) 1 Slight kidney damage with normal or increased filtration >90 2 Kidney damage with mild decrease in kidney function 60-89 3 Moderate decrease in kidney function 30-59 4 5 Severe decrease in kidney function Kidney failure requiring transplantation or dialysis End Stage Renal Disease (ESRD) 15-29 <15 Note Adapted from, "National Kidney Foundation (2002) Clinical Practice Guidelines for Chronic Kidney Disease Evaluation, Classification and Stratification," American Journal of Kidney Diseases, 39 (Suppl 1) pSl-S266 CKD has a significant impact on society and the health care system Chronic kidney disease is estimated to affect between 1 9 and 2 3 million Canadians (Levin et al, 2008) While there are no statistics available regarding prevalence of CKD in separate regions of British Columbia, the overall estimated prevalence of chronic kidney disease is 145,000 cases (BCPRA, 2008) Patients with CKD have a risk of cardiovascular disease that is ten to thirty times higher than those without CKD (Sarnak et al, 2003) CKD is often associated with anemia, cardiac disease, hypertension, and diabetes all of which require aggressive 2 chronic disease management CKD is correlated with increased length of hospital stay thereby having an impact on patients, families, society, and health services (Mix et al, 2003) Stage five CKD is also known as end stage renal disease (ESRD) At stage five, renal replacement therapy (RRT) is required to extend life (Levin et al, 2008) RRT consists of either dialysis or transplantation Hemodialysis and peritoneal dialysis are two forms of dialysis treatment In 2006, 82 1% of all new ESRD patients were started on hemodialysis treatment (Canadian Institute for Health Information, 2008) Etiology of Chronic Kidney Disease The etiology of CKD has changed over the past decade Table 2 lists the primary causes of ESRD between 1997 and 2006 Table 2 Percent Distribution of Incident ESRD Patients in Canada by Primary Cause of Renal Failure 1997-2006 Diagnosis 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Diabetes 28 9 29 9 314 32 1 33 8 33 8 34 2 34 3 35 0 34 4 Vascular 178 195 20 5 20 3 192 184 186 184 195 195 PCKD 40 47 43 45 47 39 42 43 51 8 GN 55 16 0 153 13 7 13 3 12 7 12 9 13 1 114 116 41 42 Pyelonephritis 4 7 47 45 40 43 45 39 35 Drug induced 16 17 13 18 21 21 20 18 20 17 Unknown 15 8 13 1 12 5 118 14 0 14 7 14 3 13 5 12 2 12 9 Other 9 9 11 5 9 7 10 1 10 9 10 8 95 10 1 109 11 7 Adapted from Canadian Institute for Health Information, 2008 Annual Report—Treatment of End-Stage Organ Failure in Canada, 1997 to 2006 (Ottawa, Ont CIHI, 2008) ISBN 978-1-55465-391-1 (PDF) Note PCKDPolycystic Kidney Disease, GN- Glomerulonephritis Diabetes has remained the most prevalent cause of ESRD in Canada Diabetic nephropathy was the leading cause of renal failure between 1997 and 2006 in Canada, with incidence rates increasing to 34 4% of all new cases of renal failure in 2006 (Canadian Institute for Health Information, 2008) The global increase in the incidence and prevalence 3 of diabetes over the last twenty years has had a direct impact on the incidence of CKD and the need for dialysis treatment The second most common cause of ESRD is hypertensive/vascular disease In the United States between 1990 and 2001, the incidence of ESRD related to hypertension has increased by almost 50% (Eustace & Coresh, 2005) Since 1997, the incident rates of ESRD related to vascular disease have increased minimally in Canada from 17 8 % to 19 5 % These differences may be related to better treatments of hypertension in the last decade, therefore reducing the rate of renal disease related to vascular disease Hypertension is both a cause and consequence of ESRD Over 75% of CKD patients develop hypertension related to an inability of the kidney to regulate blood pressure (Levin et al, 2008) Other causes of CKD are glomerulonephritis (GN), polycystic kidney disease (PCKD) and obstructive uropathy (Pereira, Sayegh & Blake, 2005) Glomerulonephritis is the third leading cause of ESRD in Canada Within the last decade, there have been decreasing trends in the incidence of glomerulonephritis and pyelonephritis and nearly unchanged rates of PCKD in Canada (Canadian Institute for Health Information, 2008) Hemodialysis Treatment for ESRD Hemodialysis is derived from two terms, hemo meaning "blood" and dialysis meaning "to pass through" Blood is passed through a semi-permeable membrane by diffusion to separate toxic wastes, excess water and electrolytes from the blood (American Nephrology Nurses' Association, 2007b) The overall treatment goal of hemodialysis is to correct electrolyte and fluid imbalance in addition to removing uremic toxins (Levin et al, 2008), completed by the continuous flow of blood across a dialyzer membrane (Yeun & Depner, 2005) A concurrent passage of dialysate solution along the opposite side of the semi-permeable membrane provides an environment in which loss of solutes can occur via 4 diffusion (Nesrallah, Blake, & Mendelssohn, 2005) If these goals of dialysis are met, the patient may have improved survival by reducing the effects of uremia and its complications Hemodialysis is the most common form of renal replacement therapy available to those patients with stage five CKD (American Nephrology Nurses' Association, 2007b) Conventional hemodialysis consists of 3 to 4-hour treatments, three times a week (Levin et al, 2008) In Canada, 47 7% of all hemodialysis patients are provided treatment in a hospital setting (Canadian Institute of Health Information, 2008) In 2005, 57 3 % of hemodialysis patients in B C received treatment in a hospital, while 36% received treatment in a community dialysis unit, with the remainder being dialyzed at home (Canadian Institute for Health Information) Nationally, the demand for renal services, in particular RRT, is increasing each year There were estimated 33,832 ESRD patients in Canada in 2006, which corresponds to an increase of 69 7% since 1997 (Canadian Institute for Health Information, 2008) In Canada, the incidence of ESRD and those requiring RRT among patients 75 years and older has more than doubled from 1996 to 2005 (Canadian Institute for Health Information) In British Columbia, the increase in the number of hemodialysis patients is considerable In 1996, there were 747 hemodialysis patients, while in 2005 there were 1812 hemodialysis patients (Canadian Institute for Health Information) These trends indicate the growing need for RRT in addition to the provision of efficient and fiscally responsible resource allocation of renal services within the health care system The outcome of patients on dialysis remains poor The one year survival rate of hemodialysis patients in Canada for 2004 was 81 7% (Canadian Institute for Health Information, 2008) In 2008, the one year survival rate of hemodialysis patients in British Columbia was 83% (British Columbia Provincial Renal Agency, 2008) The three year 5 survival rate for hemodialysis patients in Canada was 57 7% in 2003 (Canadian Institute for Health Information) Between 1996 and 2005, the five year survival rate for Canadian patients on dialysis was approximately 40% (Canadian Institute for Health Information) The overall five year survival rate of hemodialysis patients in Canada shows differences in ethnicity Patients of Black and Asian origin showed an unadjusted five year survival rate of 61 7% and 55 6% respectively Patients of Aboriginal and Caucasian origin had the lowest survival rates of 42 5% and 36 3% respectively (Canadian Institute for Health Information, 2008) Prevalence of ESRD among Aboriginals in Canada is 58 4% higher than in non-aboriginals (Canadian Institute for Health Information) Regional Settings This study compared two British Columbia regional hemodialysis programs consisting of the Northern Renal Program (NRP) and the Kelowna Renal Program (KRP) The NRP services all renal patients located within the Northern Health (NH) The KRP serves all patients within the Okanagan Health Service Delivery Area (OHSDA) This section will provide a comparison of demographics in a health region (NH) and a sub region of Interior Health (IH) identified as the OHSDA NH provides health care to 7 5% of the population of British Columbia (BC) encompassing a geographical area of 66 7% of BC (BC Stats, 2007a) The KRP provides renal care to the OHSDA, which serves over 3 5% of BC's population covering 3 3% of the area of BC (BC Stats, 2007b) A comparison of age, ethnicity, family income and aboriginal descent between the two health jurisdictions indicate some differences for these variables between these communities The Northern Health (NH) has a younger population than the Okanagan Health Service Delivery Area (OHSDA) in all age categories (Figure 1) The OHSDA has an aging population with 19 7 % above the age of 65 years compared to 9 6% in NH region (BC Stats, 6 2007 a,b) NH has a higher South Asian and Filipino population in comparison to the OHSDA (Figure 2) At 15 6% (BC Stats), NH also has the highest rate of aboriginal persons living in the region compared to the OHSDA and all other BC health authorities (Figure 3) The OHSDA has 69 2% of its population falling within the $20,000-$79,999 income bracket in contrast to NH at 58% (BC Stats) In the NH region, 29 1% of its population has a family income bracket over $80,000 which is higher than the OHSDA at 19 9% (Figure 4) 60 50 40 Percent Distribution of Region 30 p- jjj 20 L J B •I 10 0 t• -NH •. - OHSDA OUyears 18-24years 25 64years 65+years Age Distribution Figure 1 Comparison of age distribution in Northern Health and the Okanagan Health Service Delivery Area (BC Stats, 2007a) Percent Distribution of Region -, ,I NH OHSDA :m.-nL Single origins Chinese South Asian Filipino Japanese Other Ethnic origin Figure 2 Comparison of population ethnic origins Northern Health and the Okanagan Health Service Delivery Area (BC Stats, 2007a) Note Single origins-indicates a single ethnic origin other than ethnicity displayed on figure, e g Canadian, English, French, First Nations, other European 7 Percent of Population OHSDA Interior Health Vancouver Island Health Authority Fraser Health Vancouver Coastal Health Health Service Delivery Area or Health Authority Figure 3 A comparison of the aboriginal populations in BC (BC Stats, 2007a) 80 70 60 50 Percent of Population • NH 40 OHSDA 30 20 10 0 < $ 20 000 $ 20 000-79 999 1 >$ 80 000 Yearly family income Figure 4 A comparison of the income distribution between Northern Health and the Okanagan Health Service Delivery Area (BC Stats, 2007a) The differences between the region (NH) and the subregion (OHSDA) may have affected the outcomes of this study As the OHSDA has an aging population, the rates of hemodialysis patients within this age range were higher than NH The higher rates of Aboriginal persons in NH corresponded to higher rates of hemodialysis patients within the 8 NRP When examining the income distribution of both regions, no conclusions were drawn as to the effect on the hemodialysis population in this study Current Trends in the Management of Renal Anemia in ESRD Significance of Study Nephrology is a specialty in which health professionals work together in a system of interdependent roles to provide patient care At times, these roles will change based on organizational change or changes in the protocols or the structure of decision-making that directly meets patient needs (Sidani & Braden, 1998) Change in nursing practice has always been a part of the nephrology nursing As early as the 1960's, as technology in nephrology progressed, nephrology nurses embraced more responsibilities and autonomy with dialysis treatments (Fulton & Cameron, 1989) Often these changes were due to limited financial and human resources in the health care system Based on the current trend of increasing prevalence of ESRD, a shortage of nephrologists is evident in the health care system (American Nephrology Nurses Association, 2007a) Nephrology nurses are increasingly working to full scope of practice as clinical decision support tools are utilized Successful management of anemia through implementation of clinical practice guidelines can greatly improve patient outcomes (National Kidney Foundation, 2007) Although clinical practice guidelines may have been used prior to the anemia protocol implementation, the patient outcomes could not be attributed to a nurse sensitive intervention since the care was physician-driven and dependent For this study, Sidani's (1998) Nursing Role Effectiveness Model (NREM) was used as the framework to delineate the relationship between structure, process and outcomes in a specific nephrology nursing context This model frames the context of the nurse changing from a dependent role to an independent role in renal anemia management 9 Outcomes are an essential aspect of evaluating the contributions of a profession Many researchers describe a link between outcomes and interventions in the health care system (Donabedian, 1985, Doran, 2003, Newell, 1996, Sidani & Braden, 1998) It can be challenging to identify appropriate outcomes that result primarily from nursing care, but this is a necessary aspect of professional accountability and high quality patient care (Doran, 2003, Frauman & Gilman, 2001) Nurse sensitive outcomes are those outcomes which can be directly affected or influenced by nursing care (Frauman & Gilman, 2001) With financial restraints and limited nursing resources, it is imperative that nurses show that the care resulting from their actions leads to positive patient outcomes Evaluation of nephrology nursing interventions is commonly measured by physiological patient outcomes Patient outcomes such as the adequacy of dialysis, blood pressure, fluid status, pulse and respiratory rate are based on nurses' assessment and intervention before and during a hemodialysis run (Burrows-Hudson & Prowant, 2005, Frauman & Gilman, 2001) A measureable patient outcome for anemia management as an aspect of nephrology nursing care is achieved by maintaining target hemoglobin (Hgb) levels (Burrows-Hudson & Prowant, 2005) Appropriate assessment, decision-making and nursing interventions in anemia management produce physiological patient outcomes that can be considered nurse sensitive Hemoglobin is the best indicator anemia management Nurses can keep Hgb levels stable through the use of an algorithm Consequently, Hgb levels can be seen as a nurse sensitive outcome It is essential that nurse sensitive measures demonstrate how nursing actions influence patient outcomes The goal of renal anemia management is to provide Hgb levels within an acceptable target range Evidence shows that effective renal anemia management takes into 10 consideration the interdependent relationship between iron status, intravenous iron use, erythropoietic stimulating agent (ESA) use and Hgb These factors, properly managed are associated with reduced mortality and hospitalization among hemodialysis patients (Easom, 2006, Elliott, Pham, & Macdougall, 2008, Pisoni et al, 2004) Appropriate renal anemia management requires evaluation of iron status and ESA dosing to improve patient outcomes The use of decision support tools, such as algorithms, can assist nurses to care for the complex patient A decision-making tool is an evidence-based document used by nurses to guide the assessment, diagnosis and treatment of specific client clinical issues (College of Registered Nurses of British Columbia, 2008) These tools can support nurses as they assume increasingly independent practice in chronic disease management and reduce variation in client care related to individual practitioner practices (Dickerson, Sackett, Jones, & Brewer, 2001) The goal of using decision support tools is the achievement of positive patient outcomes in an efficient, cost effective and continuous manner Algorithms can be based on guidelines as outlined by the specialty and provide guidance in decision-making They are not a replacement for clinical judgment Algorithms or protocols can present evidence-based guidelines in visual form and enable nurses to outline aspects of disease management in order to make decisions about assessment, treatment and evaluation of care (Dickerson et al, 2001) Since 2006, algorithms in anemia management for hemodialysis patients have been mandated as standard practice in British Columbia (BC) by the British Columbia Provincial Renal Agency (BCPRA) The BC Medical Advisory Committee recommends that all BC Health Authority Renal Programs use algorithms for efficient use of erythropoietic stimulating agents and of nephrologists' time (British Columbia Provincial Renal Agency, 2007) Each health authority in British Columbia has the autonomy to develop an anemia 11 management protocol using the National Kidney Foundation Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines (British Columbia Provincial Renal Agency) The anemia management algorithm used in the NRP is based on evidence for managing renal anemia as outlined in the (KDOQI) practice guidelines (National Kidney Foundation, 2006) At the time of the study, the KRP used the KDOQI guidelines for renal anemia management based upon a traditional physician-driven approach A traditional physiciandnven approach consisted of monthly rounds with a review of patient blood work A charge nurse was present at rounds to provide blood work results and discuss each patient Adjustment of erythropoietin and intravenous iron doses was completed during rounds Any critical lab values were addressed urgently as necessary Table 3 outlines the differences between a physician-driven approach to anemia management and a nurse-driven protocol approach Table 3 Differences of Anemia Management Practice in Physician-driven and Nurse-driven Protocol Approaches Anemia management practice Physician-driven approach Nurse-driven protocol approach Responsible for assessment, Nephrologist Hemodialysis nurses monitoring and evaluation of anemia blood work (Hgb, TSAT, Ferritin levels) Procedure for adjustment of erythropoietin stimulating agents and intravenous iron doses Nephrologist adjusts medications during monthly rounds and urgently as needed Evaluation of effectiveness of Nephrologist evaluates treatments treatment during monthly rounds 12 Hemodialysis nurses adjust medications using step-wise instructions outlined in RAMP when lab work results are available Hemodialysis nurses provide ongoing evaluation of treatments using instructions in RAMP Within the NRP, the introduction of a renal anemia management protocol (RAMP) (Appendix A) resulted in a change in the role of the registered nurse in the care of the hemodialysis patient This change encompassed greater assessment, monitoring and decision-making within the scope of the registered nurse Anemia management has traditionally been the physician's role with minimal nurse input or action The RAMP transfers anemia management of hemodialysis patients to nurses upon order of the nephrologist Based on general RAMP guidelines, all patients have monthly lab work drawn to assess Hgb status Depending on Hgb status, the protocol provides step-wise instructions on how to titrate an ESA or provide iron therapy Steps for monitoring Hgb status are also provided in the protocol The targeted Hgb for this adult protocol is between 110-125 g/1 Transferrin saturation level is targeted between 20% and 50% The ferritin levels are recommended between 200-500 ug/L These targets were agreed upon by the RAMP committee based on the KDOQI guidelines at the time of printing This protocol enables the nephrology nurse to treat anemia in a timely manner Research Context This study compared patient outcomes in two settings Patients from the KRP served as a control group using data between November 2005 and November 2006 when an anemia protocol was not in use The NRP was identified as the protocol group during a time period that the protocol was in use from October 2007 to October 2008 The difference in time periods is due to the introduction of an anemia protocol in one of the community units linked to the KRP after December 2006 The KRP is considered to be a control population based on its comparable size in numbers of patients and the inclusion of two community units located in Vernon and Rutland, BC The KRP also has an organizational structure similar to the 13 NRP based on the British Columbia Provincial Renal Agency (BCPRA) renal program guidelines (British Columbia Provincial Renal Agency, 2004) The NRP provides renal services for all residents of Northern Health in BC This area encompasses the Northwest, Northern Interior and Northeast health service delivery areas in the province The primary renal management centre is located at the University Hospital of Northern British Columbia (UHNBC) in Prince George, BC Hemodialysis services consist of an in-centre hemodialysis unit at UHNBC and interdependent community dialysis units located at Terrace and Fort St John The UHNBC hemodialysis unit serves as the regional base for both community units In October 2007, there were 73 in-centre hemodialysis patients, 17 Fort St John community patients, and 14 Terrace community patients in the program (British Columbia Provincial Agency, 2008) for a total of 104 patients with approximately 60 patients on the RAMP The KRP provided renal services to all hemodialysis patients in the OHSDA The primary regional unit is in Kelowna General Hospital and two interdependent community units are located in Rutland and Vernon, BC In October 2005, there were approximately 38 Kelowna in-centre patients, 19 patients in the Rutland unit and 25 patients in the Vernon unit At this geographic location, 82 patients were not on an anemia protocol Through a retrospective, non-equivalent, group design, this study looked at the impact of implementing a decision support tool within the NRP hemodialysis population The outcome variables evaluated included Hgb levels, transferrin saturation levels, ferritin levels in addition to intravenous iron and ESA use An exploration of the relationship between interventions and outcomes in two comparable hemodialysis populations occurred 14 Research Questions The introduction of the RAMP in the NRP arose from the need to better manage Hgb levels for patients with renal anemia The trend of exceeding Hgb targets was of concern for patient safety and cost effective use of ESA's The goal of this study was to provide evidence concerning whether the use of a nurse-driven decision making tool in the hemodialysis setting could result in effective and safe patient outcomes compared to traditional physician-driven approaches to anemia management There are two research questions that were explored in this study 1) In the hemodialysis population, does using a nurse-driven anemia management protocol enable patients to maintain target Hgb levels as effectively as a non-protocol based physician-driven approach7 2) Is there a relationship between anemia management practice approach and Hgb levels, iron levels, intravenous iron use and erythropoietin use in the hemodialysis setting7 15 CHAPTER 2 Review of Literature The treatment of renal anemia requires an understanding of the impact ESAs and intravenous iron have on patient outcomes such as target Hgb levels and iron levels in the body The use of decision making tools such as algorithms or protocols, directs the nurse with steps to treat, monitor and evaluate renal anemia management Using the evaluation framework of the Nursing Role Effectiveness model (NREM), this study will apply its concepts to nurse-driven renal anemia management Renal Anemia Anemia is a condition marked by a reduction of red blood cells in the body (National Institutes of Health, 2008) Although there are many contributors to anemia, renal anemia is unique to end stage renal patients as it directly relates to erythropoietin deficiency in the body Erythropoietin is vital to the development of red blood cells Erythropoiesis is a physiologic process of maintaining oxygen levels in the body through the actions of erythropoietin (Elliott et al, 2008) The release of erythropoietin directly corresponds to oxygen levels in the body (Elliott et al) Hypoxia, which triggers a renal oxygen sensor in the kidney, leads to an increase in the production of erythropoietin resulting in the development of red blood cells (Hodges, Rainey, Lappin, & Maxwell, 2007, National Institutes of Health, 2008, Tranter, Martinez, & Rayment, 2006) By increasing the concentration of red blood cells and Hgb, the oxygen carrying capacity is improved Decreased erythropoietin (EPO) production in chronic kidney disease is the primary cause of renal anemia (Dalton & Schmidt, 2008, Drueke et al, 2006, Levin, 2007, Singh, et al, 2006) Renal anemia is a common complication of kidney failure (Collins et al, 2005, National Kidney Foundation, 2007, Singh & Hertello, 2005, Stnppoh, Navaneethan, & 16 Craig, 2006) At Stage 5 CKD (Table 1), also known as ESRD, 60-80% of patients are affected by renal anemia (Hsu, Mcculloch, & Curhan, 2002) There are various factors which can influence the development of anemia in hemodialysis patients Hemodialysis patients can experience gastrointestinal bleeding, shortened erythrocyte survival time of 30-60% of the normal 120 days, iron deficiency and increased blood losses at hemodialysis runs (O'Mara, 2008, Pisoni et al, 2004) Increased hemolysis related to the toxic effects of uremia can also exacerbate anemia in the hemodialysis patient (Pruett, Johnson, & O'Keefe, 2007) Anemia in hemodialysis patients has significant consequences Anemia is associated with decreased exercise capacity, reduced cognitive function, depression, and decreased quality of life (Ludwig & Strasser, 2001, Painter et al, 2002, Silverberg et al, 2003) Renal anemia can have a negative impact on cardiac function due to vasodilation, cardiac dilation, increased cardiac output leading to left ventricular hypertrophy, and congestive heart failure (Locatelh et al, 2007, Silverberg, lama, Wexler, & Blum, 2001) Between 1996 and 2005, cardiac failure was the leading cause of death among end stage renal disease patients in Canada (Canadian Institute for Health Information, 2008) Treatment of anemia has the potential to improve the strength and function of muscle as well as improve cognitive and brain electrophysiological function due to an increase in peripheral oxygen supply (Mason & McMahon, 1997, McMahon et al, 2000) Effective anemia management is associated with lower morbidity and mortality among hemodialysis patients (Brattich, 2006, Pisoni et al, 2004) Erythropoietin and ESA Therapy Erythropoietin (EPO) is an amino acid hematopoietic growth factor of which approximately 90 % of EPO is produced by the kidneys (Drueke et al, 2006, Elliott et al, 17 2008, O'Mara, 2008) If there is damage to the kidney, its ability to produce adequate amounts of erythropoietin is compromised (O'Mara) Renal anemia is treated by supplementing reduced erythropoietin levels in the body with erythropoietic stimulating agents (Cody et al, 2005, Dalton & Schmidt, 2008) EPO acts on the bone marrow to stimulate red blood cell production The bone marrow manufactures red blood cells through a series of events The stem cell also known as the hematopoietic progenitors differentiate into burst-forming erythroid cells and then colony-forming erythroid cells Each colony-forming erythroid cell has erythropoietin receptors and is erythropoietin dependent EPO is required to attach to the erythropoeitin receptor on the erythroid cell to continue process of producing red blood cells The erythroid cell is then differentiated into erythroblasts, reticulocytes and mature erythrocytes respectively (Elliott et al, 2008) The magnitude of increase in red blood cell concentration is controlled by the length of time EPO concentrations are maintained and not by the EPO level itself (Elliott et al, 2008) As a result, measuring EPO levels has little significance in renal anemia management The aim of renal anemia management is maintaining consistent EPO concentrations over time to impact Hgb levels The maintenance of adequate EPO levels is affected by several factors The balance between red blood cell production and destruction can be influenced by blood losses and kidney defects in oxygen sensing (Elliott, et al, 2008) Particularly in CKD, inadequate concentrations of erythropoietin and reduced life span of the red blood cell trigger an imbalance in erythropoiesis resulting in renal anemia Erythropoietic stimulating agents (ESA) are genetically engineered forms of the naturally occurring human erythropoietin hormone Epoetin alfa was the first ESA to be 18 developed and is administered 1 to 3 times per week with dosing based on weight of the patient, 50-100 units/kg (Duh, Werner, White, Lefebvre, & Greenberg, 2008) Epoetin beta is another form of ESA with similar starting doses as epoetin alfa Darbepoetin alfa is another form of erythropoeitin with a longer half life than epoetin alfa, taking 3 to 5 times longer to reach peak serum concentrations (Duh et al) Starting doses for Darbepoetin alfa are 0 45 ug/kg at weekly to every three week intervals (Duh et al) Erythropoietic stimulating agent (ESA) therapy, also known as erythropoietic hormone replacement therapy (EHRT), has become the first line of treatment for more than 90% of hemodialysis patients (National Institutes of Health, 2008, Patel, Robinson, & Singh, 2007) ESAs have changed the way renal anemia is treated and this treatment is associated with decreased mortality, morbidity, disease progression, cardiovascular risk, and an improvement in quality of life in hemodialysis patients (O'Riordan & Foley, 2000, Ritz & Eisenhardt, 2000) Prior to the use of ESAs, blood transfusions exposed the hemodialysis population to considerable risk Repeated blood transfusions do little to maintain Hgb at a target level and can cause iron overload (Bennett, 1998, Easom, 2006) Blood transfusions can also sensitize a patient to develop antibodies in the blood or viral infections, placing them at risk of unsuccessful transplant matches (Bennett) After exposure to Human Leukocyte Antigens (HLA) m previous blood transfusions, a patient can develop antibodies against HLA antigens resulting in difficulty finding a suitable transplant organ and increased risk of rejection of the organ (Magee, 2005) Blood transfusions can also introduce viral infections such as Hepatitis B, Hepatitis C, and Cytomegalovirus to the patient (Magee) Ideally, the use of ESAs can reduce the need for blood transfusions and can contribute to successful future transplantation 19 Erythropoietic stimulating agent dosing varies with clinical indication, practitioner and geographic region (Elliott et al, 2008) Results from the Dialysis Outcomes Practice Patterns study (DOPPS) of 12 countries demonstrate variable ESA dosing patterns worldwide (Pisoni et al, 2004) These differences are influenced by patient co-morbidities, hospitalization rates, national and regional practice guidelines, in addition to practice patterns (Pisoni et al) These variations account for many of the differences of patient outcomes worldwide Hyporesponse of hemodialysis patients to ESAs is identified in literature Hyporesponse to epoetin alfa is recognized as a potentially modifiable occurrence correlated to the following factors inadequate epoetin doses, blood loss, acute inflammation or infection, iron deficiency, poor nutritional status, and vitamin deficiency (Deziel, 2002, Kalanatar- Zadeh et al, 2009, National Kidney Foundation, 2006) High bone turnover disease as indicated by elevated serum alkaline phosphatase and parathyroid hormone levels has been correlated with ESA hyporesponse (Kalantar-Zadeh et al) Hyporesponse to epoetin alfa in hemodialysis patients is correlated with inflammation and malnutrition (Chawla & Knshnan, 2009) Literature shows that elevated pro inflammatory biomarkers such as c reactive protein (CRP) levels are present in some hemodialysis patients and coupled with low albumin levels, can be predictors of epoetin alfa hyporesponse (Breiterman-White, 2006, Deziel, 2002) Studies have demonstrated that high inflammatory biomarkers correlate with incidence of congestive heart failure, coronary heart disease, and vasculitis (van Tellingen et al, 2002) A patient with an inflammatory disorder, as evidenced by high ferritin levels, could require increases in epoetin alfa doses to maintain Hgb levels at acceptable targets (Drueke, 2001) Inflammation in diabetic hemodialysis patients is common as a result of the presence of atherosclerosis, proteinuria, diabetic 20 neuropathy, diabetic retinopathy, and infections (Jenq, Hsu, Huang, Chen, Lin, & Lin-Tan, 2009) The use of ESAs in the United States is on the rise and this has a significant impact on the financial resources of renal programs There has been a dramatic rise in epoetin dosing between 1993 and 2005 (National Institutes of Health, 2008) In the United States, between 1991 and 2005, the mean monthly Hgb in dialysis patients rose 24 g/L and the weekly EPO doses rose threefold to over 19,000 units (National Institutes of Health) ESA's account for 10% of ESRD costs of all patients on hemodialysis or peritoneal dialysis (National Institutes of Health) Iron The most important goal of iron therapy is to support optimal erythropoiesis to allow target Hgb levels to be reached and maintained It is crucial that careful monitoring of iron status and the dosing of iron therapy take into account the uniqueness of each patient An essential component of anemia management is the routine monitoring of iron status (Easom, 2006) KDOQI guidelines recommend iron status measurement every 3 months during stable ESA treatment (National Kidney Foundation, 2006) Regular monitoring of iron levels is a vital part of anemia management in the hemodialysis patient for several reasons It is common for hemodialysis patients undergoing ESA treatment to experience "functional iron deficiency" (Pruett et al, 2007, p 207) Iron deficiency is multi-factorial in the hemodialysis patient Iron deficiency can be due to retention of blood in the dialyzer and tubing as well as frequent laboratory tests (Singh & Hertello, 2005) Patients can fail to adequately respond to ESAs due to iron deficiency, as iron is a necessary component to complete the red blood cell production cycle (National Kidney Foundation, 2006) ESA treatment also increases erythropoiesis at rates higher than normal to support Hgb synthesis, 21 thus increasing the demand for iron faster than can be released from iron stores in the body (Pruett et al) Although the absorption of iron by the body increases as much as five times normal during erythropoietin therapy, losses from hemodialysis and blood testing may exceed gastrointestinal iron absorption (Fishbane, Frei & Maesaka, 1995, Skikne, Ahauwalia, Fergusson, Chonko, & Cook, 1998, Skikne & Cook, 1992) Transferrin saturation (TSAT) and ferritin levels are often used to diagnose and treat iron deficiency in hemodialysis patients (Singh, Coyne, Shapiro, & Rizkala, 2007) TSAT is a measure of iron stores available for red blood cell production corresponding to the circulating iron bound to transferrin (Kalantar-Zadeh et al, 1998) Greater TSAT has a positive association with effective anemia control (Pisoni et al, 2004) Serum ferritin is the amount of iron stored m the body as released by tissues (Easom, 2006, Kalantar-Zadeh, Rodriguez, & Humphreys, 2004) Ferritin levels alone are "an imprecise marker of iron status due to inflammatory factors which may interfere with synthesis and clearance of ferritin" (Easom, p 545) ESRD, hemodialysis, infections and protein energy malnutrition are factors which can activate the inflammatory response (Easom, 2006) Serum ferritin levels can increase during inflammatory disorders Inflammation iron block occurs resulting in a type of functional iron deficiency potentially disrupting erythropoiesis (Easom) Inflammation within the hemodialysis population may be as high as 40% to 60% (Kalantar-Zadeh et al, 2004) Elevated serum ferritin (500-2000pg/ml) should not be a marker for excessive iron but could be an "indication of iron plus inflammation in hemodialysis patients" (Easom, p 547) Nephrology nurses must use critical clinical thinking when interpreting KDOQI guidelines of serum ferritin of 500 p.g/ml as an upper limit for withholding intravenous (IV) iron therapy 22 (Easom) Nephrology nurses must evaluate each individual patient's response to IV iron based on ESA responsiveness, Hgb level, and clinical status Intravenous (IV) iron administration is the preferred route for hemodialysis patients There is a strong recommendation for treatment with intravenous iron within the hemodialysis setting in clinical guidelines (National Kidney Foundation, 2006) In a systematic review and meta-analyses of hemoglobin outcomes in hemodialysis patients using oral iron versus intravenous iron, the better hemoglobin response was with the patients treated with intravenous iron (Rozen-Zvi et al, 2008) Three randomized control trials comparing IV iron with oral iron administration in hemodialysis patients showed that the use of IV iron resulted in greater Hgb levels and reduced the need for higher ESA doses when compared with patients using oral iron (Fishbane, Frei, & Maesaka, 1995, Fudin, Jaichenko, Shostak, Bennett, Gotloib, 1998, Macdougall et al, 1996) In the past, iron overload was a legitimate concern in the care of hemodialysis patients Blood transfusions were common in hemodialysis patients and resulted in up to 6 g of parenteral iron per year (Easom, 2006) Iron, in the absence of erythropoiesis proved to be detrimental to the patient The greater use of ESAs coupled with the reduced need for blood transfusions has resulted in a reduced risk of iron overload in hemodialysis patients (Easom) It is unlikely that iron overload could occur with the use of appropriate intravenous iron dosing There are three forms of intravenous iron used in the RAMP (Appendix A) Iron dextran, iron sucrose and iron gluconate are parenteral forms of iron that are funded under the British Columbia Provincial Renal Medication Program The decision for which form to use is dependent on physician choice, patient tolerability and the evidence related to safety of each form 23 Iron dextran was the only form of parenteral iron available in the United States until 1999 It has proven efficacy but its safety has been questionable (Faich & Strobos, 1999) It has been reported that approximately 0 7% of iron dextran use in hemodialysis patients has resulted in life threatening anaphylactic reactions (Fishbane, Ungureanu, Maesaka, Kaupke, Lim, & Wish, 1996) Between 1976 and 1999, there were 30 deaths attributed to iron dextran use (Faich & Strobos, 1999) In a meta-analysis of studies of iron dextran use found drug intolerance rates of approximately 2 47% and anaphylaxis rates of 0 61 %, p< 0 0001 (Michael et al, 2002) It has been speculated that the high molecular weight dextran molecule rather than the iron itself induces anaphylaxis or contributes to adverse effects (Faich & Strobos, 1999, Michael et al, 2002, Sengolge, Horl, & Sunder-Plassman, 2005) Iron sucrose and iron gluconate (Ferrlecit) are increasingly being used in the hemodialysis setting due to the risk of serious adverse drug reactions related to iron dextran Iron sucrose therapy is a safe alternative to iron dextran use In a summary of four prospective studies of 130 iron dextran or iron gluconate sensitive patients, no serious adverse events (anaphylaxis or death) occurred related to iron sucrose therapy in these same patients Fourteen non-serious drug related adverse events such as diarrhea, hypotension, nausea, vomiting and constipation occurred in 8 patients (Charytan, Schwenk, Al-Saloum, & Spinowitz, 2004) Sodium ferric gluconate complex (iron gluconate or Ferrlecit) is a form of intravenous iron in which there is significantly less allergic and anaphylactic reactions occurring in comparison to iron dextran In a randomized controlled double blinded study of 2534 hemodialysis patients, there was a rate of immediate type of reaction of 0 04% (Michael et al, 2002) This indicates a statistically significant lower rate of anaphylaxis when compared to 0 61% in iron dextran use,/?=0 0001 (Michael et al, 2002) 24 Target Hemoglobin Levels Considering the impact of renal anemia on hemodialysis patients, several studies have looked at treatment and safety issues surrounding the use of ESAs For many years, the assumption in the nephrology community was that normalized Hgb levels presented positive outcomes for hemodialysis patients As a result, many studies were organized around this hypothesis A systematic review of Hgb targets in renal anemia found that many studies hypothesized that higher Hgb levels were positively associated with improved survival The results indicate otherwise Of twenty-two randomized control trials involving 3707 patients, it was found that Hgb > 130 g/L was not associated with decreased risk of mortality when compared to Hgb levels of 120 g/L Lower Hgb targets of < 100 g/L resulted in increased risk of seizure (RR 5 25, 95% CI 1 13-24 34) and a decreased risk of hypertensive episodes (RR 0 05, 95% CI 0 33-0 76, Strippoli et al, 2006, p i ) The results regarding quality of life and higher Hgb values were inconclusive due to non-validated scales of assessment and "presentation of individual positive results instead of a generalized assessment useful for analysis" (Strippoli et al, p 8) The limitations of this review include the limited number of RCT's available, the small sample sizes in RCTs and the lack of primary end points at the patient level (Strippoli et al) A meta-analysis of nine randomized controlled trials of various target Hgb of 5143 patients provided similar results (Phrommintikul, Haas, Elsik, & Krum, 2007) Higher Hgb targets demonstrated an increase of 20% in mortality, 30% in arteriovenous access thrombosis and 30% of poorly controlled blood pressure In this study, higher Hgb target levels were considered 120-160 g/L A sub-group of dialysis patients showed a relative risk of 1 11 (95 % CI 0 94-1 31, p=0 22) in the higher target Hgb group compared to the lower 25 Hgb group In addition, the risk of poorly controlled blood pressure was significantly higher in the high Hgb group than the low Hgb group (RR 1 27, 95% CI 1 08-1 50,p=0 004, Phrommintikul et al) In 2006, two studies were published that resulted in the amendment of the National Kidney Foundation guidelines indicating an upper Hgb limit of 130 g/L The Correction of Hemoglobin and Outcomes in Renal Insufficiency (CHOIR) trial studied 1432 patients with chronic kidney disease who were randomly assigned to receive epoetin targeted to achieve Hgb levels of 135 g/L or 113 g/L The primary end point was death, myocardial infarction, hospitalization related to congestive heart disease or stroke The results indicated that the higher Hgb group had a higher incidence of the primary end point and the trial was terminated early A target Hgb of 135g/L was associated with increased risk and no improvement in quality of life (Singh et al, 2006) Based on this evidence, the authors of the CHOIR study suggested that a target Hgb range of 110-120 g/L is based on lower risks and lower treatment costs The Cardiovascular Risk Reduction by Early Anemia Treatment with Epoetin beta (CREATE) study randomly assigned 603 patients with Stage 3 or 4 chronic kidney disease to two groups with target Hgb levels of 130-150 g/L or 105-115 g/L The primary endpoint was eight cardiovascular events including sudden death, myocardial infarction, acute health failure, stroke, angina or cardiac arrhythmias with twenty four hour hospitalization, and peripheral vascular disease complications After three years, the CREATE researchers found that there was no significant difference between the two groups with regard to adverse events They concluded that normalization of Hgb in CKD anemic patients does not reduce the risk of cardiovascular events The CREATE study did find a slight increase in quality of life measurement in higher Hgb group (Drueke et al, 2006) 26 Both studies (Drueke et al, 2006, Singh et al, 2006) found that the event rates were higher among the participants that responded poorly to ESAs Some critics argued that the difference in event rates was most likely due to the high burden of disease among these poor responders as opposed to the variation of Hgb within the target range (Demirjian & Nurko, 2008, Greene et al, 2005) Other critics argued that a slight increase in quality of life is of significance to the patient and this should be considered on an individual basis (Muirhead, 2007) The conclusions from these studies suggest that Hgb > 130 g/L is not recommended due to increased risks of cardiovascular adverse effects The studies also indicated that there are positive effects to quality of life when normalization of Hgb occurs A recent study on Hgb level variability did not find increased mortality among the above target range or even the higher end of target range (Gilbertson et al, 2008) Because this was an observational trial, its results have potential for more biases as compared to the CREATE and CHOIR randomized control trials and exert less influence over current treatment guidelines Other studies have provided inconsistent results as they pertain to safe target Hgb ranges The Anemia in Chronic Heart Failure Outcomes and Resources utilization study (ANCHOR) found that the ESRD population with chronic heart failure with Hgb >170 g/L levels were at the highest risk of death, and the risk was reduced as the Hgb decreased to 130 g/L level (Go et al, 2006) A randomized controlled trial of 157 hemodialysis patients within a group of predialysis and peritoneal dialysis patients demonstrated that Hgb levels of 135-160 g/L improved quality of life as measured by decrease in fatigue, depression and frustration The incidence of thrombovascular events and vascular access thrombosis in hemodialysis patients did not differ between the lower and higher target Hgb (Furuland et al, 2003) A randomized, double blind study of the effect of normalized Hgb targets (Hgb 13527 145 g/L) on hemodialysis patients without symptomatic heart disease found that it has no beneficial effect on cardiac structure as evidenced by left ventricular volume index (Parfrey et al, 2005) KDOQI guidelines from 2006 conclude that a Hgb lower limit of 110 g/L and an upper limit not > 130 g/L are the most efficacious in terms of disease management (National Kidney Foundation, 2006, p S33) On the basis of the CREATE and CHOIR studies, maintenance of Hgb levels >130g/L appears to require more epoetin, increasing costs to health care while placing patients at significant risk of adverse cardiovascular events (Steinbrook, 2006) Based on the wording of this guideline, no upper limit was determined except for avoiding Hgb >130 g/L This was considered an opinion-based guideline and has since been amended to an evidence-based guideline Current evidence supports a Hgb target range of 110g/L-120 g/L (National Kidney Foundation, 2007) Normalization of Hgb levels is a complex issue in hemodialysis patients As reported by the United States Renal Data System Annual Report, 42% of patients reaching Hgb > 140 g/L, achieve this level within six months of reaching the target level of 110 g/L (National Institutes of Health, 2008) Lack of attention to the higher target Hgb levels may lead to overuse of ESAs, contributing to overshooting targets (Collins, Ebben, & Gilbertson, 2007) Indeed, overshooting the target Hgb level was 3 to 4 times more common in 2008 than it was in 1997 (Kapoian, 2008) Higher ESA doses is associated with decreased survival by reflecting resistance to ESA treatment within coexisting inflammation and malnutrition issues (Kalantar-Zadeh, Kopple, Block, & Humphreys, 2001, Locatelh et al, 2006) Just as higher Hgb levels increase health risks for the hemodialysis patient, lower Hgb levels expose the patient to a greater risk In a study of 12 countries and dialysis outcomes, 23-29% of hemodialysis patients were below 110 g/L in Sweden, the United States, Belgium 28 and Canada (Pisoni et al, 2004) Hemodialysis patients with lower Hgb levels were potentially at the highest risk of mortality (Locatelh, Conte, & Marcelh, 1998) Hyporesponsiveness to ESAs creates a challenge to maintain the minimum Hgb target level One study indicated that 75% of ESA hyporesponsive patients with a non-functioning arteriovenous graft had evidence of a bacterial infection without symptoms The only evidence of infection was higher ferritin levels with low Hgb levels, suggesting inflammation or infection can greatly influence anemia management (Nassar, Fishbane & Ayus, 2002) Additional studies have addressed the challenge of maintaining Hgb levels within a narrow target range, indicating large variability in individual patient results (Berns et al, 2003, Lacson, Ofsthun, & Lazarus, 2003) "Achievement of Hgb within this target range is prone to fluctuation" (National Kidney Foundation, 2007, p 503), and is a common phenomenon in hemodialysis patients (Walker & Pussell, 2007) It has been reported that 90% of hemodialysis patients have cyclical Hgb results averaging 10 3 weeks and 25 g/L in range (Fishbane & Berns, 2005) The variability of Hgb levels observed in clinical practice indicates that 95% of hemodialysis patients using ESAs would obtain a Hgb interval range as high as 56 g/L (using a normal curve) (Berns et al, 2003) Hemoglobin variability can be influenced by the appropriateness of physician orders for ESA dose change, biological diversity within a population, unique patient responsiveness to ESAs, hemodilution related to fluid overload, adequacy of iron stores, bleeding, and inflammatory responses (Collins et al, 2005) According to one study of Hgb variability, a typical hemodialysis patient can be expected to have 42% of 3 month rolling average Hgb values outside of the 110-120 g/L range (Lacson, Ofsthun, & Lazarus, 2003) Variability in results among hemodialysis patients will present transient Hgb levels >130 g/L and a temporary spike does not constitute 29 a safety concern when appropriate adjustments are made to ESAs (National Kidney Foundation, 2006) Related to intrapatient Hgb variability, Lacson et al (2003) suggest that a wider target range will reduce the likelihood of staff responding aggressively to Hgb levels and decreasing Hgb fluctuations Variability of Hgb levels can be associated with increased mortality A one year retrospective study of 159,720 hemodialysis patients showed the degree of Hgb variability, low and highly variable levels were associated with increased risk of death (Gilbertson et al, 2008) Patients consistently within target hemoglobin levels, as well as those persistently over target, had lower mortality rates (Gilbertson et al) These results speak to the necessity of ensuring best outcomes for patients by maintaining Hgb levels within acceptable range Examining Hgb results require monitoring of longitudinal trends to prevent sudden reaction to isolated Hgb values, taking into account changes m a patient's condition such as infection or hospitalization (Breiterman-White, 2003) A single Hgb level often reflects a range of ESA doses over a period of time (Locatelli, Del Vecchio, & Pozzoni, 2007, National Kidney Foundation, 2007) Withholding ESAs for a Hgb level greater than target range contributes to unpredictable and variable Hgb levels downward Incrementally decreasing EPO doses provides effective clinical results as opposed to holding doses which can result in plummeting Hgb levels (Breiterman-White) There is an association between Hgb levels and hospitalization In a prospective observational study of 11,041 hemodialysis patients from 7 countries, results concluded that the risk of hospitalization decreased 9-55% over 5 years in patients with Hgb 110-120 g/L (Pisoni et al, 2004) In a study of U S hemodialysis patients treated with ESAs, Hgb levels dropped 5 3 g/L within 30 days of hospitalization as compared to 3 months prior and this was statistically significant (p< 0 001, Pisoni et al) It is evident that patients who are 30 hospitalized can experience more pronounced anemia Pisoni et al demonstrated that mortality and hospitalization risks decreased by 5-6% for every 10 g/L increase in Hgb level up to target range Clinical Decision-making Clinical decision-making is an essential skill that impacts the nurse, patient and institution The quality of clinical decision-making determines the delivery of nursing care and the quality of patient outcomes (Ames, 2006) Each nurse brings unique knowledge to a clinical problem and this plays a role in how a problem is interpreted and which clinical issues will be attended to (Jones, 1988 as cited in Bakahs & Watson, 2005) Clinical decision-making can also impact an institution The quality of a nurse's decision-making influences the outcome for the patient which can financially impact an institution (Ames) Because of the diversity of nursing, education, knowledge and experience in any health care institution, standardization of patient care can assist in providing better patient outcomes The role of the registered nurse (RN) has become more complex as changes to the health care system demand increased responsibilities for the care of the patient An emphasis on cost-effectiveness in conjunction with increased disease chromcity requires highly qualified, accountable individuals who accept responsibility for decision-making about patient care (Mrayan, 2003) A strategy to control costs while maintaining quality of care is to expand the scope of activities of the nurse Transferring medical management of a specific clinical issue creates an environment where nursing has greater responsibility and accountability for patient care (College of Registered Nurses of British Columbia, 2008) The College of Registered Nurses of British Columbia (CRNBC) strongly encourages employers to develop decision-making tools to assist nurses in their clinical practice as responsibilities and activities expand In BC, health care employers are responsible for the 31 development and initiation of any clinically-based client decision making tool that nurses use (CRNBC, 2008) Decision-Making Tools Clinical protocols are based on a standardized approach to practice and derive its origins from algorithms (Ilott, Booth, & Patterson, 2010) A clinical algorithm is a set of logical, sequential steps shown as a decision tree which directs patient care (Gerdtz & Bucknall, 1999, Miller, York, & Ryan, 2005) An algorithm in the clinical setting is linear in approach to clinical problem solving, displaying major decision points in disease management and strategies of action (Department of Veteran Affairs, 2002, Schwartz & Griffin, 1986) Using a specific, logical approach to clinical problem solving is essential to decision-making and can assist in the assessment of the patient (Department of Veteran Affairs, 2002) Algorithms and/or protocols which incorporate clinical practice guidelines provide a framework for evidence-based care within a specialty Within the nephrology specialty, the KDOQI clinical practice guidelines and recommendations are the standard guidelines for all stages of chronic kidney disease (National Kidney Foundation, 2006) Algorithms within a specialty cover the scope of a guideline to provide a summary of appropriate management decisions and strategies to address specific patient issues (Hadorn, 1994) Ideally, an algorithm is the translation of research into functional interventions, using a problem solving orientation to create steps grounded in evidence-based principles and practice (Miller et al, 2005) The advantages of clinical algorithm use in practice relates to its impact on patients, clinical practice, and institutions Algorithms can assist nurses to identify when testing is unwarranted and as a result provides more efficient patient care (Hadorn, 1994) Algorithm 32 use in clinical practice can improve patient safety related to standardization of care regardless of size or geographical remoteness of practice site (McDonald, 2007) Reducing variations by standardizing clinical practice is effective in minimizing the probability of error in judgment (Kohn, Corrigan, & Donaldson, 1999), thus placing the nurse at lower professional practice liability risk (Bucknall & Thomas, 1995, Gerdtz & Bucknall, 1999) Algorithms reduce the risk-adjusted outcomes for an organization by reducing the risk of medication error (Vanhae,cht, de Witte, & Sermeus, 2007) Within an organization, algorithm use provides a basis for establishing autonomy of practice within a safety net of mutually agreed upon actions Limitations of clinical algorithms are well-documented "The systematic use of algorithm approaches has the potential to hinder development of more flexible approaches to problem solving" (Gerdtz & Bucknall, 1999, p 55) It can be important to look outside the decision-making tree to find solutions that may apply to a specific patient or clinical situation Perhaps the most imminent concerns with using algorithms are having clinical practice dictated, losing control of practice and being monitored by others (Hartigan et al, 2003) Another drawback is the time required to delineate all of the information in the algorithm, much of which does not apply to the patient at hand (Hadorn, 1994) Other criticisms of algorithms are the questionable clinical validity of some recommendations, which are not always linked to best guideline practices or systematic reviews (Hadorn) Renal Anemia Management Protocol (RAMP) A RAMP is a type of algorithm that provides standardization of care in a hemodialysis unit The primary components of an anemia management protocol include target Hgb levels, defined parameters for use of ESAs and intravenous iron, defined and corrective actions related to causes of a hyporesponse, and clear documentation of 33 assessments, interventions and outcomes (Michael, 2005) The measure of success of any anemia management protocol is its ability to guide clinical interventions to ensure every patient has the same standard of care If these components are directed by evidence-based guidelines, the achievement will be improved patient outcomes and organizational efficiency When examining the use of protocols in the management of anemia in hemodialysis patients, the evidence points to its benefits "Increased utilization of anemia management algorithms to guide treatment decisions for ESAs and iron therapy can allow a range of renal care professionals, in acute and primary care settings, to deliver consistent and effective treatment of patients to recommended hemoglobin target" (Macdonald, 2007, p 185) Renal anemia management using a treatment algorithm can reduce variability of Hgb levels by implementing best practices and routine Hgb assessments in the care of the patient (Breiterman-White, 2003, Lacson, Ofsthun & Lazarus 2003) Standardized care and reduction of variability in Hgb levels are essential to producing positive patient outcomes in the hemodialysis population The use of a standardized anemia protocol in the nephrology setting has the potential to impact patient outcomes In a recent study of target Hgb levels and risk of hospitalization, those patients with more months below target range were less likely to have received intravenous iron and were more likely to be hospitalized or die (Ishani et al, 2008) A lack of appropriate and timely anemia treatment could be avoided with the use of a renal anemia management protocol "Institutional variability in anemia care is a potentially modifiable factor associated with the inability to achieve target hemoglobin concentrations" (Ishani et a l , p 1686) Anemia management protocols are used extensively in North America with varying results related to target Hgb and ESA use A study of facility factors which affect the 34 achievement of target Hgb found that although the protocol was effective in the initial management of anemia management, maintenance and refinement of the protocol were a necessary step to improving target ranges (Chan, Lafayette, Whittemore, Hlatky, & Moran, 2008) In a randomized controlled trial of an anemia management protocol in hemodialysis patients, the use of erythropoeitin was substantially reduced with no improvement in the achievement of target Hgb levels (Bnmble, Rabbat, McKenna, Lambert, & Carlisle, 2003) A study of hemodialysis patients in the Northern Alberta Renal program found no significant change in anemia pre-implementation to an algorithm versus post-implementation with a noted concern regarding the adherence of the algorithm throughout the study (Nhan, Jensen, & McMahon, 2007) Collins et al (2007), analyzed the likelihood of Hgb above target levels related to practice patterns Patients with Hgb > 130 g/L were assessed to determine if appropriate dose reduction in ESAs was prescribed in the month following change in Hgb level It was reported that approximately 70% of dialysis providers followed KDOQI guidelines, but this was often dependent on whether a dialysis facility was owned by a corporation, or it was hospital-based Hospital based dialysis facilities complied with guidelines more often Although anemia protocols may reduce ESA use and thus reduce costs to a health care institution, patient outcomes have not consistently improved, at times dependent on individual clinical practice and expertise in the application of these tools Protocol-based care by renal nurses has potential benefits of providing direction and structure for practice to the novice nurse, while enabling autonomy in decision-making to manage anemia in the hemodialysis patient It also creates an environment of critical thinking for the nurse to use the KDOQI guidelines for anemia, iron and ESA and assess how the interaction of these indices determines and guides decisions (Breiterman-White, 2003) Many dialysis facilities develop anemia management protocols to guide ESA dosing where 35 there are limitations to available physician and nursing time to provide anemia management (Werner & Levey, 2007) In addition to efficiently using ESAs, protocol use can assist the nurse to identify patients who are hyporesponders (Demirjian & Nurko, 2008) Protocols can contribute to the efficacy of resources such a lab testing Evaluating Patient Outcomes Outcomes assessment is a necessary component of nursing care because it provides evidence for accountability of practice (Irvine, Sidani, & McGilhs Hall, 1998) Donabedian introduced the model of health care quality using structure, process and outcomes as three quality determinants of health care (Donabedian, 1966) The Donabedian model has influenced the development of theories and models to delineate nursing outcomes research Donabedian's model focuses on the quality of health care by examining the structure of the environment and the processes that result in measurable outcomes Originally, this framework influenced studies on cost, length of stay, patient mortality, and patient satisfaction (Doran, 2003) It has now been used to address patient outcomes which can be sensitive to nursing care within a variety of health care settings (Doran) Those outcomes that are nursing sensitive are "relevant based on nurses' scope and domain of practice and for which there is empiric evidence linking nursing inputs and interventions to the outcome" (Doran, 2003, p vu) The National Quality Forum (NQF) (2004) defines nursing sensitive outcomes as outcomes that are affected, provided and /or influenced by nursing personnel, but for which nursing is not entirely responsible The relationship between nursing actions and nurse sensitive outcomes is "quantifiable but not necessarily causal" (National Quality Forum, 2004, p 2) The development of nurse sensitive outcomes began in the 1990's as a result of media and scientific studies identifying decreased quality of patient care and outcomes (Needleman, 36 Kurtzman, & Kizer, 2007), as well as the shortage of nurses withm the health care system (National Quality Forum, 2004) As a result of concerns regarding quality of care and nursing human resource shortages, many nursing outcomes studies have focused on nurse staffing, prevention of adverse events and patient safety (Aiken, Sochalski, & Anderson, 1996, Aiken, Clarke, Sloane, Sochalski, & Silber, 2002, Cho, Ketefian, Barkauskas, & Smith, 2003) Many studies have demonstrated the correlation between nurse staffing, processes of care and patient outcomes In a meta-analysis of 28 studies, there was an association between RN-to-patient ratio and patient mortality and adverse effects (Kane, Shamhyan, Mueller, Duval, & Wilt, 2007) The link to negative outcomes such as adverse effects, complications or mortality is the most published area of nursing outcomes research (Doran, 2003) There is a gap in nursing research in the area of studying nursing actions that demonstrate positive patient outcomes instead of avoiding negative outcomes Additional challenges in nursing outcomes research are related to the limited evidence directly linking patient outcomes to nursing activities in acute care settings A review of 4000 systematic reviews and 500 meta-analyses on nursing interventions and patient outcomes in acute care settings found that there is limited evidence to establish a direct association between nursing actions and patient care outcomes (Bolton, Donaldson, Rutledge, Bennett, & Brown, 2007) For example, patient satisfaction is identified as a nurse sensitive patient outcome (Doran, 2003) despite its vague definition Patient satisfaction appears to be patient specific, influenced by factors such as affective response and lack of standardized measurement Bolton et al (2004) strongly encouraged the development of standardized nursing interventions and patient outcomes to test the efficacy of nursing interventions across a patient population 37 Nursing sensitive care can be categorized into three areas (National Quality Forum, 2004) First, patient-centered outcome measures are those outcomes of care delivered to patients by nurses They can include failure to rescue, falls, restraints, pressure ulcers, central line infections and urinary tract infections from catheters Second, nurse-centered intervention measures focuses on aspects of nursing intervention and processes of care, such as nicotine counseling Third, system-centered measures focus on organizational effectiveness that influences and is influenced by nursing care System centered measures are represented by skill mix, nursing care hours and nurse turnover There are limited standardized nurse sensitive measures which are applicable in the nephrology nursing setting Within the nephrology specialty, nurses are in a key position to influence patientcentered outcomes Nephrology nurses are instrumental in improving anemia treatment and resultant outcomes because of their coordinating role in the care of the hemodialysis patient, working to reduce the clinical symptoms of anemia and thus influencing renal and cardiovascular health (Macdonald, 2007, Singh & Hertello, 2005) Nephrology nurses are actively involved in data collection and patient assessments, enabling identification and management of anemia (Bennett & Alonso, 2005) The use of a RAMP to produce target Hgb levels can be utilized to test of the efficacy of nurse-driven anemia management across the hemodialysis population There remains a need for nursing research to contribute to a greater understanding of the unique interventions of nursing and the impact these actions have on quality patient outcomes It is often difficult to separate the contributions of nurses from those of physicians and other health care providers (Naylor, 2007, Needleman, Kurtzman, & Kizer, 2007) Nursing actions are often intertwined with others' actions within the health care team 38 Determining actions unique to the nursing role can validate nurses' contribution to patient care Nursing Role Effectiveness Model The Nursing Role Effectiveness Model (NREM) was developed based on the Donabedian model (1966) to address the need to measure quality nursing care (Doran, 2003) The NREM (Figure 5) provides a framework to delineate the relationship between structure, process and outcomes within the context of nursing (Doran et a l , 2006) This model groups nursing roles into three categories and links these roles to nurse-sensitive patient outcomes Structure Process Patient Independent role Age, gender, education, type of severity of illness, comorbidities Nursing Interventions Nurse Dependent role Education, experience Organizational Staffing, staff mix, workload work environment Execution of medical orders Physician initiated treatments Outcomes f Nursing-Sensitive patient outcomes Functional status, selfcare, symptom control, safety/adverse occurrences, patient satisfaction Interdependent role Team communication, coordination of care, case management Figure 5 The Nursing Role Effectiveness Model highlighting the relationship between structure, process and outcomes Adapted from The Nursing Role Effectiveness Model (Irvine, Sidani & McGilhs Hall, 1998) The structure component consists of patients, nurses, and organizational variables that impact processes and outcomes of care (Doran, Sidani, Keating, & Doidge, 2002) Patient variables such as diagnosis, severity of illness, and co-morbidity can influence patient outcomes Nurse variables can include number of years experience, educational background 39 and clinical skills Organizational variables are those measures that emphasize nursing staff ratio, workload measurement and assignment patterns The process component identifies three roles of the nurse and categorizes these roles based on the functions or activities of the nurse, independent, dependent, interdependent roles The independent role includes those role functions such as patient assessment, decision-making, and intervention for which only nurses are responsible The nurse's dependent role includes clinical actions and judgments involved in the implementation of physician's orders and/or medical treatments The interdependent role encompasses all functions and responsibilities which nurses share with other health care professionals such as inter-professional communication and coordination of team-based patient care The outcome component of the model consists of nurse-sensitive patient outcomes Nurse-sensitive patient outcomes are those changes that occur m the patient as a result of nursing interventions (Doran, 2003) such as patient's health status, patient's perception of nursing care, and direct/indirect costs related to nursing care The NREM provides a framework for the study This study examined the relationship between structure, process and outcomes in renal anemia management Patient structural variables including age, gender, race, diagnosis, location of dialysis, number of days of dialysis, number of days of hospitalization, and co-morbidities influenced process and outcomes variables There are no nursing or organizational structural variables examined in this study The process variables differ between the control and protocol groups The control group nurses enact a dependent role to provide medically directed renal anemia management to hemodialysis patients The protocol group nurses worked within the independent role by using a nurse-driven protocol to provide renal anemia management to hemodialysis patients 40 The protocol group nurses also worked within the interdependent role when communication with the nephrologist or renal pharmacist was required according to protocol guidelines The differing process variables in either the control or protocol groups may have influenced the outcomes of the study Outcome variables in this study were represented by clinical lab values, dosage use of ESAs and intravenous iron, and cost of ESA and iron use The primary patient outcome examined in this study was Hgb levels This study also examined whether or not Hgb levels could be considered a nurse-sensitive patient outcome Other patient outcomes that were assessed in this study are transferrin saturation levels, ferritin levels, ESA use, and intravenous iron use 41 CHAPTER 3 Methods This study design was framed by the Nursing Role Effectiveness Model, illustrated in Figure 5 and discussed in the previous chapter In this study, renal anemia was managed by the nurse in two different ways The control group hemodialysis nurses utilized physician's orders to manage renal anemia (physician-driven approach) In contrast, the protocol group nurses functioned independently in decision making, assessment, intervention and follow up to provide care to renal anemia patients using a RAMP (nurse-driven protocol) An examination of the control and protocol groups was completed by using a case/control study design This study design was chosen to provide a comparison of two groups to evaluate clinical patient outcomes and costs of a nurse-driven protocol approach to renal anemia management in contrast to an established physician-driven approach to anemia management Study Design A retrospective, non-equivalent case control group design was used to determine whether the process of using a nurse-driven RAMP was associated with equally effective patient outcomes when compared to a physician-driven anemia management approach The study time interval for the control group was October 2005-October 2006 The study time period for the NRP was from October 2007-October 2008 Using two groups from different settings was intended to increase the representativeness of the results Sources of Data This study used data collected and maintained by the British Columbia Provincial Renal Agency (BCPRA) from the Patient Record/Registration and Outcome Management Information System (PROMIS) database The BCPRA is under the Provincial Health Services Authority It is an agency that improves the health of individuals in British 42 Columbia with kidney disease by implementing province-wide solutions to the specialized needs of renal patients in collaboration with BC health authorities and community partners BCPRA manages personal health information under the BC Freedom of Information and Protection of Privacy Act (FOIPPA) PROMIS is the health information system for the BCPRA and the only province-wide integrated registry for renal patients in Canada This database incorporates extensive demographic and clinical data on all hemodialysis patients in BC Data collected from renal units provides information for patient management, renal unit management, research, continuous quality improvement, and outcomes-based planning (British Columbia Provincial Renal Agency, 2009a) It incorporates clinical tools to support direct patient care in addition to supporting resource allocation The BCPRA uses PROMIS to provide an information link between renal health care professionals and the renal care community in British Columbia The PROMIS database provides patient-specific information such as patient demographic information, treatment characteristics, underlying chronic conditions causing CKD, date of registration, lab data, and medication profiles Unit clerks in the regional hemodialysis units enter this information into the PROMIS database for all BC hemodialysis patients Sample A non-probability sample of eligible participants was drawn from all hemodialysis patients in the NRP on the RAMP and all hemodialysis patients in the KRP not on a protocol The control group consisted of a sample of hemodialysis patients («=64) with renal anemia in the KRP for a 1 -year period in which a standardized nurse administered anemia management protocol was not in use The protocol group included all hemodialysis patients («=43) in the NRP that were evaluated and treated using the RAMP Using the information available from 43 the NRP hemodialysis unit records, a list of those patients who were treated using the RAMP in October 2007 was compiled by the renal pharmacist This list, in addition to personal health numbers, was sent electronically to the BCPRA There was no disclosure of this list or information to the researcher or any party other than the BCPRA After the list had been sent, the BCPRA also received an electronic list of exclusion and inclusion critena, as well as variables to study, from the researcher Matching by gender, age and diagnosis was not possible due to a limited number of patients eligible for the study Figure 6 illustrates the selection process to determine eligibility of the control and protocol groups for the study Control Group Protocol Group 82 hemodialysis patients 104 hemodialysis patients 18 patients excluded due to mortality or not on hemodialysis for the full one-year study period 61 patients excluded from study due to mortality, not on protocol, or not on hemodialysis for the full one-year study period 64 control hemodialysis patients in study 43 protocol hemodialysis patients in study Figure 6 Selection process for control and protocol groups All patients that died at any period during the study were excluded All chronic hemodialysis patients registered with the BCPRA and who had Hgb levels and at least one transferrin saturation reading in the database were included in the study Study Variables To provide descriptive comparisons of the control and protocol groups, data regarding the following variables were examined age, gender, race, diagnosis, co-morbidities, location of dialysis treatments, and length of time on dialysis Patient outcomes were measured 44 clinically by lab values and dosages of renal anemia medications such as ESA and intravenous iron The lab data specific to this study were Hgb, transferrin saturation, and ferritin levels Hemoglobin is the protein component of the red blood cell The Hgb was measured as g/L in this study Hemoglobin lab values were measured monthly for each patient Transferrin saturation is a lab value that indicates the amount of transferrin attached to iron and it is measured in percentage For this study, transferrin saturation was the primary measurable indicator of available iron in a patient The transferrin saturation level was measured every three months Ferritin levels were also used as an iron indicator Ferritin levels were measured intermittently based on need for assessment of iron status in the patients Renal anemia medications were evaluated based on patient usage A patient's ESA requirement was defined as the monthly dose of ESA, either Epoetin alfa or Darbepoetin alfa, administered during the study The intravenous iron requirement was defined as the monthly dose of iron, either iron dextran, iron saccharate (iron sucrose), or ferrous gluconate (ferrlecit) administered during the study Data Analysis Data was analyzed using SPSS 17 0 software Descriptive baseline statistics of gender, race, age, diagnosis, days of hospitalization, co-morbidities, and duration of hemodialysis were presented using n, means and/or standard deviations for the control and protocol groups To determine if significant differences existed between groups in Hgb levels, /-tests were completed Repeated measure testing was not feasible as there were missing data in the protocol group, limiting the number of consecutive monthly Hgb readings available for this test Additional data analysis was conducted using chi square test for categorical Hgb levels Transferrin saturation levels were compared in each group using 45 means and z-scores to determine the rate of patients reaching target levels Ferritin levels were compared using means and relating these values to standard guideline target ferritin levels A comparison of intravenous iron use per month for each group was completed by averaging the overall use of iron To determine the differences between the mean dose of ESAs between the control and protocol groups, a Mest was performed The cost of intravenous iron use on a monthly basis was calculated by multiplying the mean dose of iron each month per group and the iron cost per mg The cost of epoetin alfa was calculated by multiplying the average per-person monthly dose and the epoetin alfa cost per unit Cost Effectiveness A comparison of cost effectiveness of the two approaches to anemia management involved reviewing the control and protocol groups' usage of renal anemia medications Evaluation of the costs of intravenous iron and ESA's enabled the researcher to provide a comparison of the groups Calculating costs included determining the total average monthly dosage of epoetin alfa per person and multiplying by the cost per unit dose The costs of intravenous iron were calculated based on the average monthly dose of the group multiplied by the cost per mg The cost per unit of medication is based on the BCPRA summary of monthly costs of medications (2009) available on the BCPRA website (BCPRA, 2009) Ethics Aggregate data from the BCPRA PROMIS database was utilized for this study Consent for the use of any data in the PROMIS database was signed at initial point of renal care in British Columbia (Appendix B) and therefore every patient in this study had a signed consent Preliminary permission to use this database was granted to the researcher pending ethics approval (Appendix C) Ethics approval was granted by the University of Northern 46 British Columbia Research Ethics Board and the Northern Health Research Review committees (Appendices D & E) This study consisted of the "secondary use of information that was anonymous, anonymized or de-identified/coded and where the research team had no access to the code"(Canadian Institutes of Health Research, Natural Sciences and Engineering Council of Canada, Social Sciences and Humanities Research Council of Canada, 2008, p 49) There was no possibility of patient contact as all the data necessary were present in this database and sent to the researcher in codified form No personal information could be re-identified by the researcher after coding had occurred and therefore the privacy of the participants was protected All data was secured in a locked filing cabinet or by means of secure passwordprotected electronic data files Destruction of files and database information generated for this study will take place no later than five years after data collection All paper-based information will be shredded while all computer-based information will be deleted five years after study completion Information for dissemination of results (presentation or publication) will be provided in a non-identifiable format to ensure the anonymity of subjects Procedures This study was conducted using the RAMP from NRP as an intervention in the protocol group The RAMP transfers anemia management of hemodialysis patients to nurses upon order of the nephrologist (Appendix A) This protocol was implemented as a unit-wide policy change in anemia management in the NRP All nurses were trained to use the protocol by attending two training sessions One of the sessions included a guest speaker from Fraser Health Authority A review of case studies with a resource nurse and a pharmacist provided 47 an opportunity to demonstrate the use of the protocol There was no dedicated protocol resource nurse available so nurses were working in pairs to use the RAMP All patients had monthly lab work drawn on a designated date to monitor Hgb levels The lab work also included iron studies indicated by transferrin saturation (TSAT) and ferritin levels at beginning of protocol, every three months thereafter and as needed Depending on Hgb status, the protocol provides step-wise instructions on how to titrate an erythropoietic stimulating agent (ESA) or provide iron therapy Steps for monitoring Hgb status were also provided in the protocol If the Hgb was 126 g/L or higher with a >20 g/L increase since last Hgb, the physician was notified The targeted Hgb for this protocol was between 110-125 g/1 Transferrin saturation level was targeted between 20% and 50% Ferritin levels were drawn every three months and as needed These targets were agreed upon by the protocol committee based on the KDOQI guidelines at the time of printing 48 CHAPTER 4 Results Comparison of Control and Protocol Groups This study presented two groups in comparison The control group consisted of hemodialysis unit patients from three centres in the KRP, Kelowna General Hospital (KGH), Rutland, and Vernon, treated without a RAMP The protocol group consisted of hemodialysis patients from three centres in the NRP, University Hospital of Northern British Columbia (UHNBC), Fort St John, and Terrace Northwest, treated using a RAMP Throughout this chapter, the control group data will be presented first This will allow a clear comparison of the group without the use of a RAMP to the protocol group Table 4 shows the distribution of patients in each group Table 4 Distribution of Patients in the Control and Protocol Groups Group n Control KGH 27 Rutland 16 Vernon 21 Total 64 Protocol UHNBC 29 Fort St John 7 Northwest 7 Total 43 Percent 422 25 0 32 8 100 0 67 4 16 3 16 3 100 0 Classification of descriptive data in the control and protocol groups is shown in Tables 5, 6 and 7 Table 5 includes a brief summary of patient demographic characteristics of gender, age and race data The term gender was defined as a patient's biological sex Caucasian and First Nations groups are displayed as these are the predominant races in each group Table 6 shows the age distribution of patients in each group 49 Table 5 Gender, Race, and Age of Patients in Control and Protocol Groups Variable Control Protocol Gender n (%) 34(53 1) 30 (69 8) Male 13(30 2) Female 30 (46 9) Age (yrs) 61 6 ± 13 8 Mean ± SD 62 0 ±15 4 64 65 Range Race n(%) Caucasian 51 (79 7) 29 (67 4) First Nations 2 (3 1) 12 (27 9) Note Values enclosed in parentheses indicate percentage Patients in the control group (n=64) had a mean age of 62 0 with a standard deviation of 15 4 years Approximately 84% of all the patients in the control group were in the age bracket of 41-80 years old The protocol group (n=43) had age data very similar to the control group with a mean age of 61 6 years and a standard deviation of 13 8 years The age distribution of the protocol group was very similar to the control group with approximately 84% of all patients in the age bracket of 41-80 years old Table 6 Age Distribution of Control and Protocol Groups Age (years) Control n(%) 1 (16) 11-20 21-30 2 (3 1) 31-40 2 (3 1) 41-50 9 (14 1) 51-60 12 (18 8) 61-70 13 (20 3) 71-80 21 (32 8) 81-90 4 (6 3) 91-100 0 (0) Protocol /?(%) 0 (0) 1 (2 3) 3 (7 0) 5 (116) 11 (25 6) 12 (27 9) 9 (20 9) 1 (2 3) 1 (2 3) There were significant differences between the groups for gender split or diagnosis There was a considerably greater percentage (70%) of males in the protocol group (Figure 7), while there was an approximate gender split of 50% in the control group Race data 50 indicated large differences between the groups The protocol group had almost a third of its patients identified as First Nations while the control group had less than 4% (Figure 8) The control group had approximately 10% of patients in an unidentified other race category 80 70 ! 60 H 50 Percent 4 0 _, 30 ' H control • Protocol 20 J 10 • " •V . .: Male Female Figure 7 Comparison of gender split in control and protocol groups 90 , 80 Percent 70 ' 60 . 50 40 a Control 30 2o 20 l 10 ! Protocol m85B8paiapfej BMflOTsPgggggl B8888SSpagg8 0 -> Caucasian First Nations Asian (Oriental, Indian, Filipino) Unknown Race Figure 8 Comparison of race in control and protocol groups Hospitalization days, hemodialysis days, diagnosis and co-morbidity data are presented in Table 7 The hospitalization data are measured by days hospitalized during the study Hemodialysis days pre-study are displayed as mean days Diagnosis was categorized 51 based on the most prevalent causes of kidney disease with an "other" category which did not identify a specific diagnosis in the database Co-morbidities were also reported as frequency and percent incidence Table 7 Hospitalization, Hemodialysis Days, Diagnosis and Co-morbidities for the Control and Protocol Groups Variable value Control Protocol Days of Hospitalization 0 days n (%) 33(516) 19(44 2) 1-20 days n (%) 19 (29 7) 14 (32 6) >21 daysn(%) 12(18 7) 10(23 2) Hemodialysis days pre study Median 882 790 Mean 1222 1235 1-500 days 23 (35 9) 14 (32 6) 501-1000 days 14(21 9) 14 (32 6) 1001-1500 days 9(14 1) 7(16 3) 1501-2000 days 8(12 5) 3 (7 0) 2001-2500 days 4(6 3) 2(4 7) >2500 days 6(9 4) 3(7 0) Diagnosis Diabetic Nephropathy 15 (23 4) 21 (48 8) Renal Vascular disease 15 (23 4) 6(14 0) Chronic Renal Failure 13(20 3) 5(116) Polycystic Kidney Disease 3(4 7) 2(4 7) Glomerulonephritis 3(4 7) 3(7 0) Obstructive Uropathy 1(16) 2(4 7) IgA Nephropathy 1(16) 2(4 7) Other 13(20 3) 2(4 7) Comorbidity Diabetes 19 (29 7) 28(65 1) LVH 37 (57 8) 34 (79 1) CHF 40 (62 5) 40 (93 0) CVA 43 (67 2) 41 (95 3) MI 44 (68 8) 41 (95 3) DYSL 13(20 3) 15(34 9) HTN 52(813) 41 (95 3) Note LVH-=left ventricular hypertrophy, CHF=Congestive Heart Failure, CVA= Cerebrovascular Accident, MI=Myocardial Infarction, DYSL= Dyshpidemia, HTN= Hypertension 52 Nearly 50% of patients in the control group were not hospitalized during the study period Over 70% of the patients had 1500 hemodialysis days or less pre-study Diabetic nephropathy, renal vascular disease and chronic renal failure were the leading primary diagnoses for the group, with just over 20% incidence respectively The control group had approximately 20% of patients within the "other" diagnosis category The primary comorbidity of the control group was hypertension (81 3%) Myocardial infarction and cerebral vascular accident were present in just over two-thirds of the patients in the control group Approximately 45% of patients in the protocol group had no hospitalization days Over 80% of patients had 1500 hemodialysis days or less Nearly 50% of the protocol group had a primary diagnosis of diabetic nephropathy The protocol group had about 10% less incidence of renal vascular disease and chronic renal failure compared to the control group The protocol group had a high incidence of cardiac co-morbidities Over 90% of patients in the protocol group had hypertension, myocardial infarction, cerebral vascular accident and congestive heart failure identified as co-morbidities The control and protocol groups were similar in total hospitalization days, number of hemodialysis days pre-study, and incidence of some primary diagnosis The percentage of patients with no hospitalizations days at 51 6 and 44 2 percent for control and protocol groups respectively were not importantly different (Table 7) Most of the patients in both groups had less than 1000 days of hemodialysis with similar mean days of over 1200 days The median hemodialysis days pre-study differed for the control and protocol groups at 882 and 790 days respectively Both groups had similar incidence of polycystic kidney disease, glomerulonephritis, obstructive uropathy, and IgA nephropathy 53 Based on diagnosis and co-morbidity incidence, there are considerable differences The control group had approximately 25% less incidence of diabetic nephropathy compared to the protocol group (Table 7) The control group had a 9% greater incidence of renal vascular disease than the protocol group The protocol group had over 90% of patients with cardiac co-morbidities such as myocardial infarction, congestive heart failure and left ventricular hypertrophy compared to the control group with approximately 30 % less incidence (Figure 9) The protocol group had over 30% more incidence of diabetes as a comorbidity compared to the control group (Figure 9) Percent 90 80 ' 70 -1 60 J 50 ' 40 J i Control 30 20 10 0 - I • # $ £ & * Protocol / J> £> <$ / / Cormorbidity Figure 9 A comparison of incidence of co-morbidities in the control and protocol groups Note LVH- Left Ventricular Hypertrophy, CHF- Congestive Heart Failure, CVA- Cerebrovascular Accident, MI- Myocardial Infarction Data Analysis Outcome Data Hemoglobin Levels Initial data analysis focused on comparison of Hgb means between the control and protocol group Group statistics demonstrated that the means between the groups were 54 similar (Figure 10) The largest difference of means was 4 g/L Data from more than 14 protocol group patients was missing for the months of December, March and July The control group had much less missing data with an average of 3 cases per month and a maximum of 6 cases missing in August The protocol group number of patients ranges from 21-43 and the control group number ranges from 58-61 Mean Hemoglobin g/L 140 135 130 125 120 115 110 105 100 95 90 »Control Hgb Mean Protocol Hgb Mean •*-« > O C _Q *- **- -^ C "^ Q0 +-» Month Figure 10 A comparison of hemoglobin level means between the control and protocol groups An independent, non directional, Miest was performed to compare differences of mean Hgb levels between the protocol group and control group A Levene's test for equality of variances indicated ap > 05 value for each month, suggesting that the variability in the groups was consistent and therefore equal variances was assumed in all cases (Appendix F) Calculation of Cohen's d effect size indicated that every month had trivial or small, effect size results (Appendix F) For purposes of this study, trivial effect size in a two-group test of mean differences was defined as <0 20, a small effect size was between 0 20 and 0 49, a medium effect size was between 0 50 and 0 79, and a large effect size was 0 80 or greater (Cohen, 1988) The Mest for equality of means indicated ap > 05 for every month, and in 55 addition to effect sizes, no significant statistical differences in Hgb means were identified between the groups (Appendix D) Following the /-test, Hgb data were regrouped into three categories below target, 75109g/L, target, 110-125 g/L, above target, 126-145 g/L These categories were based on Hgb values that indicate below target, target and above target according to KDOQI guidelines used to develop the renal anemia management protocol (Appendix G) To determine whether the Hgb categorical outcomes of the groups were statistically different, a chi square test of independence was conducted In every category, the observed and expected values were similar The p values were > 05 every month indicating both groups were not statistically different from each other For every month there was no indication that the ratios of below target, target or above target differed between the protocol and control groups Although categorizing the data into above or below targets (110-125 g/L) was used in the previous testing, neither category could be considered more unsafe for the patient than the other Categorizing the data into two categories to indicate on or off target was useful for further testing to determine if there were differences in the control and protocol group Hgb means The on target category was defined as Hgb values of 110-125 g/L, while off target category was defined as all Hgb values outside the target category range To determine if the Hgb observed values of the groups were statistically different, chi square testing was performed For all months there was no evidence of statistically significant differences between the groups, with on target or off target values/? > 05 in all cases (Appendix H) Because of the consistent chi-square statistics indicating the control and protocol groups were not statistically different from each other, no significant differences were found between the groups based on categorical comparison of on and off target Hgb 56 To improve sensitivity of the testing, a repeated measures test under the General Linear Model was attempted There were limited number of protocol patients with consecutive monthly Hgb levels present in the database (n=S) therefore conducting a repeated measures statistical test was not practical Transferrin Saturation Levels Transferrin saturation results were reviewed to determine if there were differences between the groups Transferrin saturation (TSAT) levels have a direct impact on the efficacy of erythropoietin stimulating agents as well as being a key indicator to assess iron needs (Kalantar-Zadeh et al, 1998, Pisoni et al, 2004) The protocol group using RAMP had TSAT levels drawn every three months routinely, periodic levels drawn dependent on the TSAT level, and follow up levels drawn after the use of intravenous iron The control group had monthly TSAT levels drawn The number of patients with TSAT levels differs considerably each month, particularly in the protocol group (Tables 8 and 9) Table 8 Control Group Mean Transferrin Saturation Levels and z Scores Month N Target Mean Control Mean SD z-Score October 61 >0 2 0 13 0 07 102 November 59 >0 2 0 13 0 06 1 10 December 61 >0 2 0 07 0 59 0 16 January 60 >0 2 0 72 0 16 0 06 February 60 >0 2 0 06 0 72 0 16 March >0 2 60 0 15 0 07 0 75 April 61 >0 2 0 17 0 10 0 27 May 59 >0 2 0 58 0 16 0 07 June >0 2 60 0 17 0 07 0 47 July 61 >0 2 0 74 0 15 0 06 August >0 2 55 0 15 0 06 0 82 September 60 >0 2 0 17 0 07 0 53 % Patient 15 14 28 24 24 23 39 28 32 22 21 30 To compare the TSAT means of the control and protocol groups, each mean was compared to the number of standard deviations above or below the target, using the KDOQI 57 guidelines value of TSAT >0 2 as the target mean Each z-score was calculated based on the number of standard deviations from the target The control group z-scores were consistently above zero while the protocol group z-scores were below zero The z-score was converted to a percentage based on assumed normality of distribution The percentage of patients reaching the target TSAT levels are presented in Table 8 and 9 Table 9 Protocol Group Mean Transferrin Saturation Levels and z-Scores Month n Target Mean Protocol Mean SD z-Score >0 2 0 34 October 8 0 17 -0 86 November 39 >0 2 0 28 011 -0 73 December 9 >0 2 0 35 021 -0 72 32 January >0 2 0 27 0 08 -0 87 34 February >0 2 031 0 10 -1 10 14 March >0 2 0 32 0 19 -0 62 April 14 >0 2 0 15 0 35 -100 May 36 >0 2 0 28 0 12 -0 65 June 9 >0 2 0 30 0 22 -0 47 July -0 91 13 >0 2 0 28 0 09 August 27 >0 2 031 0 12 -0 85 September 14 0 14 >0 2 0 30 -0 69 % Patients 81 77 76 81 86 73 84 74 68 82 80 75 The percentages of patients reaching targets indicate that the control group had TSAT means that were consistently below acceptable targets (Figure 11) The control group had an interquartile range of 22-28% of patients reaching target mean TSAT levels while the protocol group had an interquartile range of 75-81 % of patients reaching those levels With less serum iron saturation testing and the use of the RAMP, the protocol group had reached acceptable TSAT levels 58 ioo - 90 - 80 Percent 5040 ... 20 - - -- - -- . . _ . _ _ . — :__ — : % p at ients Reaching TSAT Target g I | I • 1 ~~B — I — I — I — I —B — I —• — I ~ Protocol Group % Patients Reaching TSAT Target u Contro| Group 10 0 Figure 11 The percentage of patients reaching target transferrin saturation levels in the control and protocol groups Ferritin Levels Ferritin is a protein that stores iron that is released and transported to necessary areas in the body Ferritin levels were not consistently available on a monthly basis in the control group The protocol group had ferritin levels drawn based on the RAMP recommendations The ferritin monthly mean levels are shown in Figure 12 KDOQI guidelines recommend maintaining ferritin levels at 200-500 pg/ml during ESA treatment to avoid iron deficiency in hemodialysis patients (National Kidney Foundation, 2006) The protocol group had monthly mean ferritin levels above 600 ug/L throughout the study No ferritin levels were available in the control group in May and August Control group monthly mean levels were below 500 pg/L in each month except November Data on ferritin levels were low for patients in the protocol group and not consistently available for patients in the control group, and therefore statistical testing to compare the means was not feasible 59 1400 1200 1000 Mean Ferritin levels (ug/L) 800 600 Control 400 Protocol 200 0 ^ / / 05) in mean values between the control and protocol groups in the 61 darbepoetin dosage category This is likely a type II error because the /-test could not detect significant differences, despite the evidence of medium effect sizes in seven out of twelve months The /-test is lacking sensitivity due to small sample sizes Table 10 Effect Size of the Difference Between the Means of Control and Protocol Darbepoetin Alfa Dosages Month October November December January February March April May June July August September Mean difference -89 07 -101 32 -115 82 -130 28 -79 60 -79 60 -50 65 -123 05 -101 32 -86 85 -86 85 -115 82 SD 13170 139 36 139 36 135 70 156 56 156 56 149 58 149 58 149 58 145 32 145 32 145 32 Cohen's d 0 67 0 72 0 83 0 96 051 051 0 33 0 82 0 68 0 60 0 60 0 80 Effect Size Medium Medium Large Large Medium Medium Small Large Medium Medium Medium Large Note p> 05 for all months analyzed A majority of patients in both groups were using epoetin alfa as an erythropoietin stimulating agent Equal variances were assumed in all months with the exception of April, May and June (Appendix I) Effect sizes for 9 out of 12 months demonstrated small effect sizes (Appendix I) An independent, non directional /-test was performed to determine if there were differences between the means of epoetin alfa dosages in the control and protocol dosages For each month, there were no significant differences (p > 05) between the mean dosages of epoetin alfa of the control and protocol groups (Table 11) 62 Table 11 The t-testsfor the Equality of Means in Epoetin Alfa Dosages Between the Control and Protocol Groups / Month Equal variances df P 93 88 October mean epoetin alfa dose assumed 0 153 93 November mean epoetin alfa dose assumed 0 082 90 84 December mean epoetin alfa dose 0 197 90 assumed 91 87 January mean epoetin alfa dose assumed -0 159 62 February mean epoetin alfa dose assumed -0 497 88 88 47 March mean epoetin alfa dose assumed -0 719 27 April mean epoetin alfa dose not assumed -1209 59 -1214 57 27 May mean epoetin alfa dose not assumed 25 June mean epoetin alfa dose not assumed -1 176 53 22 July mean epoetin alfa dose assumed 90 -1243 90 23 August mean epoetin alfa dose assumed -1215 89 27 September mean epoetin alfa dose assumed -1 113 To examine the average use of epoetin alfa per person in the control («=55) and protocol («=37) groups, the sum of the mean averages were divided by the number of patients using epoetin alfa in the group Figure 14 indicates that the average use at the beginning of the study was very similar for the control and protocol group at approximately 48,000 and 50,000 units per person respectively The use of epoetin alfa in the control group decreased progressively by approximately 3,000 units in May and then trended to levels similar to October results by the end of the study The use of epoetin alfa in the protocol group trended upwards by almost 10,000 units per person, per month by the end of the study 63 70000 r 60000 t 50000 \ 40000 Average Epoetin Alfa dose (units) 30000 Control Group l Protocol Group 20000 10000 0 PV /^V ^ °vv * ^ ^w J& Figure 14 The average dose of epoetin alfa per person in the control and protocol groups Confounding Variables To examine the effect any extraneous variables may have on the ESA dosages between the groups, /-tests were performed to determine if there were statistically significant differences between potential confounding variables There were noticeable differences in gender, race and co-morbidity frequency in the groups Further non directional /-tests show that there were significant differences (p< 05) between the incidence of diabetes, cerebrovascular accidents, congestive heart failure, myocardial infarction and hypertension between the control and protocol groups There were no significant differences of dyshpidemia incidence means between the control and protocol groups An examination of the entire group of patients was performed to determine if there were differences in race and co-morbidities The groups were tested together due to the low numbers of First Nations patients in the control group There were no significant differences (p> 05) in the total group of patients (both control and protocol combined) between Caucasian and First Nations means for all co-morbidity incidence 64 Next, a /-test to show if there were any significant gender differences in the control and protocol groups for use of Epoetin Alfa was completed Results indicated p > 05 for every month, therefore no significant differences were detected A /-test to determine if there were significant differences in each group between Caucasians and First Nations use of Epoetin Alfa was performed thereafter There were no significant differences between these races and Epoetin Alfa dosages in all the months in the control group and 10 out of 12 of the months in the protocol group (p > 05) An attempt to find significant differences between the iron dosages and co-morbidity incidence was not feasible due to low numbers of patients using each preparation of intravenous iron in each group The tables for each of these /-tests were not included in this study Economic Analysis An analysis of the cost effectiveness of the protocol was assessed, by examining the costs of intravenous iron and epoetin alfa treatments A comparison of the control and protocol group provided information regarding the cost of a non-protocol and protocol approach to renal anemia management All costs for medications are based on BCPRA 2009 summary of monthly drug costs (British Columbia Provincial Renal Agency, 2009b) Intravenous Iron When comparing monthly expenses for intravenous iron, the control group had elevated costs related to higher use of iron and the type of iron used The cost of iron saccharate is $36 63/100 mg, iron dextran is $16 83/100 mg and iron gluconate is $19 96/100 mg (British Columbia Provincial Renal Agency, 2009b) Over the time period of the study, the total intravenous iron cost of the control group was approximately $17,000 more than the protocol group (Fig 15) 65 Cost of Iron 1500 ($) Control group 1000 Protocol group 500 0 ^ ^ ^ * < ^ ^ f O ^ „c? *«-«f * *°~ 05) in the Hgb levels between the groups In addition, when the groups were separated into below and above target categories, there were no significant differences between the protocol and control groups In this study, hemodialysis patients in the protocol group maintained Hgb target levels as effectively as hemodialysis patients in the control group These findings provide some evidence to indicate that the use of nurse-driven anemia protocols can contribute to safe patient outcomes in the hemodialysis setting, as it pertains to maintaining target hemoglobin levels Research Question Two Is there a relationship between an anemia management practice approach and Hgb levels, iron levels, intravenous iron use, and erythropoietin use in the hemodialysis setting9 Throughout the study, iron saturation levels tended to reach target levels for the protocol group The measure of iron saturation levels available for red blood cell production is not a direct indicator of effective anemia control but evidence shows a strong association between 68 transferrin saturation (TSAT) levels and anemia control (Pisoni et al, 2004) KDOQI guidelines recommend TSAT levels > 0 20 to prevent iron deficiency anemia (National Kidney Foundation, 2006) For the protocol group, approximately 78% of their patients reached TSAT levels of 20, while approximately 25% of the control group reached that level The RAMP has specific interventions targeted at treating low TSAT levels The low rates of acceptable TSAT levels in the control group did not appear to affect the achievement of target Hgb levels Nothing in literature reviewed for this study addressed the impact of low TSAT on target Hgb, with the exception that TSAT levels do not have a direct cause and effect relationship with hemoglobin level, they are only associated between the variables Further studies examining TSAT levels and their relationship with target Hgb levels in hemodialysis patients are required to determine acceptable TSAT levels for renal anemia management The control group had monthly TSAT levels drawn, which reflected intensive monitoring, but did not appear to contribute to better health outcomes The TSAT levels remained far below target levels throughout the study The protocol group was tested for TSAT levels every three months and randomly depending on intravenous iron use Although the protocol group was tested less frequently, the patient outcomes were much better, with the added benefit of less discomfort to the patient and greater cost savings The cost of serum TSAT testing is $19 86 (A Arsenault, economics assistant, BCMA, personal communication, April 27, 2010) Monthly TSAT testing appears to be a costly and unnecessary intervention for patients Ferritin levels differed noticeably between the control and protocol groups Ferritin levels should be maintained at 200-500 ug/L to prevent iron deficiency in hemodialysis patients treated with an ESA (National Kidney Foundation, 2006) In this study, the control 69 group had ferritin levels consistently below 250 ug/L with the exception of two months where data was unavailable In spite of these two absent data points, the average ferritin level for the duration of the study was close to 300 ug/L Unfortunately there was sporadic data available in the control group due to testing based on individual needs and not as a scheduled unit routine As a result, statistical testing was not feasible This finding speaks to the need to monitor ferritin levels in practice, based on a more standardized protocol such as the RAMP utilized in this study The protocol group had monthly ferritin levels greater than 600 ug/L The average ferritin level for the protocol group was close to 800 ug/L Evidence shows that hemodialysis patients often have elevated ferritin levels related to inflammation, which interferes with the synthesis and clearance of ferritin (Easom, 2006) Elevated ferritin levels pose a problem when using stringent protocol guidelines The RAMP uses ferritin levels of >500 ug/L as an indication of iron overload although in the protocol group, this may be associated with underlying inflammation related to cardiac co-morbidities in the patients Further research is recommended to determine if there are statistically significant differences in ferritin levels between the groups and the implications that these differences may have for anemia management As well, further investigation into validity of using ferritin as an iron status indicator in the hemodialysis population is suggested The use of intravenous iron was also an indicator of anemia management that differed between the groups Although different intravenous iron preparations were used in each group, a comparison of elemental iron in each preparation allowed a comparison of iron use Iron doses were greater in the control group resulting in an overall use that was nearly 10 g more than the protocol group The higher iron doses in the control group were inconsistent with the low TSAT levels Results from this study suggest that iron deficiency was present 70 in the control group despite treatment with intravenous iron Low transferrin saturation despite treatment indicates a need to further investigate the relationship between transferrin saturation and iron deficiency and its impact on anemia management in hemodialysis patients When examining the repercussions of using intravenous iron at higher rates, most health care institutions focus on costs The costs of intravenous iron are substantial, depending on the type of iron in use The cost of iron saccharate ($36 63/1 OOmg) is much higher than iron gluconate ($19 96/1 OOmg) and iron dextran ($16 83/1 OOmg) (British Columbia Provincial Renal Agency, 2009b) The control group used iron saccharate exclusively, contributing to the inflated costs of iron treatment in comparison to the protocol group, which used iron gluconate and iron dextran The overall protocol group intravenous iron cost over the 12 month study period was $17000 less than the overall control group cost These are significant cost savings that could be related the type of iron used, in addition to the use of a protocol that facilitated the appropriate treatment of iron deficiency Another variable which showed differences between groups was the overall use of erythropoietin stimulating agents Although there was no statistically significant differences (p > 05) between the means of epoetin alfa dosages of each group, the average dose per person showed trends that indicate slightly greater use in the protocol group In the last 8 months of the study, the protocol group displayed an increased use of nearly 10000 units per person compared to the control group The protocol group had over $8000 higher costs per person for the use of epoetin alfa compared to the control group during the study The high cost of this medication and the trend of increased use of epoetin alfa present a concern about the financial implications of standardized dosing in a RAMP such as the one used for this study 71 The increased use of epoetin alfa in the protocol group is an indication of hyporesponse to the drug and hyporesponse was considered a confounding variable in this study Evidence demonstrates that hyporesponse to ESAs are associated with inflammation The higher incidence of cardiovascular co-morbidities in the protocol group could also provide an explanation for hyporesponse to epoetin alfa The influence of prevalence of cardiac co-morbidities was not considered a confounding variable at the study design stage Furthermore, the higher incidence of diabetes in the protocol group and the ensuing inflammatory process related to infection, microvascular disease, and atherosclerosis could play a role in hyporesponse to epoetin alfa Hyporesponse to epoetin alfa in conjunction with higher ferritin levels in the protocol group suggests the need for further investigation to determine if the protocol group had a higher incidence of inflammatory disorders or processes as compared to the control group Further need for prospective trials to examine the relationship between the presence of cardiac comorbidities in hemodialysis patients and response to ESA's are also indicated Another potential confounding variable is the higher First Nations population in the protocol group Although this study attempted to explore a relationship between ethnicity and ESA response through correlational analysis, limited numbers of patients from First Nations background in the control group prevented this Future research is warranted to examine if there is a relationship between First Nations ethnicity and ESA response The higher population of First Nations people in Northern British Columbia is a predictable finding from this study The challenges that the Northern Renal program faces is unique in the province Northern British Columbia has nearly 16% of its population identified as aboriginal located at 80 reserve First Nations communities (BC Stats, 2007a, Tabobondung, 2007) The First Nations populations in Northern British Columbia are both 72 numerous and diverse compared to the rest of the province The overall prevalence of diabetes among First Nations populations in BC is at least 40% higher than that of the general BC population The prevalence of diabetic nephropathy is much higher in First Nations than in the general population, and the rates range from 25-60% following 15 to 20 years with diabetes (Whiteside, 1994) The risk of developing ESRD from diabetes in First Nations populations is three times higher than the general population and of those with ESRD, the relative risk of starting dialysis is 6 5 times higher than the general population (Young, Kaufert, McKenzie, Hawkins, & O'Neil, 1989) This finding indicates a need to account for the type of population that the NRP serves to assist in appropriate health service planning, funding and policy development in this region This study explored patient structural variables and the independent or dependent role of the nurse resulting in nurse sensitive outcomes As indicated previously, the patient variables, such as co-morbidities, may have a significant influence on the patient outcomes of ESA use There is no conclusive evidence to indicate that higher ESA use in the protocol group is directly related to the independent role of the nurse in renal anemia management The higher ESA use may be related to deficiencies in the protocol in regards to dosing for patients with cardiac co-morbidities In contrast, the patient outcome of increased use of intravenous iron could be related to the dependent role of the nurse and lack of decisionmaking tools to properly treat iron deficiency The complexity of the interrelationship between patient outcomes such as Hgb, ferritin, TSAT, iron and ESA use, as well as patient variables such as diagnosis, race and the presence of co-morbidities complicates evaluating the process variables of the independent and dependent roles of the nurses In spite of challenges of examining the roles of nurses in this study, the protocol nurses are primarily responsible for ensuring that patients reach target Hgb levels Due to iron levels outside of 73 the parameter of the RAMP, in this case higher ferritin levels in the protocol group, physician or pharmacist consultation was required The NREM provided a framework to determine how nurse-driven anemia management contributes to nurse sensitive patient outcomes It was useful to conceptualize renal anemia management through a structure, process and outcome model that specifically examines the impact that nursing actions have on patient outcomes The structural component was applicable to this study by focusing on the patient variables which can impact patient outcomes In this study, patient co-morbidities and ethnicity had the potential of influencing patient outcomes This study did not examine nursing or organizational structural components as outlined by the NREM The influence of individual nursing variables such as education and experience, as well as organizational variables of staffing or hospital policy on patient outcomes in this study are unknown at this time Nursing research to test the NREM in the hemodialysis setting, to further explore nursing and organizational variables and their impact on patient outcomes is suggested Examining the process component by delineating the roles that nurses acquire to provide renal anemia management to the hemodialysis patient indicated that the control group maintained a dependent role while the protocol group used a combination of independent and interdependent roles The NREM also showed the relationship between the structure and process and its impact on outcomes In this study, physiological patient biomarkers such as Hgb were identified as potential nurse sensitive outcomes, directly influenced by nursing roles and patient variables Examining nurse sensitive outcomes in the hemodialysis setting informs the health care system of the contributions of nephrology nursing practice 74 In this study, the protocol nurses were the renal care professionals that provided most of the treatment care for the RAMP resulting in physiological patient outcomes The activities associated with the independent role of the nurse included assessment of lab work, assessment of patient's condition, decision-making regarding a plan of care and intervention according to protocol guidelines Therefore, using a nurse-driven RAMP in hemodialysis patients may have resulted in Hgb levels identified as a clinical nurse sensitive patient outcome 75 CHAPTER 6 Conclusion This study examined patient outcomes pertaining to the use of a RAMP to care for hemodialysis patients The study provided some evidence that the use of a nurse-driven protocol is as effective as a non-protocol, physician-based practice approach to renal anemia management Each group reached and maintained acceptable target Hgb ranges during the course of the study The study also demonstrated that the relationship between the practice approach to anemia management and Hgb levels, iron levels, intravenous iron use and ESA use cannot be evaluated effectively based on the complexity of the interrelationship between the variables Nurses at the NRP were meeting nephrology nursing standards by utilizing a RAMP The change in decision-making structure had enabled nurses to work to a full scope of practice in managing renal anemia by increasing their responsibilities in the process of assessment, planning and evaluation of hemodialysis patients Through the use of a decision support tool, namely the RAMP, nurses at the NRP were maintaining acceptable target Hgb levels in their hemodialysis patients Although this tool enabled nurses to provide the necessary actions to maintain Hgb levels, evidence from this study also indicated a need to examine the use and dosing of intravenous iron preparations and their relationship to iron level indicators in addition to ESA dosing and its impact on Hgb levels The relationship between these indicators was found to be more complex than anticipated at the study outset A link between these indicators was assumed but results from this study demonstrated otherwise For example, the literature provided a strong association between adequate TSAT levels and target Hgb levels although results from the control group in this study demonstrated otherwise Also, the influence of 76 confounding variables such as race, diagnosis and co-morbidity incidence may have added to the complexity of examining target Hgb levels in the hemodialysis population The higher incidence of cardiac co-morbidities and diabetes in the hemodialysis patients treated by the NRP challenges the idea that a standardized protocol fits all patients Inflammatory processes related to cardiac co-morbidities or diabetes could play a role in the clinical effectiveness and financial efficacy of anemia management protocols The incidence of erythropoietin hyporesponse coupled with higher ferritin levels creates new issues of financial feasibility for components of this protocol, particularly standardized dosing for all patients Prospective trials investigating alternate dosing patterns based on the incidence of inflammatory conditions are needed to provide evidence that will direct decisions about the use of ESA's This study examined how nurse-driven approach compares to physician-driven practice approach to renal anemia management Evaluation of nurse-driven protocols is necessary to determine how the role of the nurse evolves with practice changes and the resulting impact on patient care outcomes Using the NREM, the changes in the role of the nurse were delineated in each approach to determine if nurse sensitive patient outcomes could be identified in the protocol approach This study demonstrated that Hgb levels may be attributed to the independent role of nursing and be identified as a nurse sensitive patient outcome in nurse-driven renal anemia management Anemia management protocols are commonly being used in the hemodialysis setting to provide efficient, cost effective renal care Results from this study suggest that using a protocol directs nurses in providing renal anemia management in an efficient manner, potentially reducing the need for frequent lab testing and nephrologist consultation Cost effectiveness of aspects of the nurse-driven protocol was questionable Although iron doses 77 were lower and also less costly than the control group, hyporesponse to the ESA could potentially be more expensive The use of a RAMP suggests that it is an effective tool with potentially higher costs The need to review ESA and iron dosing in the protocol as it relates to patient variables such as cardiac or other inflammatory processes is encouraged It is evident that more investigation is necessary to establish the relationship between iron and epoetin dosing and cost A nursing limitation of this study is that the actions of the nurses were not directly measured to determine the time, education level, experience, and workload of each nurse and its influence on decision-making Although the use of the protocol was as effective as traditional physician practice, the impact on the nurses' workload and time management was not examined Further research is suggested to directly measure nursing experience, education and time as it relates to the use of the protocol and examine its connection with the nurse There are three further limitations to this study First, only data available in the BCPRA PROMIS database was used As a result, data entry may be subject to error, omissions may be present and limited standard terminology of variable terms in the database all contribute to the quality of the data in this study Second, there were insufficient numbers of patients eligible to provide adequate power, particularly in the protocol group A post hoc power analysis of/-test difference between two independent means, n=T06,/?= 0 05 and effect size = 5 indicated a power of 71% was achieved Low patient numbers eliminated the possibility of matching subjects based gender, race and diagnosis They also reduced correlation testing possibilities, particularly when gender or race was posed as potential confounding variables Although higher sample sizes could have improved power in this study and reduced the risk of a Type II error, a false negative conclusion, the retrospective 78 nature of the study in two clinical settings with limited patient populations poses as an obstacle to increasing power of the study Third, medical practitioner preferences played a role in cost effectiveness Preference of type of intravenous iron ordered by physicians increased costs significantly for the control group Identifying the most cost effective therapy for renal anemia in the hemodialysis setting without compromising efficacy or patient safety is an important objective for health care systems The health care system demands that programs demonstrate the impact nursing practice changes have on patient outcomes Nursing has a responsibility to ensure that their practices provide safe patient care in addition to meeting national standards Evaluating practice through patient outcomes achievement is one way to examine nursing practice Evaluating patient outcomes in the health care system is an essential part of determining the impact of practice change The complexity of the hemodialysis patient contributes to the difficulty of examining renal anemia outcomes when implementing a protocol This study demonstrated that the use of a nurse-driven RAMP is as effective as a traditional physician approach in ensuring that target hemoglobin levels were reached The challenge is to look at the multifaceted issue of renal anemia and determine if these protocols require modification to better meet the needs of this patient population 79 REFERENCES Aiken, L H , Sochalski, J , & Anderson, G F (1996) Downsizing the hospital nursing workforce Health Affairs, 15(4), 88-92 Aiken, L H , Clarke, S P , Sloane, D M , Sochalski, J , & Silber, J H (2002) Hospital nurse staffing and patient mortality, nurse burnout and job dissatisfaction Journal of the American Medical Association, 288(16), 1987-1993 American Nephrology Nurses' Association (2007a) ANNA position statements Nephrology Nursing Journal, 34(3), 319-335 American Nephrology Nurses' Association (2007b) Hemodialysis fact sheet [Brochure] Pitman, NJ ANNA Ames, E (2006) Practice standards for quality clinical decision-making in nursing Curatioms, 29(1), 62-72 Bakahs, N A , & Watson, R (2005) Nurses' decision-making in clinical practice Nursing Standard, 19(23), 33-39 BC Stats (2007a) Socioeconomic profiles health authorities In Health Authority Northernstatistical profile Retrieved March 30, 2009 from http //www bcstats gov be ca/data/sep/ha/ha_5 pdf BC Stats (2007b) Socioeconomic profiles health service delivery area Retrieved March 22, 2009 from http //www bcstats gov be ca/data/sep/hsda/hs_main asp Bennett, L (1998) The anemia research nurse in effective multi-disciplinary management of patients on erythropoietin EDTNA/ERCA Journal of Renal Care, 24(3), 38-39 Bennett, L & Alonso, M A (2005) Current practice in renal anaemia management two multinational surveys EDTNA/ERCA Journal, 31, 99-103 80 Berns, J S , Elzein, H , Lynn, R I , Fishbane, S , Meisels, I S Deoreo, P B (2003) Hemoglobin variability in epoetin-treated hemodialysis patients Kidney International, 64, 1514-1521 Bolton, L B , Donaldson, N E , Rutledge, D N , Bennett, C , & Brown, D (2007) The impact of nursing interventions Overview of effective interventions, outcomes, measures and priorities for future research Medical Care Research and Review, 64(2, Suppl l)pp 123S-143S Brattich, M (2006) Morbidity and mortality in patients on dialysis the impact of hemoglobin levels Nephrology Nursing Journal, 33(1), 64-69 Breiterman-White, R (2003) Hemoglobin variability impact on anemia management practices Nephrology Nursing Journal, 30,456-459 Breiterman-White, R (2006) C-reactive protein and anemia Implications for patients on dialysis Nephrology Nursing Journal, 33(5), 555-558 Bnmble, K S , Rabbat, C G , McKenna, P , Lambert, K , & Carlisle, E J (2003) Protocohzed anemia management with erythropoietin in hemodialysis patients a randomized controlled trial Journal of the American Society of Nephrology, 14, 2654-2661 British Columbia Provincial Renal Agency (2004) BC Provincial Renal Agency Guidelines for renal programs from http //www bcrenalagency ca/kidneysvcs/ProgramGuide/htm British Columbia Provincial Renal Agency (2007) Standard and Guidelines In Protocol For Delegation of Anemia Management Retrieved August 28, 2008 from http //www bcrenalagency ca 81 British Columbia Provincial Renal Agency (2008) At a Glance 2007-2008 BC Renal Agency Scorecard In At a Glance 2007-2008 BC Renal Agency Scorecard Retrieved August 14, 2008 from http //www bcrenalagency ca/kidneysvcs/Kidney+Care+Statistics htm British Columbia Provincial Renal Agency (2009a) Collection of personal information Kidney patients Retrieved March 30, 2009 from http //www bcrenalagency ca/NR/rdonlyres/86969F55-FB28-46C7-9305490F3B2FA729/32824/PrivacvBrochurefoiWebMar09pdf British Columbia Provincial Renal Agency (2009b) Summary of monthly drug costs In Pharmacy and Formulary Retrieved November 24, 2009 from http //www bcrenalagency ca/professionals/PharmForm/default htm Bucknall, T K , & Thomas, S A (1995) Decision-making in critical care 73(2), 10-17 Burrows-Hudson, S , & Prowant, B F (Eds ) (2005) ANNA Nephrology nursing standards of practice and guidelines for care (6 ed) Pitman, NJ American Nephrology Nurses' Association (Original work published 1977) Canadian Institute for Health Information (2008) 2007 Annual Report - Treatment of End Stage Organ Failure in Canada, 1996 to 2005 Ottawa CIHI Canadian Institutes of Health Research, Natural Sciences and Engineering Research Council of Canada, Social Sciences and Humanities Research Council of Canada (December 2008) Tri-Council Policy Statement Ethical Conduct for Research Involving Humans, Draft 2nd Edition Retrieved May 20, 2009 from http //www pre ethics gc ca/policy-politnque/docs/TCPS-Draft2-eng pdf 82 Chan, K E , Lafayette, R A , Whittemore, A S , Hlatky, M A , & Moran, J (2008) Facility factors dominate the ability to achieve target haemoglobin levels in haemodialysis patients Nephrology Dialysis Transplantation, 23, 2948-2956 Charytan, C , Schwenk, M H , Al-Saloum, M M , & Spinowitz, B S (2004) Safety of iron sucrose in hemodialysis patients intolerant to other parenteral iron products Nephron Clinical Practice, 96, 63-66 Chawla, L S & Krishnan, M (2009) Causes and consequences of inflammation on anemia managment in hemodialysis patients Hemodialysis International, 13, 222-234 Cho, S H , Ketefian, S , Barkauskas, V H & Smith, D G (2003) The effects of nurse staffing on adverse events, morbidity, mortality and medical costs Nursing Research, 52(2), 71-79 Cody , J Daly, C , Campbell, M , Donaldson, D , Khan, I , Rabindranath, K, Vale, L , Wallace, S , & MacLeod, A (2005) Recombinant human erythropoietin for chronic renal failure in pre dialysis patients Cochrane Database of Systematic Reviews, 3, pp 1-51 Cohen, J (1988) Statistical power analysis for the behavioral sciences (2nd ed ) Hillsdale, NJ Lawrence Erlbaum Associates College of Registered Nurse of British Columbia (2008) Scope of Practice for Registered Nurses Standard, Limits, Conditions (Pub no 433) Vancouver CRNBC, Collins, A J , Brenner, R M , Ofman, J J , Chi E M , Stuccio-White, M , Krishnan, M , Solid, C , Ofsthun, N J & Lazarus, J M (2005) Epoetin Alfa use in patients with ESRD An analysis of recent US prescribing patterns and hemoglobin outcomes American Journal of Kidney Diseases, 46(3), 481-488 83 Collins, A J , Ebben, J P , & Gilbertson, D T (2007) EPO adjustments in patients with elevated hemoglobin levels provider practice patterns compared with recommended practice guidelines American Journal of Kidney Diseases, 49, 135-142 Dalton, C , & Schmidt, R (2008) Diagnosis and treatment of anemia of chronic kidney disease in the primary care setting a primer for nurse practitioner Journal for Nurse Practitioners, 4(3), 194-200 Demirjian, S & Nurko, S (2008) Anemia of chronic kidney disease When normalcy becomes undesirable Clevelend Clinic Journal of Medicine, 75(5), 353-356 Department of Veteran Affairs (2002) Clinical Practice Guidelines In Frequently asked questions Retrieved July 9, 2008 from Http //www opq med va gov/cpg/faqs asp Deziel, S M (2002) Anemia management in patients with chronic conditions that affect erythropoiesis Case study of the anemic patient Nephrology Nursing Journal, 29(6), 582-585 Dickerson, S S , Sackett, K , Jones, J M & Brewer, C (2001) Guidelines for evaluating tools for clinical decision-making Nurse Educator, 26(5), 215-220 Donabedian, A (1966) Evaluating the quality of medical care Milbank Memorial Fund Quarterly, 44(3), 166-203 Donabedian, A (1985) The Methods and Findings of Quality Assessment and Monitoring An Illustrated Analysis (Vol 3) Ann Arbor, Michigan Health Administration Press Doran, D (Ed ) (2003) Nursing-Sensitive Outcomes State of science Sudbury, MA Jones and Bartlett Publishers 84 Doran, D , Harrison, M B , Laschinger, H , Hirdes, J , Rukholm, E , Sidani, S (2006) Relationship between nursing interventions and outcome achievement in acute care settings Research in Nursing and Health, 29, 61-70 Doran, D , Sidani, S , Keating, M , & Doidge, D (2002) An empirical test of the nursing role effectiveness model Journal of Advanced Nursing, 38(\), 29-39 Drueke, T (2001) Hyporesponsiveness to recombinant human erythropoietin Nephrology, Dialysis & Transplantation, 16, 25-28 Drueke, T B , Locatelli, F , Clyne, N , Eckardt, K , Macdougall, I C , Tsakins, D , Burger, H , & Scherhad, A (2006) Normalization of hemoglobin level in patients with chronic kidney disease and anemia The New England Journal of Medicine, 355(20), 2071-2084 Duh, M S , Werner, J R, White, L A , Lefebvre, P , & Greenberg, P E (2008) Management of anaemia A critical and systematic review of the cost effectiveness of erythropoiesis-stimulating agents Pharmacoeconomics, 26(2), 99-120 Easom, A (2006) The challenges of using serum ferritin to guide IV iron treatment practices in patients on hemodialysis with anemia Nephrology Nursing Journal, 33(5), 543551 Eknoyan, G & Levin, N W (2002) Clinical practice guidelines fro chronic kidney disease Evaluation, classification and stratification American Journal of Kidney Diseases, 39(2, Suppl l)pp S14-S266 Elliott, S , Pham, E , & Macdougall, I C (2008) Erythropoietins A common mechanism of action Experimental Hematology, 36, 1573-1584 85 Eustace, J A , & Coresh, J (2005) Chronic kidney disease Definition and epidemiology In J Pereira, M Sayegh & P Blake (Eds ), Chronic Kidney Disease, Dialysis & Transplantation (2nd ed , pp 1-19) Philadelphia Elsevier Saunders Faich,G , & Strobos, J (1999) Sodium ferric gluconate complex in sucrose Safer intravenous iron therapy than iron dextrans American Journal of Kidney Diseases, 33, 464-470 Fishbane, S & Berns, J S (2005) Hemoglobin cycling in hemodialysis patients treated with recombinant human erythropoietin Kidney International, 68, 1337-1343 Fishbane, S , Frei, G L & Maesaka, J (1995) Reduction in recombinant human erythropoietin doeses by the use of chronic intravenous iron supplementation American Journal of Kidney Diseases, 26, 41-46 Fishbane, S , Ungureanu, V D , Maesaka, J K , Kaupke, C J , Lim, V , & Wish, J (1996) The safety of intravenous iron dextran in hemodialysis patients American Journal of Kidney Diseases, 28(4), 529-534 Frauman, A C , & Gilman, C M (2001) Identification and measurement of nurse sensitive outcomes in pediatric nephrology nursing Nephrology Nursing Journal, 28(4), 395399 Fudin, R Jaichenko, J Shostak, A , Bennett, M , Gotloib, L (1998) Correction of uremic iron defiiciency anemia in hemodialyzed patients A prospective study Nephron, 79, 299-305 Fulton, B J , & Cameron, E M (1989) Perspectives on our beginnings 1962-1979 American Nephrology Nurses' Association Journal, 16(3), 201-203 86 Furuland, H , Linde, T , Ahlmen, J , Christensson, A , Stombom, U , & Damelson, B G (2003) A randomized controlled trial of haemoglobin normalization with epoetin alfa in pre-dialysis and dialysis patients Nephrology Dialysis Transplantation, 18,353361 Gerdtz, M F ,& Bucknall, T K (1999) Why we do the things we do applying clinical decision-making frameworks to triage practice Accident and Emergency Nursing, 7(1), 50-57 Gilbertson, D T , Ebben, J P , Foley, R N , Weinhandl, E D , Bradbury, B D , & Collins, A J (2008) Hemoglobin level variability associations with mortality Clinical Journal of the American Society of Nephrologists, 3, 133-138 Go, A S , Yang, J , Ackerson, L M , Lepper, K , Robbins, S Massie, B M & Shlipak, M G (2006) Hemoglobin level, chronic kidney disease and the risks of death and hospitalization in adults with chronic heart failure The anemia in chronic heart failure Outcomes and resource utilization (ANCHOR) study Circulation, 773,1323 Greene, T , Daugirdas, J , Depner, T , Allon, M , Beck, G , & Chumlea,C et al (2005) Association of achieved dialysis dose with mortality in the hemodialysis study an example of "dose targeting bias" Journal of the American Society of Nephrology, 75(11), 3371-3380 Hadorn, D C (1994) Use of algorithms in clinical guideline development In Clinical Practice Guideline Development Methodology Perspectives (pp 93-104) Rockville, MD US Agency for Health Care Policy and Research 87 Hartigan, M , Cesta, T , Mapes, D , Burrows-Hudson, S , Prathikanti, R , Lamb, G et al (2003) AMDP The anemia management demonstration project Development, implementation and testing of a multidisciphnary action plan for hemodialysis patients in the community Care management, 9(2), 19-28 Hodges, V M , Rainey, S , Lappin T R , & Maxwell, A P (2007) Pathophysiology of anemia and erythocytosis Critical Review of Oncology/Hematology, 64, 139-158 Hsu, C Y , Mcculloch, C E , Curhan, G C (2002) Epidemiology of anemia associated with chronic renal insufficiency among adults in the United States results from the Third National Health and Nutrition Examination survey Journal of the American Society of Nephrology, 73(2), 504-510 Ilott, I , Booth, A , & Patterson, M (2010) How do nurses, midwives and health visitors contribute to protocol-based care 9 International Journal of Nursing Studies, 47(6), 770-780 Irvine, D , Sidani, S & McGilhs Hall, L (1998) Linking outcomes to nurses' roles in health care Nursing Economics, 16( 2), 58-64 Ishani, A , Solid, C A , Weinhandl, E D , Gilbertson, D T , Foley, R N , & Collins, A J (2008) Association between number of months below K/DOQI haemoglobin target and risk of hospitalization and death Nephrology Dialysis Transplantation, 23, 16821689 Jenq, C C , Hsu, C , Huang, W , Chen, K , Lin, J , & Lin-Tan, D (2009) Serum ferritin levies predict all-cause and mfection-cause 1-year mortality in diabetic patients on maintenance hemodialysis The American Journal of the Medical Sciences, 337(3), 188-194 88 Jones, J (1988) Clinical reasoning in nursing Journal of Advanced Nursing, 73(2), 185-192 Kalantar-Zadeh, K , Kleiner, M , Dunne, E , Ahern, K , Nelson, M , Koslowe, R , & Luft, F C (1998) Total iron-binding capacity-estimated transferrin correlates with the nutritional subjective global assessment in hemodialysis patients American Journal of Kidney Diseases, 31(2), 263-2 Kalantar-Zadeh, K , Kopple, J D , Block, G , & Humphreys, M H (2001) A malnutritioninflammation score is correlated with morbidity and mortality in maintenance hemodialysis patients American Journal of Kidney Diseases, 38, 1251-1263 Kalantar-Zadeh, K , Lee, G H , MillerJ E , Streja, E , Jing, J , Robertson, J , et al (2009) Predictors of hyporesponsiveness to erythropoiesis-stimulating agents in hemodialysis patients American Journal of Kidney Diseases, 53(5), 823-834 Kalantar-Zadeh, K , Rodriguez, R A & Humphreys, M H (2004) Association between serum ferritin and measures of inflammation, nutrition and iron in haemodialysis patients Nehprology Dialysis Transplantation, 19, 141-149 Kane, R L , Shamliyan, T A , Mueller, C , Duval, S , & Wilt, T J (2007) The association of registered nurse staffing levels and patient outcomes systematic review and metaanalysis Medical Care, 45(12), 1195-1204 Kapoian, T (2008) Challenge of effectively using erythropoiesis-stimulating agents and intravenous iron American Journal of Kidney Diseases, 52(6, Suppl l)pp S21-S28 Kohn, L T , Corrigan, J M ,& Donaldson, M S (1999) To Err is Human Building a Safer Health System Washington, DC National Academic Press 89 Lacson, E , Ofsthun, N , & Lazarus, J M (2003) Effects of variability in anemia management on hemoglobin outcomes in ESRD American Journal of Kidney Diseases, 41(\), 111-124 Levin, A (2007) The treatment of anemia in chronic kidney disease Understandings in 2006 Current Opinion in Nephrology and Hypertension, 16, 261-21 \ Levin, A , Hemmelgarn, B , Culleton, B , Tobe, S , McFarlane, P , Ruzicka, M et al (2008) Guidelines for the management of chronic kidney disease Canadian Medical Association Journal, 779(11), 1154-1162 Locatelli, F , Andrulli, S , Memoh, B , Maffei, C Del Vecchio, L, Aterini, S , De Simon, W , Mandalan, A , Brunon, G , Amato, M , Cianciaruso, B & Zoccah, C (2006) Nutritional-inflammation status and resistance to erythropoietin therapy in haemodialysis patients Nephrology Dialysis Transplantation, 21, 991-998 Locatelli, F , Conte, F , & Marcelh, D (1998) The impact of hematocrit levels and erythropoetin treatment on overall and cardiovascular mortality and morbidity the experience of the Lombardy Dialysis Registry Nephrology, Dialysis, Transplantation, 13, 1642-1644 Locatelli, F , Del Vecchio, L , & Pozzoni, P (2007) Treating anemia at different stages of renal disease Journal of Nephrology, 20(12, Suppl 12) pp S33-S38 Ludwig, H & Strasser, K (2001) Symptomology of anemia Seminars in Oncology, 28(S), pp 7-14 Macdonald, J (2007) Addressing the challenges of renal anaemia How nurses can make a difference Journal of Renal Care, 33(4), 182-186 90 Macdougall, I C , Tucker, B , Thompson, J , Tomson, C R , Baker, L R , & Raine, A E (1996) A randomized controlled study of iron supplementation in patients treated with erythropoietin Kidney International, 50, 1694-1699 Magee, C C (2005) Evaluation of donors and recipients InB Pereira, M Sayegh, & P Blake (Eds ), Chronic Kidney Disease, Dialysis, & Transplantation (2nd ed , pp 623631) Philadephia Elsevier Saunders Mason, K , & McMahon, L P (1997) Normalization of haemoglobin in haemodialysis patients a comparative study [Abstract] Nephrology, 3(suppl 1), S305 McMahon, L P , Mason, K , Skinner, S L , Burge, C M , Gngg, L E , & Becker, G J (2000) Effects of haemoglobin normalization on quality of life and cardiovascular parameter in end-stage renal failure Nephrology Dialysis Transplantation, 15(9), 1425-1430 Michael, B , Coyne, D W , Fishbane, S , Folkert, V , Lynn, R , Nissenson, A R , Agarwal, R,Eschbach, J W , Fadem, S Z , Trout, J R , Strobos, J ,& Warnock, D G (2002) Sodium ferric gluconate complex in hemodialysis patients Adverse reactions compared to placebo and iron dextran Kidney International, 61, 1830-1839 Michael, M (2005) Anemia management protocols providing consistent hemoglobin outcomes Nephrology Nursing Journal, 32(4), 423-426 Miller, T , York, C , & Ryan, M (2005) Utilizing algorithms and pathways of care in allied health care practice The Internet Journal of Allied Health Sciences and Practice, 3(2), 1-18 91 Mix, T C , St Peter, W L , Ebben, J , Xue, J , Pereira, B J , Kausz, A , Collins, A J (2003) Hospitalization during advancing chronic kidney disease American Journal of Kidney Diseases, 42(5), 972-81 Mrayyan, M (2003) Nurse autonomy, nurse job satisfaction and client satisfaction with nursing care Their place in nursing data sets Nursing Leadership, 16(2), 74-83 Muirhead, N (2007) Renal policy Target hemoglobin time for change9 Nephrology News and Issues, 21(1), 57 Nassar, G M , Fishbane, S Ayus, J C (2002) Occult infection of old nonfunctioning arteriovenous grafts A novel cause of erythropoietin resistence and chronic inflammation in hemodialysis patients Kidney International, (Suppl 80) pp S49-S54 National Institutes of Health (2008) United States Renal Data System 2007 Annual Data Report American Journal ofKidney Diseases, 51(1, Suppl l)pp SI-S319 National Kidney Foundation (2002) Clinical practice guidelines for chronic kidney disease evaluation, classification and stratification American Journal of Kidney Diseases, 3P(Suppl l)pp S1-S266 National Kidney Foundation (2006) KDOQI clinical practice guidelines and clinical practice recommendations for anemia in chronic kidney disease American Journal of Kidney Diseases, 47(5, Suppl 3)pp S11-S145 National Kidney Foundation (2007) KDOQI clinical practice guideline and clinical practice recommendations for anemia in chronic kidney disease 2007 update of Hgb target American Journal of Kidney Diseases, 50(3), 471-530 92 National Quality Forum (2004) National voluntary consensus standards for nursing sensitive care An initial performance measure set Washington, DC National Quality Forum Naylor, M D (2007) Advancing the science in the measurement of healthcare quality influenced by nurses Medical Care Research and Review, 64(2, Suppl 1) pp 144S169S Needleman, J , Kurtzman, E T , & Kizer, K W (2007) Performance measurement of nursing care Medical Care Research and Review, 64(2, Suppl l)pp 10S-43S Nesrallah, G E , Blake, P , & Mendelssohn, D (2005) Modality options for end-stage renal disease care In B Pereira, M Sayegh & P Blake (Eds ), Chronic Kidney Disease, Dialysis & Transplantation (2nd ed , pp 9-306) Philadelphia Elsevier Saunders Newell, M (1996) Using Case Management to Improve Health Outcomes Gaithersburg, Md Aspen Nhan, J , Jensen, L , & McMahon, A (2007) Evaluation of an anemia algorithm in chronic hemodialysis patients The CANNT Journal, 17(3), 48-58 O'Mara, N (2008) Anemia in patients with chronic kidney disease Diabetes Spectrum, 21(1), 12-19 O'Riordan, E & Foley, R N (2000) Effects of anaemia on cardiovascular status Nephrology, Dialysis, Transplantation, 75(Suppl 3) pp 19-22 Painter, P , Moore, G , Carlson, L , Paul, S , Myll, J , & Phillips, W (2002) Effects of exercise training plus normalization of hematocrit on exercise capacity and healthrelated quality of life American Journal of Kidney Diseases, 39, 257-265 93 Parfrey, P S , Foley, R N , Wittreich, B H , Sullivan, D J , Zagari, M J , & Frei, D (2005) Double-blind comparison of full and partial anemia correction in incident hemodialysis patients without symptomatic heart disease Journal of the American Society of Nephrology, 16(1), 2180-2189 Patel, T V , Robinson, K & Singh, A K (2007) Is it time to reconsider subcutaneous administration of epoetin9 Nephrology News & Issues, 27(11), 57-67 Pereira, B , Sayegh, M , & Blake, P (Eds ) (2005) Chronic kidney disease, Dialysis & Transplantation (2nd ed) Philadelphia Elselvier Saunders Phrommintikul, A , Haas, S J , Elsik, M , & Krum, H (2007) Mortality and target hemoglobin concentrations in anaemic patients with chronic kidney disease treated with erythropoietin a meta-analysis Lancet, 369, 381-388 Pisoni, R , Bragg-Gresham, J L , Young, E W , Akizawa, T Asano, Y , Locatelh,F , Bommer, J , Cruz, J M , Kerr, P G , Mendelssohn, D C , Held, P J , & Port, F K (2004) Anemia management and outcomes from 12 countries in the dialysis outcomes and practice patterns study(DOPPS) American Journal of Kidney Diseases, 44(1), 94-111 Pruett, B , Johnson, S , & O'Keefe, N (2007) Improving IV iron and anemia management in the hemodialysis setting A collaborative CQI approach Nephrology Nursing Journal, 34(2), 206-213 Ritz, E , & Eisenhardt, A (2000) Early epoetin treatment in patients with renal insufficiency Nephrology, Dialysis, Transplantation, 75(Suppl 3) pp 40-44 Rozen-Zvi, B , Gafter-Gvilh, A , Paul, M , Leibovici, L , Shpilberg, O , & Gafter, U (2008) Intravenous versus oral iron supplementation for the treatment of anemia in CKD 94 Systematic review and meta-analysis American Journal of Kidney Diseases, 52(5), 897-906 Sarnak, M J , Levey, A S , Schoolwerth, A C , Coresh, J , Culleton, B , Hamm, L L , et al (2003) Kidney disease as a risk factor for development of cardiovascular disease a statement from the American Heart Association Councils on kidney in cardiovascular disease, high blood pressure research, clinical canology and epidemiology and prevention Circulation, 108(11), 2154-2169 Schwartz, S ,& Griffin, T (1986) Medical Thinking The Psychology of Medical Judgement and Decision-making New York Springer-Verlag Sengolge, G , Horl, W H , & Sunder-Plassman, G (2005) Intravenous iron therapy welltolerated, yet not harmless European Journal of Clinical Investigation, 35(Suppl 3) pp 46-51 Sidani, S & Braden, C J (1998) Evaluating Nursing Intervention A theory-driven approach Thousand Oaks Sage Publications Silverberg, D S , laina, A , Wexler, D , & Blum, M (2001) The pathological consequences of anemia Clinical Laboratory and Haematology, 23, 1-6 Silverberg, D S , Wexler, D , Blum, M , Tchebiner, J Z , Sheps, D , & Keren, G , et al (2003) The effect of correction of anaemia in diabetics and non-diabetics with severe resistant congestive heart failure and chronic renal failure by subcutaneous erythropoietin and intravenous iron Nephrology, Dialysis, Transplantation, 18, 141146 95 Singh A K &Hertello, P (2005) The benefits of IV iron therapy in treating anemia in patients with renal disease and cormorbid cardiovascular disease Nephrology Nursing Journal, 32(2), 199-206 Singh, A K , Coyne, D W , Shapiro, W , & Rizkala, A R (2007) Predictors of the response to treatment in anemic hemodialysis patients with high serum ferritin and low transferrin saturation Kidney International, 71, 1163-1171 Singh, A K , Szczech, L , Tang, K L Barnhart, H Sapp, S , Wolfson, M , & Reddan, D (2006) Correction of anemia with epoetin alfa in chronic kidney disease The New England Journal of Medicine, 355(20), 2085-2098 Skikne, B & Cook, J (1992) The effect of enhanced erythropoiesis on iron absorption Journal of Laboratory Clinical Medicine, 120, 746-751 Skikne, B , Ahauwalia, N , Fergusson, B, Chonko, A & Cook J (1998) Effects of recombinant human erythropoietin on iron absorption in chronic renal failure Blood, 92, 24B Steinbrook, R (2006) Haemoglobin concentrations in chronic kidney disease The Lancet, 368,2191-2193 Strippoli, G , Navaneethan, S , & Craig, J (2006) Haemoglobin and haematocrit targets for the anemia of chronic kidney disease (Review) Cochrane Database of Systematic Reviews, 4, 1-59 Tabobondung, L (2007) Northern Health Aboriginal Health Services Plan 2007-2010 Retrieved June 19, 2010, from http //www northernhealth ca/Portals/0/Your_Health/Programs/Abonginal_Health 96 Tranter, S , Martinez, Y , & Rayment, G (2006) A nurse-initiated iron management protocol for patients on hospital hemodialysis Renal Society of Australasia Journal, 2(1), 3032 van Tellingen, A , Grooteman, M P , Schoorl, M , Bartels, P C , Schoorl, M , van der Ploeg, T , et al (2002) Intercurrent clinical events are predictive of plasma c-reactive protein levels in hemodialysis patients Kidney International, 62(2), 632-638 Vanhaecht, K , de Witte, K , & Sermeus, W (2007) The care process organization triangle a framework to better understand how clinical pathways work Journal of Integrated Care Pathways, 77(2), 54-61 Walker, R & Pussell, B A (2007) Fluctuations in haemoglobin levels in haemodialysis patients receiving intravenous epoetin alfa or intravenous darbepoetin alfa Nephrology, 72,448-451 Werner, D E ,& Levey, A S (2007) Dialysis facility ownership and epoetin dosing in hemodialysis patients an overview American Journal of Kidney Diseases, 50(3), 340-353 Whiteside, C (1994) Diabetic nephropathy Successful treatment depends upon early diagnosis Diabetes News, 2(1), 8 Yeun, J Y ,& Depner, T A (2005) Principles of hemodialysis InB Pereira, M Sayegh, P Blake (Ed ), Chronic Kidney Disease, Dialysis, & Transplantation (2nd ed , pp 307340) Philadelphia Elsevier Saunders Young, T K , Kaufert, J M , McKenzie, J K , Hawkins, A , & O'Neil, A (1989) Excessive burden of end state renal disease among Canadian Indians A national survey American Journal of Public Health, 79(6), 756-758 97 Appendix A RENAL ANEMIA MANAGEMENT PROTOCOL RP010 ASSESS HEMOGLOBIN STATUS Target Hgb 110 125 Notify physician if increase tn Hgb greater than 20 from last blood work Follow subsequent orders ASSESS IRON STATUS Refer to page 3 for Infufer and Venofer Refer to page 4 for Ferrlecit Hgb 126 or Higher and/or Large Increase in Hgb {less than 20) The following protocol on order of physician transfers anemia management of hemodialysis patients to non physician start 0 e RNs and pharmacists) The following protocol is intended to serve as a guide and cannot replace clinical judgment The recommendations included may be inappropnate for specific clinical situations (e g patients with hemochromatosis PRCA allergy to IV iron or erythropoietic agents or patients who feel better with higher hemoglobins etc) When in doubt please ask a physician or renal pharmacist Hgb 110 - 125 and Hgb Stable (no rise or fall in Hgb greater than 20) Receiving EHRT? Hgb 109 or Lower and/or Large Decrease in Hgb (greater than20) Notify (vID for Large Decrease in Hgb > a S3 EHRT on hold/discontinued 00 £> No EHRT required Continue routine monitoring of Hgb every 4 weeks Reduce dose provided there have been no dose reductions in the previous S weeks See dose adjustment schedule (Note if there has been a dose reduction in the previous 5 weeks maintain current dose) Recheck Hgb in 2 weeks Continued on page 2 HOLD EHRT Collect Hgb in two weeks and then Hgb status Restart EHRT at e reduced dose based on dosage before hold SeedD5» adjustironi schedule Continue routine monilonrg of Hgb every 4 weeks NoEHRT required Continue routine monitoring of Hgb every 4 weeks Maintain dose of EHRT and continue routine monitoring of Hgb every 4 weeks ASSESS IRON STATUS Refer to page 3 for Infufer and Venofer Refer to page 4 for Ferrlecit NOTE Erythropoietic hormone replacement therapy (EHRT) refers to both Aranesp and Eprex 72 Restart EHRT at reduced dose based on dose before hold See dose adjustment schedule Notify physician if fall En Hgb greater than 20 from last btoodwork Recheck Hgb in 4 weeks NORTHERN RENAL PROGRAM Hgb 109 or lower and/or Large decrease in Hgb (greater than 20) Notify physician if Hgb is less than 90 or if fall in Hgb is greater than 20 from last Woodwork Follow subsequent orders ACCESS IRON STATUS Refer to page 3 for Dexiron and Venofer Refer to page 4 for Ferrlecit NO Notify physician for initiation of erythropoietic hormone replacement therapy SO Increase dose provided there has been no dose increase in the previous 5 weeks* See dose adjustment schedule Notify nephrologist if patient s hemoglobin is not within target after 3 consecutive dose increases NOTE If them has been a dose reduction in the previous 5 weeks maintain current dose Target Hgb 110 - 125 should be achieved within 2 to 4 months of initiating therapy Continue monitonng of Hgb every 4 weeks or as clinically indicated Monitor Hgb every 4 weeks or as clinically indicated (no earlier than 2 weeks after a dose change) ASSESS IRON STATUS Refer to page 3 for Infufer and Venofer Refer to page 4 for Ferneat *lf Aranesp™ dose is 150mcg/week or if Eprex® dose is 10 000 units 3x per week for B weeks or more and hemoglobin less than 110 contact nephrologist Follow this algorithm f o r Iron Dextran and Iron Sucrose (must be given after a test dose) (Refer t o page 4 f o r Ferrlecit) ASSESS IRON STATUS- NOTE If TSat and ferritin values indicate both an overload and Ihe need for iron replacement (i e TSat less than 20% and Ferritin greater than 500) please contact physician for further orders TSat greater than 50% OR Ferritin greater than 500 TSat less than 20% AND OR Femtin less than 200 REPLACEMENT HOLD PO/IV IRON Collect iron bloodwork in 4 weeks and ASSESS IRON STATUS TSat over 20% Ferntm 200 - 500 OF IRON Initiate IV iron at a dose of 100mg2x/wkfor5 weeks Check iron bloodwork 2 week after last dose and ASSESS IRON STATUS NOTE Patient not to receive more than 4 courses or iron replacement in a row Notify physician if iron indicies remain low after 20 consecutive weeks of iron 100 mg 2x/wk If just completed If receiving maintenance iron continue current maintenance dosage 100 m g 2 x / w k f o r 5 weeks initiate maintenance IV iron at 100 mg q 2 weeks If Iron Is currently on hold restart Iron at half the frequency but same dose (e g 100 mg q2 monthly) This dose becomes the patient s new maintenance dose MONITOR TSat and ferritin every 3 months and 'If iron bloodwork ever appears very unusual compared to previous results (e g with replacement of Iron stores TSat aoes from areater than 20% to less than 50%) reoeat bloodwork and reassess iron Note Iron maintenance dose refers to 100 q monthly unless otherwise indicated Infufer refers to Iron Dextran Venofer refers to Iron Sucrose If n o t receiving IV iron initiate maintenance IV iron at 100 mg q 2 weeks NORTHERN RENAL PROGRAM Follow this algorithm for F e r r l e c i t (Refer t o page 3 for Iron Dextran a n d Iron Sucrose) ASSESS IRON STATUS" NOTE If TSat and ferritin values indicate both an overload and Ihe need for iron replacement (i e TSat less than 2 0 % and Ferritin greater than 500) please contact physician for further orders TSat greater than 50% OR Ferntin greater than 500 OR AND Ferntin less than 200 REPLACEMENT HOLD PO/IV IRON Collect iron bloodwork in 4 weeks and ASSESS IRON STATUS TSat over 2 0 % TSat less than 20% Ferritin 200 - 500 OF IRON Initiate IV iron at a dose of 125 mg 2x/wk for 5 weeks Check iron bloodwork 2 weeks after last dose and ASSESS IRON STATUS NOTE Patient not to receive more than 4 courses or iron replacement in a row Notify physician if iron indlcies remain low after 20 consecutive weeks of iron 126 mg 2x/wk if receiving maintenance i r o n continue current maintenance dosage If j u s t completed 125mg 2x/wkfor 5 w e e k s initiate maintenance IV iron at 62 5 mg q 2 weeks If iron is currently on hold restart Iron at half the frequency but same dose (e g 125 mgq monthly) This dose becomes the patient s new maintenance dose MONITOR TSat and ferntin every 3 months and *lf Iron bloodwork ever appears very unusual compared to previous results (e g with replacement of iron stores TSat aoes from areater than 20% to less than 50%) reoeat bloodwork and reassess iron Note Iron maintenance dose refers to 125 q monthly unless otherwise indicated If not receiving IV iron initiate maintenance IV iron at 125 mg q 2 weeks NORTHERN RENAL PROGRAM DOSE ADJUSTMENT SCHEDULE FOR PATIENTS USING EPREX (ERYTHROPOIETIN) Increase Dose To Reduce Dose To 2,000 units/wk 3,000 units/wk Discontinue Eprex 3,000 units/wk 4,000 units/wk 2,000 units/wk 4,000 units/wk 6,000 units/wk 3,000 units/wk 6,000 units/wk 4,000 units 2x/wk 4,000 units/wk 4,000 IU 2x/wk 5,000 IU 2x/wk 6,000 lU/wk 5,000 IU 2x/wk 6,000 IU 2x/wk 4 000 IU 2x/wk 6,000 IU 2x/wk 8,000 IU 2x/wk 5,000 IU 2x/wk 8,000 IU 2x/wk 9,000 IU 2x/wk 6,000 IU 2x/wk 9,000 IU 2x/wk 10,000 IU 2x/wk 8,000 IU 2x/wk 10,000 IU 2x/wk 8,000 IU 3x/wk 9,000 IU 2x/wk 8,000 IU 3x/wk 10,000 IU 3x/wk 10,000 IU2x/wk 10,000 IU 3x/wk Contact nephrologist 8,000 IU 3x/wk Curt en t Dose The following chart provides the necessary changes for most dose increments or dose reductions If the patient's dosing does not fall into one of the intervals, contact the nephrologist Eprex is available in multidose vials (20,000 units/mL) . • • • • • . . 2,000 units (0 1 mL) 3,000 units (0 15 mL) 4,000 units (0 2 mL) 5,000 units (0 25 mL) 6,000 units (0 3 mL) 8, 000 units (0 4 mL) 9,000 units (0 45 mL) and 10,000 units (0 5 mL) Developed by PGRH Renal Pharmacist Approved by Regional P & T Committee on May 17, 2007 May proposal Page 77 of 80 ~^(" northern health NORTHERN RENAL PROGRAM DOSE ADJUSTMENT SCHEDULE FOR PATIENTS USING ARANESP (DARBEPOIETIN) The following chart provides the necessary changes for most dose increments or dose reductions If the patient's dosing does not fall into one of the intervals, contact the nephrologist The following Aranesp pre-filled syringes are available • 10 meg • 20 meg • 30 meg • 40 meg • 50 meg « 60 meg • 80 meg • 100 meg and • 150 meg Increase Dose To Reduce Dose To 10 mcg/q2wk 10 meg/wk Discontinue Aranesp™ 10 meg/wk 20 meg/wk 10 mcg/q2wk 20 meg/wk 30 meg/wk 10 meg/wk 30 meg/wk 40 meg/wk 20 meg/wk 40 meg/wk 50 meg/wk 30 meg/wk 50 meg/wk 60 meg/wk 40 meg/wk 60 meg/wk 80 meg/wk 50 meg/wk 80 meg/wk 100mcg/wk 60 meg/wk 100 meg/wk 150 meg/wk 80 meg/wk 150 meg/wk Contact nephrologist 100 meg/wk Cm rent Dose/WK Appendix B 3 C R e n a l A g = n C / Scorn MO-16 IOS S u r i r d i i r V .incline- 3C \r,z i TJ i60J,63l 3-1 • rax i W H . l u t mail acpriucneos _cc . Consent for Use of Data -_Patient Record and Outcome Management information System As an individual requiring renal services in the Province of British Columbia, I, understand that information regarding my clinical, laboratory, and treatment regimens will be entered in a provincial database This information will not be accessible by any members of the public and will remain anonymous This information may be used for statistical purposes, but I will not be identified as an individual at any time The purpose of this information is to ensure state of the art delivery of care and to provide my caregivers with efficient, timely and accurate information about my health today and into the future By signing this form I acknowledge that my caregivers, the statistical analysts and data managers in charge of the Patient Information System may have access to my personal information SIGNATURE DATE WITNESS 104 Appendix C From Lee Er (Renal) [ler@bcpra ubc ca] Sent September-03-08 9 41 AM To sushila@shaw ca Subject RE Anemia management study To Whom It May Concern, This is to confirm that any access or release of PROMIS data for research purposes will be proceeded upon the approval of the ethics board on the project A copy of the ethics' approval must be faxed to BC Renal Agency at 604-806-8005 Only data related to the research project for the prespecified hospital/region, without any subject identifier information (e g PHN, Names, Date of Birth), will be released Thank you, Mirek Piaseczny Director of Information and Statistics BC Renal Agency An agency of the Provincial Health Services Authority 570 14 - 1081 Burrard Street Vancouver, BC,V6Z1Y6 tel (604)806-8647 fax (604)806-8005 105 Appendix D UNIVERSITY OF NORTHERN BRITISH COLUMBIA RESEARCH ETHICS BOARD MEMORANDUM To* CC* Sushila Saunders Martha MacLeod From: Henry Harder, Chair Research Ethics Board Date. June 8, 2009 Re: E2009.0602.094 Anemia management protocols in the care of hemodialysis patients Examining patient outcomes Thank you for submitting the above-noted proposal to the Research Ethics Board proposal has been approved Your We are pleased to issue approval for the above named study for a period of 12 months from the date of this letter Continuation beyond that date will require further review and renewal of REB approval Any changes or amendments to the protocol or consent form must be approved by the Research Ethics Board Good luck with your research Sincerely, Henry Harder 106 Appendix E Northern Health Corporate Office 600-299 Victoria Street, Prince George, BC V2L 5B8 Telephone (250) 565-2649, Fax (250) 565-2640, www northernhealth ca June 29, 2009 File #RRC-2009-0020 Sushila Saunders 4435 Otway Rd Prince George BC V2M 6X6 Dear Sushila RE: Anemia management protocols in the care of hemodialysis patients: Examining patient outcomes On behalf of the Northern Health Research Review Committee, I would like to thank you for your submission titled "Anemia management protocols in the care of hemodialysis patients Examining patient outcomes " Your study has met the requirements of the Northern Health Research Review Committee and you may proceed Enjoy your work' We look forward to hearing about your findings Sincerely, Suzanne Johnston, Chair, NH Research Review Committee Vice President, Academic Affairs and Chief Nursing Officer SJ/sw CC Laurie Ledger File 107 Appendix F Levene's test for Equality of Variances for Hgb t-test Month October mean Hgb 1 127 November mean Hgb 1 975 December mean Hgb 341 January mean Hgb 042 February mean Hgb 071 March mean Hgb 1559 April mean Hgb 456 May mean Hgb 687 June mean Hgb 611 July mean Hgb 1029 August mean Hgb 321 September mean Hgb 3 146 291 163 561 837 790 215 501 409 436 313 572 079 Effect Size of the Difference Between the Means of Hemoglobin Levels Month Mean difference SD Cohen's d October -140 15 18 09 November 14 00 -43 03 December 3 43 13 17 26 January 14 12 02 -27 February 3 63 14 10 26 March 01 14 16 01 April 13 19 14 -1 85 May 14 37 93 06 June 99 14 07 07 July 12 70 -197 16 August 12 07 14 -169 September 12 71 24 -3 11 108 Independent Samples t-test for Equality of Means Month t October mean Hgb Equal variances -488 assumed November mean Hgb Equal variances -165 assumed December mean Hgb Equal variances 1009 assumed January mean Hgb -097 Equal variances assumed February mean Hgb Equal variances 1326 assumed March mean Hgb Equal variances 003 assumed April mean Hgb Equal variances -700 assumed May mean Hgb Equal variances 340 assumed June mean Hgb Equal variances 346 assumed July mean Hgb Equal variances -723 assumed August mean Hgb Equal variances -635 assumed September mean Hgb Equal variances -1265 assumed 109 df 100 p (2 tailed) 627 100 870 80 316 101 923 100 188 86 997 94 485 102 735 94 730 88 472 87 527 99 209 Appendix G Hemoglobin Categories-October Control group Protocol group Count Expected Count % within case % within Oct Hgb % of Total Count Expected Count % within case % withm Oct Hgb % of Total 75-109 110-125 22 27 212 27 7 37 3 45 8 57 4 61 1 126-145 10 10 0 169 58 8 Total 59 59 0 100 0 59 0 22 0 14 14 8 34 1 38 9 27 0 20 19 3 48 8 42 6 10 0 7 70 17 1 41 2 59 0 41 410 100 0 41 0 14 0 20 0 70 41 0 75-109 18 16 7 29 5 64 3 176 10 113 24 4 35 7 98 110-125 30 32 9 49 2 54 5 29 4 25 22 1 610 45 5 24 5 126-145 13 11 4 21 3 68 4 12 7 6 76 14 6 316 59 Total 61 610 100 0 59 8 59 8 41 41 0 100 0 40 2 40 2 Hemoglobin Categories-November Control group Protocol group Count Expected Count % within case % within NovHgb % of Total Count Expected Count % within case % within NovHgb % of Total Hemoglobin Categories- December Control group Protocol group Count Expected Count % within case % within DecHgb % of Total Count Expected Count % within case % within DecHgb % of Total 75-109 18 198 310 66 7 22 8 9 72 42 9 33 3 114 110-125 27 25 0 46 6 79 4 34 2 7 90 33 3 20 6 89 126-145 13 13 2 22 4 72 2 16 5 5 48 23 8 27 8 63 Total 58 58 0 100 0 73 4 73 4 21 210 100 0 26 6 26 6 Hemoglobin Categories-January Control group Protocol group Count Expected Count % within case % within Jan_Hgb % of Total Count Expected Count % within case % within JanHgb % of Total 111 75-109 17 169 28 8 60 7 173 11 11 1 28 2 39 3 112 110-125 25 24 7 42 4 610 25 5 16 16 3 41 0 39 0 16 3 126-145 17 175 28 8 58 6 173 12 11 5 30 8 414 12 2 Total 59 59 0 100 0 60 2 60 2 39 39 0 100 0 39 8 39 8 Hemoglobin Categories- February Control group Protocol group Count Expected Count % within case % within FebHgb % of Total Count Expected Count % within case % within FebHgb % of Total 75-109 14 14 5 24 6 56 0 110-125 30 30 2 52 6 57 7 126-145 13 12 2 22 8 619 Total 57 57 0 100 0 58 2 14 3 11 105 26 8 44 0 30 6 22 218 53 7 42 3 13 3 8 88 195 38 1 58 2 41 41 0 100 0 41 8 11 2 22 4 82 41 8 75-109 17 159 28 3 73 9 110-125 26 29 0 43 3 619 195 6 71 22 2 26 1 29 9 16 13 0 59 3 38 1 19 5 5 68 185 22 7 69 0 27 27 0 100 0 310 69 184 57 310 Hemoglobin Categories- March Control group Protocol group Count Expected Count % within case % within Mar_Hgb % of Total Count Expected Count % within case % within MarHgb % of Total 126-145 Total 17 60 152 60 0 28 3 100 0 69 0 77 3 Hemoglobin Categories- April Control group Protocol group Count Expected Count % within case % within AprHgb % of Total Count Expected Count % within case % within AprHgb % of Total 75-109 15 144 25 0 65 2 110-125 31 30 6 51 7 63 3 126-145 14 15 0 23 3 58 3 Total 60 60 0 100 0 62 5 15 6 8 86 22 2 34 8 32 3 18 184 50 0 36 7 14 6 10 90 27 8 41 7 62 5 36 36 0 100 0 37 5 83 18 8 104 37 5 110-125 126-145 14 29 30 9 119 48 3 23 3 55 8 70 0 Total 60 60 0 100 0 59 4 Hemoglobin Categories-May Control group Protocol group Count Expected Count % withm case % within MayHgb % of Total Count Expected Count % within case % within MayHgb % of Total 75-109 17 172 28 3 58 6 16 8 12 11 8 29 3 414 28 7 23 21 1 56 1 44 2 13 9 6 81 14 6 30 0 59 4 41 410 100 0 40 6 11 9 22 8 59 40 6 Hemoglobin Categories-June Control group Protocol group Count Expected Count % within case % within JunHgb % of Total Count Expected Count % within case % within JunHgb % of Total 75-109 110-125 126-145 Total 22 24 14 60 22 1 23 4 14 5 60 0 36 7 40 0 23 3 100 0 62 9 64 9 60 9 63 2 23 2 25 3 14 7 63 2 13 13 35 12 9 13 6 9 85 37 1 37 1 25 7 100 0 37 1 35 1 39 1 36 8 13 7 13 7 95 36 8 75-109 22 22 4 36 1 66 7 110-125 26 25 8 42 6 68 4 126-145 13 12 9 213 68 4 Total 61 610 100 0 67 8 24 4 11 106 37 9 33 3 28 9 12 122 414 316 14 4 6 61 20 7 316 67 8 29 29 0 100 0 32 2 122 36 7 133 42 2 67 211 32 2 100 0 35 0 Hemoglobin Categories-July Control group Protocol group Count Expected Count % within case % within JulHgb % of Total Count Expected Count % within case % within JulHgb % of Total % of Total 114 Hemoglobin Categories-August Control group Protocol group Count Expected Count % within case % within AugHgb % of Total Count Expected Count % within case % within AugHgb % of Total 75-109 22 20 0 37 9 73 3 110-125 27 27 3 46 6 65 9 126-145 9 10 7 15 5 56 3 Total 58 58 0 100 0 66 7 25 3 8 10 0 27 6 26 7 310 14 13 7 48 3 34 1 10 3 7 53 24 1 43 8 66 7 29 29 0 100 0 33 3 92 16 1 80 33 3 75-109 24 20 1 39 3 72 7 110-125 25 29 9 410 510 126-145 12 110 19 7 66 7 Total 61 610 100 0 610 24 0 9 12 9 23 1 27 3 25 0 24 19 1 61 5 49 0 12 0 6 70 154 33 3 610 39 39 0 100 0 39 0 90 24 0 60 39 0 Hemoglobin Categories-September Control group Protocol group Count Expected Count % within case % within Sep_Hgb % of Total Count Expected Count % within case % within SepHgb % of Total Chi-Square Test of Lower than, on Target and Higher than Target Hemoglobin Data Month n Chi-Square p (2 sided) df value October 133 2 945 100 1454 2 November 102 483 December 1270 2 530 79 044 2 978 January 98 February 98 174 2 917 March 87 1951 2 377 262 2 877 April 96 May 101 1223 2 542 June 95 099 2 951 2 July 90 029 985 August 87 1393 2 498 September 4 200 2 122 100 116 Appendix H On/Off Target Hemoglobin Categories-October 110-125 Control group Count 27 Expected Count 28 1 % within case 44 3 % within October target 57 4 % of Total 26 5 Protocol group Count 20 Expected Count 189 % within case 48 8 % within October target 42 6 % of Total 196 On/Off Target Hemoglobin Categories-November 110-125 Control group Count 30 Expected Count 32 9 % within case 49 2 % within November 54 5 target % of Total 29 4 Count Protocol group 25 Expected Count 22 1 % within case 610 % within November 45 5 target % of Total 24 5 117 <110or >125 34 32 9 55 7 61 8 33 3 21 22 1 51 2 38 2 20 6 Total 61 610 100 0 59 8 59 8 41 41 0 100 0 40 2 40 2 <110or>125 31 28 1 50 8 66 0 Total 61 610 100 0 59 8 30 4 16 189 39 0 34 0 59 8 41 41 0 100 0 40 2 15 7 40 2 On/Off Target Hemoglobin Categories December 110-125 27 Control group Count 25 3 Expected Count 44 3% % within case % within December target 79 4% 32 9% % of Total 7 Protocol group Count 87 Expected Count 33 3% % within case % within December target 20 6% 8 5% % of Total <110or>125 34 35 7 55 7% 70 8% 41 5% 14 12 3 66 7% 29 2% 17 1% Total 61 610 100 0% 74 4% 74 4% 21 210 100 0% 25 6% 25 6% On/Off Target Hemoglobin Categories-January Control group Count Expected Count % within case % within January target % of Total Protocol group Count Expected Count % within case % within January target % of Total 110-125 25 24 3 41 0% 61 0% 24 3% 16 16 7 38 1% 39 0% 15 5% 118 <110or>125 36 36 7 59 0% 58 1% 35 0% 26 25 3 61 9% 41 9% 25 2% Total 61 610 100 0% 59 2% 59 2% 42 42 0 100 0% 40 8% 40 8% On/Off Target Hemoglobin Categories-February Control group Count Expected Count % within case % within Feb target % of Total Protocol group Count Expected Count % withm case % within Feb target % of Total 110-125 30 30 6 50 0% 57 7% 29 4% 22 214 52 4% 42 3% 21 6% <110or>125 30 29 4 50 0% 60 0% 29 4% 20 20 6 47 6% 40 0% 19 6% Total 60 60 0 100 0% 58 8% 58 8% 42 42 0 100 0% 41 2% 41 2% <110 or > 125 Total 35 319 57 4% 61 610 100 0 On/Off Target Hemoglobin Categories March Control group Count Expected Count % within case 110-125 26 29 1 42 6% % % within March target % of Total Protocol group Count Expected Count % within case 61 9% 29 5% 16 12 9 59 3% 76 1% 39 8% 11 14 1 40 7% 69 3% 69 3% 27 27 0 100 0 % % withm March target % of Total 38 1% 18 2% 119 23 9% 12 5% 30 7% 30 7% On/Off Target Hemoglobin Categories-April 110-125 31 Control group Count 30 6 Expected Count % within case 51 7% % within April target 63 3% % of Total 32 3% Protocol group Count 18 184 Expected Count % within case 50 0% % within April target 36 7% % of Total 18 8% <110or>125 29 29 4 48 3% 61 7% 30 2% 18 176 50 0% 38 3% 18 8% Total 60 60 0 100 0% 62 5% 62 5% 36 36 0 100 0% 37 5% 37 5% 110-125 -<110or>125 32 29 30 5 30 5 47 5% 52 5% 61 5% 55 8% 27 9% 30 8% 20 23 21 5 215 53 5% 46 5% 44 2% 38 5% Total 61 610 100 0% 58 7% 58 7% 43 43 0 100 0% 41 3% On/Off Target Hemoglobin Categories-May Control group Count Expected Count % within case % within May target % of Total Protocol group Count Expected Count % within case % within May target 120 On/Off Target Hemoglobin Categories June Control group Count Expected Count % within case % within June target % of Total Protocol group Count Expected Count % within case % within June target % of Total 110-125 <110or>125 24 37 23 5 37 5 39 3% 60 7% 62 7% 64 9% 25 0% 38 5% 22 13 13 5 215 37 1% 62 9% 35 1% 37 3% 13 5% 22 9% Total 61 610 100 0% 63 5% 63 5% 35 35 0 100 0% 36 5% 36 5% On/Off Target Hemoglobin Categories July Control group Protocol group Count Expected Count % within case % within July target % of Total Count Expected Count % within case % within July target % of Total 110-125 <110or>125 26 35 25 8 35 2 42 6% 57 4% 68 4% 67 3% Total 61 610 100 0% 67 8% 28 9% 12 12 2 41 4% 31 6% 38 9% 17 16 8 58 6% 32 7% 67 8% 29 29 0 100 0% 32 2% 13 3% 18 9% 32 2% 121 On/Off Target Hemoglobin Categories August 110-125 <110or>125 Control group 27 Count 31 26 7 Expected Count 31 3 46 6% % withm case 53 4% % within August 65 9% 64 6% target % of Total 30 3% 34 8% Protocol group 14 Count 17 14 3 16 7 Expected Count 45 2% % within case 54 8% 34 1% % within August 35 4% target 15 7% % of Total 19 1% On/Off Target Hemoglobin Categories September 110-125 Control group 25 Count 29 6 Expected Count 41 0% % within case % within September 51 0% target 24 8% % of Total Protocol group 24 Count 194 Expected Count % within case 60 0% % within September 49 0% target 23 8% % of Total 122 Total 58 58 0 100 0% 65 2% 65 2% 31 310 100 0% 34 8% 34 8% <110or>125 36 314 59 0% 69 2% Total 61 610 100 0% 60 4% 35 6% 16 20 6 40 0% 30 8% 60 4% 40 40 0 100 0% 39 6% 15 8% 39 6% Chi-Square Test of On/Off Categorical Hemoglobin Data Month n Chi-Square value October 102 0 201 November 102 1373 December 82 0 769 January 103 0 087 February 102 0 056 March 88 2 076 April 96 0 025 May 104 0 357 June 96 0 046 July 90 0 012 August 89 0016 September 101 3 498 123 df p (2 sided) ][ I ]I I I 1 1 I 1 I 1[ 11 654 241 381 769 813 150 874 550 831 911 900 061 Appendix I Levene's Test for Equality of Variances for Epoetin Alfa Mean Dose in Control and Protocol Groups Month Equal variances assumed 1454 231 October mean Epoetin Alfa dose 2 262 assumed 136 November mean Epoetin Alfa dose assumed 2 639 108 December mean Epoetin Alfa dose 154 assumed 2 063 January mean Epoetin Alfa dose assumed 3 857 053 February mean Epoetin Alfa dose assumed 2 850 095 March mean Epoetin Alfa dose not assumed 7 677 007 April mean Epoetin Alfa dose not assumed 8 364 005 May mean Epoetin Alfa dose not assumed 032 4 749 June mean Epoetin Alfa dose 3 072 assumed 083 July mean Epoetin Alfa dose assumed 2 246 137 August mean Epoetin Alfa dose assumed September mean Epoetin Alfa dose 2 240 138 Effect Size of the Difference Between the Means of Epoetin Alfa Mean Month difference SD Cohen's d 1 14 October 3 00 38 November 1 12 3 11 36 1 52 December 3 08 49 January -125 3 19 39 February -4 08 3 24 126 March -5 86 3 34 175 April -102 3 28 32 May -103 331 31 -1 11 34 June 3 30 July -107 32 3 35 August -105 3 40 31 September -9 41 3 38 2 78 124