- Email: [email protected]
Anemia and blood transfusion practices in the critically ill: A prospective cohort review
Anemia and blood transfusion practices in the critically ill: A prospective cohort review Jissy Thomas, RN, MN,a Louise Jensen, RN, PhD,a Susan Nahirniak, MD, FRCP (C),b and R.T. Noel Gibney, MB, FRCP (C)c
BACKGROUND: Nearly 75% of critically ill patients develop anemia in the intensive care unit (ICU). Anemia can be treated with red blood cell (RBC) transfusions, although evidence suggests that lower hemoglobin levels are tolerated in the critically ill. Despite such recommendations, variation exists in clinical practice. METHODS: A prospective cohort was assessed for anemia and RBC transfusion practices in 100 consecutive adults admitted to our General Systems ICU. RESULTS: The prevalence of anemia in this cohort was 98%. Mean blood loss via phlebotomy was 2510.3 mL per patient per day. The RBC transfusion rate for the ICU stay was 40%, increasing to 70% in patients whose ICU stay was >7 days. The mean pretransfusion level of hemoglobin was 7.350.47 mg/dL for the total cohort, and 8.20.65 mg/dL for those with a history of cardiovascular disease. CONCLUSION: Anemia was common in this critically ill cohort, with hemoglobin levels continuing to drop with ICU stay. Pretransfusion hemoglobin levels were lower than reported by others, yet the RBC transfusion rate was comparable. There was no association between anemia and phlebotomy practices in our ICU. (Heart LungÒ 2010;39:217–225.)
N
early 75% of critically ill patients develop anemia in the intensive care unit (ICU).1 Among the causes of anemia in the critically ill,2 phlebotomy for diagnostic testing is a major source of blood loss.3 Technical improvements in diagnostic measurements and restraints in blood sampling have resulted in less phlebotomy blood loss and resultant iatrogenic anemia.4,5 However, volumes of iatrogenic blood loss and rates of anemia vary along with institutional practices. More than 50% of ICU patients receive at least one unit of red blood cells (RBCs) to maintain hemoglobin and hematocrit levels at >10 mg/dL and 30%, respectively.6 This 10/30 rule, first proposed in 1942 by From the aFaculty of Nursing, University of Alberta, Edmonton, Alberta, Canada; bDepartment of Laboratory Medicine and Pathology, University of Alberta Hospital, Alberta, Edmonton, Alberta, Canada; and cDivision of Critical Care Medicine, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada. Corresponding author: Louise Jensen, RN, PhD, Faculty of Nursing, University of Alberta, Edmonton, Alberta T6G 2G3, Canada. E-mail: [email protected] 0147-9563/$ – see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.hrtlng.2009.07.002
HEART & LUNG VOL. 39, NO. 3
Adams and Lundy, was accepted as the transfusion trigger.7 Historically, anemia was assumed risky, particularly among patients with cardiovascular disease, and RBC transfusion was deemed beneficial. However, the [TRICC] trial suggests that lower hemoglobin levels are well-tolerated in most critically ill patients, including those with cardiovascular disease.8,9 Therefore, blood should be transfused considering factors such as age, disease severity, and specific diagnoses.6,7,10,11 Yet in a United Kingdom survey of transfusion practices with the critically ill, Boralessa et al12 reported RBC transfusion triggers of 9 to 10 mg/dL. Another survey of European ICU transfusion practices found that the mean pretransfusion hemoglobin level of patients not actively bleeding was 8.5 mg/dL.2 This was similar to the finding of Corwin et al.13 Thus despite the recommendations of the TRICC trial,9 there is variation in clinical practice.6,7,11,14 In a contemporary Canadian survey, Hebert et al15 reported that critical-care physicians have adopted a lower transfusion threshold of 7 mg/dL. However, this change in practice was not uniform. Furthermore, Turgeon et al16 found significant variation in transfusion practices among Canadian anesthesiologists within specific patient
www.heartandlung.org
217
Anemia and blood transfusion in the critically ill
Thomas et al
Table I Subject characteristics (n=100) Characteristics
Frequency
Age 18-30 years 31-50 years 51-70 years 71-90 years
7 25 49 19
Gender Male Female
57 43
Admission diagnosis Respiratory disorders Gastrointestinal disorders Neurologic disorders Cardiovascular disorders Genitourinary disorders Multisystem disorders Other
41 17 8 7 3 9 15
subgroups. Concerns regarding this and the limited blood supply led to an evaluation of blood transfusion practices in our General Systems ICU.
BACKGROUND Two main factors contribute to anemia in the critically ill: insufficient production of red blood cells,1,6,17 and blood loss.4,18 Underproduction anemia is commonly referred to as the anemia of chronic inflammatory disease. It is attributable to abnormal serum erythropoietin (EPO) concentrations, with a minimal reticulocyte response to endogenous EPO because of inhibition of the EPO gene by inflammatory mediators such as interleukin-1a, tumor necrosis factor-a, tumor growth factor-b, and interleukin-1b.1,18 Second, iatrogenic anemia occurs because of blood loss through blood draws, occult gastrointestinal bleeding, and invasive techniques. Tarpey and Lawler19 investigated ICU patients, and found a mean blood loss per day of 66 mL, and a total blood loss of 336 mL. Dale and Pruett20 studied patients in a medical ICU and in a medical unit, and found that ICU patients
218
www.heartandlung.org
Characteristics
Comorbidities Coronary Artery Disease history Hypertension Liver insufficiency Diabetes mellitus COPD Chronic renal failure CHF Other Admission hemoglobin 6.0-8.0 mg/dL 8.1-10.0 mg/dL 10.1-12.0 mg/dL 12.1-14.0 mg/dL 14.1-18.2 mg/dL Admission APACHE II cores 1-20 21-30 >30 Length of ICU stay 1-3 days 4-7 days 8-12 days 13-20 days 21-28 days
Frequency
16 14 12 8 5 4 3 32 7 32 32 22 7 75 21 4 26 39 19 9 7
had more blood draws per day (3 vs 1.5), and a higher mean total blood loss (550 mL vs 208 mL). More recently, von Ahsen et al4 found that diagnostic blood loss ranged from 20 to 41 mL, accounting for 17% of total blood loss. Similarly, Ba et al21 reported a phlebotomy blood loss of 40 mL per day, and Vincent et al22 in the Anemia and Blood Transfusion in Critical Care (ABC) Trial documented an average daily phlebotomy volume of 41 mL. A primary treatment of anemia in the critically ill has been RBC transfusions. However, studies reviewing RBC transfusion practices, based on a variety of criteria, estimate a proportion of unnecessary RBC transfusions ranging from 4% to 66%.2,23 Vincent et al2 reported RBC transfusion rates of 37% in European ICUs, and 73% with ICU length of stay >1 week, with a mean pretransfusion hemoglobin level of 8.5 mg/ dL. Corwin et al24 reported that 85% of patients with an ICU stay of >1 week received at least one unit of RBCs during their stay. In a later study,25 they found that most RBC transfusions in the ICU were not associated with acute blood loss, but rather a constant transfusion rate of 2 to 4units/week. In the Anemia
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
120
10
100
8
80
6
60
4
40
2
20
0
Number of Patients
Mean Hemoglobin (mg/dL)
Mean Hemoglobin Per Patient Per Day (mg/dL)
12
0 1
3
5
7
9
11 13 15 17 19 21 23 25 27 ICU Days
Mean Hemoglobin (mg/dL)
Number of Patients
Fig 1 Mean hemoglobin per patient per day (mg/dL).
Table II Phlebotomy practices Mode
Median
Phlebotomy rate per patient Per day 5.3 Total LOS 15.0 Blood loss per patient (mL) Per day 18.25 Total LOS 117.0 Phlebotomy discard volume per patient (mL) Per day 7.7 Total LOS 21.0
3.5 19.5
Mean (SD)
3.7 (1.8) 29.6 (29.3)
Range
2-15 3-155
22.7 156.5
24.7 (10.3) 223.6 (183.4)
8-59 32-944
12.4 68.0
13.2 (6.3) 106.0 (103.5)
3-54 0-528
LOS, length of stay.
and Blood Transfusion in Critical Care (CRIT) trial,26 70% of ICU patients developed anemia within 48 hours of admission, and 44% received at least one unit of RBCs during their ICU stay, with a mean pretransfusion hemoglobin level of 8.6 mg/dL. The decision to transfuse is usually based on patient factors such as volume status, acuteness of anemia, severity of symptoms, age, and presence of comorbid conditions, and particularly cardiovascular disease.27 Hebert et al28 found that critically ill patients with cardiac disease exhibited a trend toward increased mortality when hemoglobin levels were <9.5 mg/dL, compared with other diagnoses. Rivers
HEART & LUNG VOL. 39, NO. 3
et al29 found that early intervention in severe sepsis and septic shock, by transfusing RBCs to a hematocrit level $30%, led to reduced in-hospital mortality, a higher central venous oxygen saturation, a lower lactate level, and a lower base deficit. Precise indications for RBC transfusion in other situations remain controversial in the critically ill because of a lack of data from clinical trials. Existing guidelines for RBC transfusion and other blood components are derived from animal experiments, and from observations of patients undergoing resuscitation, surgery, and anesthesia.30,31 Adam and Lundy32 first recommended the administration of RBCs for hemoglobin levels of 8 to 10 mg/dL in
www.heartandlung.org
219
Anemia and blood transfusion in the critically ill
Thomas et al
Mean Number of Blood Draws Per Patient Per Day 120
7
100
5 80 4 60 3 40
Number of Patients
Mean Number of Blood Draws
6
2 20
1
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days
Mean Number of Blood Draws
Number of Patients
Fig 2 Mean number of blood draws per patient per day.
60
120
50
100
40
80
30
60
20
40
10
20
Number of Patients
Mean Blood Loss (mls)
Mean Phlebotomy Blood Loss Per Patient Per Day
0
0 1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days Mean Blood Loss (mls)
Number of Patients
Fig 3 Mean phlebotomy blood loss per patient per day.
the perioperative period, when surgical patients with hemoglobin levels of <10 mg/dL were not observed to do well. In the 1980s, a majority of surveyed anesthesiologists believed a preoperative hemoglobin
220
www.heartandlung.org
level of 9 to 10 mg/dL to be mandatory.33,34 Surgenor et al7 recommended maintaining a hemoglobin level of 10 to 12 mg/dL in patients with cardiovascular disorders.
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
Mean RBC Transfusions Per Patient Per Day (Units) 3.5
120
100
2.5 80 2 60 1.5 40
Number of Patients
Mean RBC Transfusions (Units)
3
1
20
0.5
0
0 1 2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days Mean RBC Transfusion (Units)
Num b er of Patients
Fig 4 Mean RBC transfusions per patient per day (units).
To address the RBC transfusion needs of the critically ill, Hebert et al conducted a series of trials to evaluate the effects of restrictive and liberal RBC transfusion strategies on mortality and morbidity.9,28,35 The largest of these was the TRICC Trial,9 with 838 critically ill patients whose hemoglobin levels were <9 mg/dL within 72 hours after admission to the ICU: 418 were randomly assigned to a restrictive strategy of RBC transfusion (hemoglobin <7 mg/dL), and 420 were assigned to a liberal strategy (RBC transfusion if hemoglobin levels were <10 mg/dL). The 30-day mortality was similar in both groups, although the rates were significantly lower with the restrictive transfusion strategy among patients with Acute Physiology and Chronic Health Evaluation (APACHE) II score of #20, and among patients <55 years of age, but not among patients with clinically significant cardiac disease. They concluded that a restrictive RBC transfusion strategy was at least as effective as, and possibly superior to, a liberal strategy in the critically ill, with the exception of acute myocardial infarction and unstable angina patients. Although that was one of the largest randomized, controlled trials to evaluate outcomes in critically ill patients treated with a restrictive transfusion approach, blood administration practices have not changed much in the last decade. Furthermore, the recent Sepsis Occurrence in Acutely Ill Patients Study36 did not find that blood transfusions were associated with increased
HEART & LUNG VOL. 39, NO. 3
mortality in critically ill patients. Clearly, evidence is controversial regarding transfusion safety or a specific transfusion trigger.
PURPOSE OF STUDY Our primary objective was to examine current RBC transfusion practices for critically ill adults admitted to our General Systems ICU. We addressed the specific questions: 1) What is the prevalence of anemia? 2) What is the rate of RBC transfusion? 3) What are the reported indications for RBC transfusions? 4) What are the factors associated with anemia and RBC transfusions?
METHODS A cohort was used to examine RBC transfusion practices in our ICU, and to assess the relationship between anemia and RBC transfusion practices in critically ill adults. Consecutive patients (n = 100) admitted to the ICU over a 2-month period were included. Excluded were patients: 1) less than 18 years of age, 2) who remained in the ICU for less than 24 hours, 3) with chronic endstage renal disease receiving exogenous erythropoietin, 4) of the Jehovah’s Witnesses, who declined blood products, and 5) with confirmed primary hematologic disease, including recent bone marrow transplantation or pancytopenia after chemotherapy. Ours is a 30-bed adult
www.heartandlung.org
221
Anemia and blood transfusion in the critically ill
Thomas et al
Table III RBC transfusion practices RBC transfusions per patient (units) Mode
Per day Total length of stay
1.4 2
Median
1.5 2
Mean (SD)
Range
1.5 (.4) 1.66 (3.25)
1-2 0-19
RBCs units per transfusion order
Per day Total length of stay
Mode
Median
1.25 2
1.4 2
Mean (SD)
Range
1.38 (.34) 2 (.5)
1-2 2-2
Pretransfusion hemoglobin (mg/dL) and number (percentage) of transfusions
<6.0 6.0-7.0 7.1-8.0 8.1-9.0 9.1-10.0
3 (3.6%) 29 (35%) 32 (38.5%) 15 (18%) 4 (5%) Units per order and units transfused
Units per order
1 2 3 4 5 Total
Frequency (%)
26 (31) 52 (63) 3 (4) 1 (1) 1 (1) 83 (100)
General Systems ICU in a large Canadian academic health center that admits approximately 1200 patients per year with a mean APACHE II score of 22. It is staffed by intensivists, medical residents, nurses, respiratory therapists, physiotherapists, dieticians, pharmacists, social workers, service attendants, and students. The ICU does not have RBC transfusion protocols. Data-collection sheets were added to the patient’s health record for nurses to prospectively record date/time of blood collection, discard volumes, volumes of blood drawn for diagnostic purposes, and blood-transfusion events. Variables included rate of anemia (hemoglobin <12 mg/dL for women, and <14 mg/dL for men, per our laboratory), frequency of phlebotomy per patient per day, associated volume of blood loss via phlebotomy per patient per day, rate of RBC transfusions, number of RBC units (volumes ranging from 250 to 350 mL, with hematocrit levels at 50% to 80%) per transfusion, admission hemoglobin level, pretransfusion hemoglobin level (hemoglobin level within 3 hours before transfusion), and primary indication for RBC
222
www.heartandlung.org
Number of units
26 114 9 4 5 158
transfusion, as reported by the attending physician. Each patient’s age, gender, admission diagnosis, comorbidities, admission APACHE II scores, and length of ICU stay were also recorded. Approval for this study was obtained from our Health Research Ethics Board. Descriptive statistics were computed for all variables. To determine if a relationship existed between phlebotomy blood loss, anemia, and RBC transfusions, and to examine relationships between age, gender, comorbidities, illness severity (APACHE II score), and length of ICU stay in the rate of anemia and RBC transfusions, Pearson’s correlations were used.
RESULTS Patient characteristics The mean age of patients was 56.5 14.8 (SD) years (range, 18 to 85 years), with 19% (n = 19) above 70 years of age. Men accounted for 57% of this cohort. The most frequent admitting diagnosis was of complications related to the respiratory system
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
Table IV Indications for RBC transfusions Indication for transfusion
Table V Number of RBC units transfused, based on admission diagnosis
Frequency Percentage
Anemia Acute bleeding Cardiac history Hypoxemia APACHE II score
38 33 7 3 2
46 40 8 4 2
(41%). Of the comorbidities documented, a cardiac history accounted for 16% of patients. The mean admission APACHE II score was 15 7, (SD) with a range of 3 to 36. The mean ICU length of stay was 7.7 6.6 (SD) days, with a range of 1 to 28 days; 35% stayed longer than 7 days (Table I).
Admission diagnosis
Respiratory disorder Multisystem disorder Gastrointestinal disorder Cardiovascular disorder Neurologic disorder Genitourinary disorder Other
Number of RBC units Percentage
54
34
32
20.25
32
20.25
18
11.5
6
4
2
1
14
9
Course of anemia The mean admission hemoglobin level was 10.8 0.22 (SD) mg/dL (range, 6.3-18.2 mg/dL), with 71% having an admission hemoglobin level <12 mg/dL, and 39% having an admission hemoglobin level <10 mg/dL (Table I). The majority of patients (86%) were anemic on admission to the ICU. The incidence of anemia during their stay in the ICU was 98%. By ICU day 8, 97% of the patients were anemic, and by ICU day 13, 100% were anemic (Fig 1). The number and volume of blood samples drawn varied widely (Table II). The median number of phlebotomies per patient was 19.5 for total ICU stay, with a mean of 3.5 1.04 (SD) per patient per day (Fig 2). The mean blood loss per phlebotomy was 7.6 6.3 (SD) mL, with a mean of 25 10.3 (SD) mL per patient per day, and 223.6 183.4 (SD) mL per patient for total ICU stay. The mean blood loss volume was greatest on the day of admission, averaging 51.4 17.1 (SD) mL, and the lowest on the day of discharge, averaging 16.6 11.3 (SD) mL (Fig 3).
Blood transfusion practices The mean number of RBC units transfused per patient per day was 1.5 0.4 (SD), for a mean of 1.66 0.5 (SD) during the total ICU stay, peaking on ICU days 5 and 27 (Fig 4). The mean number of RBC units per transfusion order per day was 1.38 0.34 (SD), and per total ICU stay, 2 0.5 (SD). There were 83 RBC transfusions during the study period, and in total, 158 RBC units were transfused (Table III). The majority (63%) of RBC transfusion orders were for 2 units per transfusion. The mean pretransfusion
HEART & LUNG VOL. 39, NO. 3
hemoglobin level was 7.3 0.47 (SD) mg/dL for this cohort. The mean pretransfusion hemoglobin level was 7.5 4.1 (SD) mg/dL for those with acute bleeding, and 7.2 0.54 (SD) mg/dL for those without acute bleeding. Some (39%) RBC transfusions occurred in the pretransfusion hemoglobin range of 7.1 to 8.0 mg/dL, with pretransfusion hemoglobin levels exceeding 9.0 mg/dL in only 5% of cases. Patients with a cardiac history were transfused at a hemoglobin level of 8.2 0.65 (SD) mg/dL, for a transfusion rate of 8%. Table IV summarizes the indications for RBC transfusion. Patients admitted to the ICU with respiratory disorders received up to 34% of transfusions, whereas those with cardiovascular disorders received 11.5% of transfusions (Table V). The RBC transfusion rate during the total ICU stay was 40%, with a daily transfusion rate of 13.6%. Patients whose length of stay was >7 days had an overall RBC transfusion rate of 70%. Transfusion rates were consistent across age groups, with 37.5% in patients aged <50 years, and 42.5% in those aged >50 years. There was no difference in RBC transfusion rate based on gender. Over 39% of patients received their first RBC transfusion within their first 3 days in ICU. The RBC transfusion rate was 11% at admission, and peaked on ICU day 19 at 71%. The mean length of stay for those who received RBC transfusions was 11.45 days, and for those who did not receive RBC transfusions, 5.2 days. Other blood products transfused were plasma (45%), albumin (35%), platelets, and cryoprecipitate during the study period. The rate of transfusion of other blood
www.heartandlung.org
223
Anemia and blood transfusion in the critically ill
products was 18% on ICU day 1, which peaked on ICU days 18 and 19 at 43%.
DISCUSSION The prevalence of anemia in this ICU cohort was similar to rates reported in other studies.2,4 Furthermore, patients with lower admission hemoglobin levels became more anemic as their length of ICU stay progressed (r = .66, P < .01). Patients with lower admission hemoglobin levels had a longer ICU stay (r = .16, P < .01) and higher admission APACHE II scores (r = .17, P < .01). Gender, age, admission diagnoses, comorbidities, and phlebotomy practices did not influence the rate of anemia. Pearl and Pohlman15 and Parillo1 documented the impact of nutritional deficiencies, decreased renal function, sepsis, and inappropriately low erythropoietin levels in the critically ill on anemia. It is not known whether an inadequate erythropoietin response contributed to the anemia in this cohort. Only 2% of patients who were anemic received oral iron therapy, whereas none received folate, vitamin B12, or exogenous erythropoietin. Anemia in the ICU setting was also attributed to frequent phlebotomies.24 In our ICU, the mean phlebotomy blood loss per patient per day was 25 10.3 (SD) mL, and anemia increased with length of ICU stay. The RBC transfusion rate in our ICU was 40%. As length of stay in the ICU increased, the RBC transfusion rate increased (r =.78, P < .01). Patients with an ICU stay >7 days had an overall transfusion rate of 70%, with a peak on ICU day 19. This is similar to rates reported in other studies.2,26 Severity of illness on ICU admission influenced RBC transfusion rates (r =.24, P < .05), with 62% of RBC transfusions occurring in patients with APACHE II scores of 13 to 24, and 19% in patients with APACHE II scores >25. Phlebotomy practices also influenced RBC transfusion rates. As the volume of blood loss increased through phlebotomies, the total number of RBC transfusions also increased (r =.80, P < .01). Of those patients who required RBC transfusions, 70% had a phlebotomy-related blood loss of >300 mL for their total ICU stay, which is equivalent to one unit of RBCs. Patients who received >5 units of RBCs during their ICU stay had a mean phlebotomy-related blood loss of >700 mL. The mean pretransfusion hemoglobin level was 7.35 .47 (SD) mg/dL for this cohort. The pretransfusion level of hemoglobin for patients with a cardiac history was 8.2 .65 (SD) mg/dL. These levels are lower than reported in other studies.2,7,11,26 This may be attributable to the findings of the TRICC trial.9 In our ICU, the pre-RBC transfusion hemoglobin levels were similar for those with
224
www.heartandlung.org
Thomas et al
and without acute bleeding, suggesting that RBC transfusion decisions were determined by hemoglobin levels rather than clinical diagnosis per se. Gender, age, and comorbidities did not play a significant role in reported RBC transfusion decisions.
Limitations Physicians in the ICU were aware that phlebotomy and RBC transfusion practices were being recorded, which could have led to temporary modifications in behavior. There was no validation of the accuracy of the reported volume of blood drawn for phlebotomy. Moreover, blood loss via hemodialysis filters and dialyzers was not documented, and RBC transfusions in surgical patients did not include transfusion events that may have occurred in the operating room. Clinical outcomes of patients transfused at lower pretransfusion levels of hemoglobin were not assessed.
CONCLUSION Anemia occurred in 98% of critically ill patients, and hemoglobin levels continued to drop with ICU stay. Despite phlebotomy-related blood loss, there was no correlation between anemia and phlebotomy practices. However, this finding should not preclude strategies to reduce phlebotomy-related blood loss. These include using multichannel microchemistry instruments for point-of-care testing, pediatric tubes for blood collection, inline blood-conservation devices to eliminate discard volumes, the timing and batching of phlebotomies, elimination of routine ICU standing orders for phlebotomies, and a daily assessment of frequency of phlebotomies. In this cohort, the pretransfusion hemoglobin level was 7.35 .47 mg/dL, a value lower than reported in other studies. However, the RBC transfusion rate was 40%, a rate similar to that in other studies.2,4,22 The management of anemia in the ICU may need to emphasize the complementary roles of decreasing phlebotomy-related blood loss, increasing RBC production with exogenous hematopoietic agents, and decreasing RBC transfusions, by using lower hemoglobin transfusion triggers. REFERENCES 1. Parillo J. Anemia management, and evidence-based critical care medicine. Crit Care Med 2001;29(Suppl):S139. 2. Vincent JL, Baron J, Reinhart K, et al. Anemia and blood transfusion in critically ill patients. JAMA 2002;288:1499-507. 3. Brown M, Whalen P. Red blood cell transfusion in critically ill patients: emerging risks and alternatives. Crit Care Nurse 2000;20(Suppl):1-14. 4. von Ahsen N, Muller C, Serke S, Frei U, Eckardt K. Important role of nondiagnostic blood loss and blunted erythropoietic
MAY/JUNE 2010
HEART & LUNG
Thomas et al
5.
6. 7. 8. 9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
response in the anemia of medical intensive care patients. Crit Care Med 2001;29:141-59. Zimmerman JE, Seneff MG, Sun X, Wagner DP, Knaus WA. Evaluating laboratory usage in the intensive care unit: patient and institutional characteristics that influence frequency of blood sampling. Crit Care Med 1997;25:737-48. Pohlman AS, Carven JH, Lindsay K. Conserving blood in the intensive care unit. Crit Care Nurse 2001;21(Suppl):1-14. Surgenor SD, Hampers MJ, Corwin HL. Is blood transfusion good for the heart? Crit Care Med 2001;29:442-4. Hebert PC, Yetisir E, Martin C, et al. Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med 2001;29:227-34. Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409-17. Canadian Medical Association. Expert working group: guidelines for red blood cell and plasma transfusion for adults and children. Can Med Assoc J 1997;156(Suppl 11): S1-24. Hebert PC, Wells G, Martin C, et al. Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998;26:482-7. Boralessa H, Rao M, Morgan C, et al. A survey of physicians’ attitudes to transfusion practice in critically ill patients in the UK. Anaesthesia 2002;57:584-8. Corwin HL, Gettinger A, Pearl RG, et al. Efficacy of recombinant human erythropoietin in critically ill patients: a randomized controlled trial. JAMA 2002;288:2827-35. Carson JL. Should patients in intensive care units receive erythropoietin [editorial]? JAMA 2002;288:2884-6. Pearl RG, Pohlman A. Understanding and managing anemia in critically ill patients. Crit Care Nurse 2002;26(Suppl):1-12. Hebert PL, Fergusson DA, Stather D, et al. Revisiting transfusion practices in critically ill patients. Crit Care Med 2005;33:7-12. Turgeon AF, Fergusson DA, Doucette S, et al. Red blood cell transfusion practices amongst Canadian anesthesiologists: a survey. Can J Anesth 2006;53:344-52. Corwin HL. Anemia of the critically ill: ‘‘acute’’ anemia of chronic disease. Crit Care Med 2000;28:3098-9. Tarpey J, Lawler PG. Iatrogenic anemia: a survey of venesection in patients in the intensive therapy unit. Anaesthesia 1990;45:396-8. Dale JC, Pruett SK. Phlebotomy: a minimalist approach. Mayo Clin Proc 1993;68:249-55.
HEART & LUNG VOL. 39, NO. 3
Anemia and blood transfusion in the critically ill 21. Ba VN, Bota DP, Melot C, Vincent J. Time course of hemoglobin concentrations in nonbleeding intensive care unit patients. Crit Care Med 2003;31:406-10. 22. Vincent JL, Baron J-F, Reinhart K, et al. Anemia and blood transfusion in critically ill patients. JAMA 2002;288: 1499-507. 23. Hebert PC, Schweitzer I, Calder L, Blajchman M, Giulivi A. Review of the clinical practice literature on allogenic red blood cell transfusion. Can Med Assoc J 1997;56(Suppl):S9-26. 24. Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion in the ICU: is there a reason? Chest 1995;108:767-71. 25. Corwin HL. Anemia in the critically ill: the role of erythropoietin. Semin Hematol 2001;38(Suppl 7):24-32. 26. Corwin HL, Gettinger A, Pearl RG, et al. The CRIT Study: anemia and blood transfusion in the critically ill: current clinical practice in the United States. Crit Care Med 2004;32:39-52. 27. Sherk PA, Granton JT, Kapral MK. Red blood cell transfusion in the intensive care unit. Intensive Care Med 2000;26:344-6. 28. Hebert PC, Wells G, Tweeddale M, et al. Does transfusion practice affect mortality in critically ill patients? Am J Respir Crit Care Med 1997;155:1618-23. 29. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368-77. 30. American College of Physicians. Practice strategies for elective red blood cell transfusion. Ann Intern Med 1992;116:403-6. 31. American Society of Anesthesiologists Task Force on Blood Component Therapy. Practice guidelines for blood component therapy. Anesthesiology 1996;84:732-47. 32. Adam RC, Lundy JS. Anesthesia in cases of poor risk: some suggestions for decreasing the risk. Surg Gynecol Obstet 1942;74:1011-9. 33. Stehling LC, Esposito B. An analysis of the appropriateness of intraoperative transfusion. Anesth Analg 1989;68(Suppl): S278-9. 34. Stehling LC, Ellison N, Faust RJ, Grotta AW, Moyers JR. A survey of transfusion practices among anesthesiologists. Vox Sang 1987;52:60-2. 35. Hebert PC, Wells G, Marshall J, et al. Transfusion requirements in critical care: a pilot study. JAMA 1995;273:1439-44. 36. Vincent JL, Sakr Y, Sprung C, et al. Are blood transfusions associated with greater mortality rates? Results of the Sepsis Occurrence in Acutely Ill Patients Study. Anesthesiology 2008; 108:31-9.
www.heartandlung.org
225
BACKGROUND: Nearly 75% of critically ill patients develop anemia in the intensive care unit (ICU). Anemia can be treated with red blood cell (RBC) transfusions, although evidence suggests that lower hemoglobin levels are tolerated in the critically ill. Despite such recommendations, variation exists in clinical practice. METHODS: A prospective cohort was assessed for anemia and RBC transfusion practices in 100 consecutive adults admitted to our General Systems ICU. RESULTS: The prevalence of anemia in this cohort was 98%. Mean blood loss via phlebotomy was 2510.3 mL per patient per day. The RBC transfusion rate for the ICU stay was 40%, increasing to 70% in patients whose ICU stay was >7 days. The mean pretransfusion level of hemoglobin was 7.350.47 mg/dL for the total cohort, and 8.20.65 mg/dL for those with a history of cardiovascular disease. CONCLUSION: Anemia was common in this critically ill cohort, with hemoglobin levels continuing to drop with ICU stay. Pretransfusion hemoglobin levels were lower than reported by others, yet the RBC transfusion rate was comparable. There was no association between anemia and phlebotomy practices in our ICU. (Heart LungÒ 2010;39:217–225.)
N
early 75% of critically ill patients develop anemia in the intensive care unit (ICU).1 Among the causes of anemia in the critically ill,2 phlebotomy for diagnostic testing is a major source of blood loss.3 Technical improvements in diagnostic measurements and restraints in blood sampling have resulted in less phlebotomy blood loss and resultant iatrogenic anemia.4,5 However, volumes of iatrogenic blood loss and rates of anemia vary along with institutional practices. More than 50% of ICU patients receive at least one unit of red blood cells (RBCs) to maintain hemoglobin and hematocrit levels at >10 mg/dL and 30%, respectively.6 This 10/30 rule, first proposed in 1942 by From the aFaculty of Nursing, University of Alberta, Edmonton, Alberta, Canada; bDepartment of Laboratory Medicine and Pathology, University of Alberta Hospital, Alberta, Edmonton, Alberta, Canada; and cDivision of Critical Care Medicine, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, Alberta, Canada. Corresponding author: Louise Jensen, RN, PhD, Faculty of Nursing, University of Alberta, Edmonton, Alberta T6G 2G3, Canada. E-mail: [email protected] 0147-9563/$ – see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.hrtlng.2009.07.002
HEART & LUNG VOL. 39, NO. 3
Adams and Lundy, was accepted as the transfusion trigger.7 Historically, anemia was assumed risky, particularly among patients with cardiovascular disease, and RBC transfusion was deemed beneficial. However, the [TRICC] trial suggests that lower hemoglobin levels are well-tolerated in most critically ill patients, including those with cardiovascular disease.8,9 Therefore, blood should be transfused considering factors such as age, disease severity, and specific diagnoses.6,7,10,11 Yet in a United Kingdom survey of transfusion practices with the critically ill, Boralessa et al12 reported RBC transfusion triggers of 9 to 10 mg/dL. Another survey of European ICU transfusion practices found that the mean pretransfusion hemoglobin level of patients not actively bleeding was 8.5 mg/dL.2 This was similar to the finding of Corwin et al.13 Thus despite the recommendations of the TRICC trial,9 there is variation in clinical practice.6,7,11,14 In a contemporary Canadian survey, Hebert et al15 reported that critical-care physicians have adopted a lower transfusion threshold of 7 mg/dL. However, this change in practice was not uniform. Furthermore, Turgeon et al16 found significant variation in transfusion practices among Canadian anesthesiologists within specific patient
www.heartandlung.org
217
Anemia and blood transfusion in the critically ill
Thomas et al
Table I Subject characteristics (n=100) Characteristics
Frequency
Age 18-30 years 31-50 years 51-70 years 71-90 years
7 25 49 19
Gender Male Female
57 43
Admission diagnosis Respiratory disorders Gastrointestinal disorders Neurologic disorders Cardiovascular disorders Genitourinary disorders Multisystem disorders Other
41 17 8 7 3 9 15
subgroups. Concerns regarding this and the limited blood supply led to an evaluation of blood transfusion practices in our General Systems ICU.
BACKGROUND Two main factors contribute to anemia in the critically ill: insufficient production of red blood cells,1,6,17 and blood loss.4,18 Underproduction anemia is commonly referred to as the anemia of chronic inflammatory disease. It is attributable to abnormal serum erythropoietin (EPO) concentrations, with a minimal reticulocyte response to endogenous EPO because of inhibition of the EPO gene by inflammatory mediators such as interleukin-1a, tumor necrosis factor-a, tumor growth factor-b, and interleukin-1b.1,18 Second, iatrogenic anemia occurs because of blood loss through blood draws, occult gastrointestinal bleeding, and invasive techniques. Tarpey and Lawler19 investigated ICU patients, and found a mean blood loss per day of 66 mL, and a total blood loss of 336 mL. Dale and Pruett20 studied patients in a medical ICU and in a medical unit, and found that ICU patients
218
www.heartandlung.org
Characteristics
Comorbidities Coronary Artery Disease history Hypertension Liver insufficiency Diabetes mellitus COPD Chronic renal failure CHF Other Admission hemoglobin 6.0-8.0 mg/dL 8.1-10.0 mg/dL 10.1-12.0 mg/dL 12.1-14.0 mg/dL 14.1-18.2 mg/dL Admission APACHE II cores 1-20 21-30 >30 Length of ICU stay 1-3 days 4-7 days 8-12 days 13-20 days 21-28 days
Frequency
16 14 12 8 5 4 3 32 7 32 32 22 7 75 21 4 26 39 19 9 7
had more blood draws per day (3 vs 1.5), and a higher mean total blood loss (550 mL vs 208 mL). More recently, von Ahsen et al4 found that diagnostic blood loss ranged from 20 to 41 mL, accounting for 17% of total blood loss. Similarly, Ba et al21 reported a phlebotomy blood loss of 40 mL per day, and Vincent et al22 in the Anemia and Blood Transfusion in Critical Care (ABC) Trial documented an average daily phlebotomy volume of 41 mL. A primary treatment of anemia in the critically ill has been RBC transfusions. However, studies reviewing RBC transfusion practices, based on a variety of criteria, estimate a proportion of unnecessary RBC transfusions ranging from 4% to 66%.2,23 Vincent et al2 reported RBC transfusion rates of 37% in European ICUs, and 73% with ICU length of stay >1 week, with a mean pretransfusion hemoglobin level of 8.5 mg/ dL. Corwin et al24 reported that 85% of patients with an ICU stay of >1 week received at least one unit of RBCs during their stay. In a later study,25 they found that most RBC transfusions in the ICU were not associated with acute blood loss, but rather a constant transfusion rate of 2 to 4units/week. In the Anemia
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
120
10
100
8
80
6
60
4
40
2
20
0
Number of Patients
Mean Hemoglobin (mg/dL)
Mean Hemoglobin Per Patient Per Day (mg/dL)
12
0 1
3
5
7
9
11 13 15 17 19 21 23 25 27 ICU Days
Mean Hemoglobin (mg/dL)
Number of Patients
Fig 1 Mean hemoglobin per patient per day (mg/dL).
Table II Phlebotomy practices Mode
Median
Phlebotomy rate per patient Per day 5.3 Total LOS 15.0 Blood loss per patient (mL) Per day 18.25 Total LOS 117.0 Phlebotomy discard volume per patient (mL) Per day 7.7 Total LOS 21.0
3.5 19.5
Mean (SD)
3.7 (1.8) 29.6 (29.3)
Range
2-15 3-155
22.7 156.5
24.7 (10.3) 223.6 (183.4)
8-59 32-944
12.4 68.0
13.2 (6.3) 106.0 (103.5)
3-54 0-528
LOS, length of stay.
and Blood Transfusion in Critical Care (CRIT) trial,26 70% of ICU patients developed anemia within 48 hours of admission, and 44% received at least one unit of RBCs during their ICU stay, with a mean pretransfusion hemoglobin level of 8.6 mg/dL. The decision to transfuse is usually based on patient factors such as volume status, acuteness of anemia, severity of symptoms, age, and presence of comorbid conditions, and particularly cardiovascular disease.27 Hebert et al28 found that critically ill patients with cardiac disease exhibited a trend toward increased mortality when hemoglobin levels were <9.5 mg/dL, compared with other diagnoses. Rivers
HEART & LUNG VOL. 39, NO. 3
et al29 found that early intervention in severe sepsis and septic shock, by transfusing RBCs to a hematocrit level $30%, led to reduced in-hospital mortality, a higher central venous oxygen saturation, a lower lactate level, and a lower base deficit. Precise indications for RBC transfusion in other situations remain controversial in the critically ill because of a lack of data from clinical trials. Existing guidelines for RBC transfusion and other blood components are derived from animal experiments, and from observations of patients undergoing resuscitation, surgery, and anesthesia.30,31 Adam and Lundy32 first recommended the administration of RBCs for hemoglobin levels of 8 to 10 mg/dL in
www.heartandlung.org
219
Anemia and blood transfusion in the critically ill
Thomas et al
Mean Number of Blood Draws Per Patient Per Day 120
7
100
5 80 4 60 3 40
Number of Patients
Mean Number of Blood Draws
6
2 20
1
0
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days
Mean Number of Blood Draws
Number of Patients
Fig 2 Mean number of blood draws per patient per day.
60
120
50
100
40
80
30
60
20
40
10
20
Number of Patients
Mean Blood Loss (mls)
Mean Phlebotomy Blood Loss Per Patient Per Day
0
0 1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days Mean Blood Loss (mls)
Number of Patients
Fig 3 Mean phlebotomy blood loss per patient per day.
the perioperative period, when surgical patients with hemoglobin levels of <10 mg/dL were not observed to do well. In the 1980s, a majority of surveyed anesthesiologists believed a preoperative hemoglobin
220
www.heartandlung.org
level of 9 to 10 mg/dL to be mandatory.33,34 Surgenor et al7 recommended maintaining a hemoglobin level of 10 to 12 mg/dL in patients with cardiovascular disorders.
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
Mean RBC Transfusions Per Patient Per Day (Units) 3.5
120
100
2.5 80 2 60 1.5 40
Number of Patients
Mean RBC Transfusions (Units)
3
1
20
0.5
0
0 1 2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
ICU Days Mean RBC Transfusion (Units)
Num b er of Patients
Fig 4 Mean RBC transfusions per patient per day (units).
To address the RBC transfusion needs of the critically ill, Hebert et al conducted a series of trials to evaluate the effects of restrictive and liberal RBC transfusion strategies on mortality and morbidity.9,28,35 The largest of these was the TRICC Trial,9 with 838 critically ill patients whose hemoglobin levels were <9 mg/dL within 72 hours after admission to the ICU: 418 were randomly assigned to a restrictive strategy of RBC transfusion (hemoglobin <7 mg/dL), and 420 were assigned to a liberal strategy (RBC transfusion if hemoglobin levels were <10 mg/dL). The 30-day mortality was similar in both groups, although the rates were significantly lower with the restrictive transfusion strategy among patients with Acute Physiology and Chronic Health Evaluation (APACHE) II score of #20, and among patients <55 years of age, but not among patients with clinically significant cardiac disease. They concluded that a restrictive RBC transfusion strategy was at least as effective as, and possibly superior to, a liberal strategy in the critically ill, with the exception of acute myocardial infarction and unstable angina patients. Although that was one of the largest randomized, controlled trials to evaluate outcomes in critically ill patients treated with a restrictive transfusion approach, blood administration practices have not changed much in the last decade. Furthermore, the recent Sepsis Occurrence in Acutely Ill Patients Study36 did not find that blood transfusions were associated with increased
HEART & LUNG VOL. 39, NO. 3
mortality in critically ill patients. Clearly, evidence is controversial regarding transfusion safety or a specific transfusion trigger.
PURPOSE OF STUDY Our primary objective was to examine current RBC transfusion practices for critically ill adults admitted to our General Systems ICU. We addressed the specific questions: 1) What is the prevalence of anemia? 2) What is the rate of RBC transfusion? 3) What are the reported indications for RBC transfusions? 4) What are the factors associated with anemia and RBC transfusions?
METHODS A cohort was used to examine RBC transfusion practices in our ICU, and to assess the relationship between anemia and RBC transfusion practices in critically ill adults. Consecutive patients (n = 100) admitted to the ICU over a 2-month period were included. Excluded were patients: 1) less than 18 years of age, 2) who remained in the ICU for less than 24 hours, 3) with chronic endstage renal disease receiving exogenous erythropoietin, 4) of the Jehovah’s Witnesses, who declined blood products, and 5) with confirmed primary hematologic disease, including recent bone marrow transplantation or pancytopenia after chemotherapy. Ours is a 30-bed adult
www.heartandlung.org
221
Anemia and blood transfusion in the critically ill
Thomas et al
Table III RBC transfusion practices RBC transfusions per patient (units) Mode
Per day Total length of stay
1.4 2
Median
1.5 2
Mean (SD)
Range
1.5 (.4) 1.66 (3.25)
1-2 0-19
RBCs units per transfusion order
Per day Total length of stay
Mode
Median
1.25 2
1.4 2
Mean (SD)
Range
1.38 (.34) 2 (.5)
1-2 2-2
Pretransfusion hemoglobin (mg/dL) and number (percentage) of transfusions
<6.0 6.0-7.0 7.1-8.0 8.1-9.0 9.1-10.0
3 (3.6%) 29 (35%) 32 (38.5%) 15 (18%) 4 (5%) Units per order and units transfused
Units per order
1 2 3 4 5 Total
Frequency (%)
26 (31) 52 (63) 3 (4) 1 (1) 1 (1) 83 (100)
General Systems ICU in a large Canadian academic health center that admits approximately 1200 patients per year with a mean APACHE II score of 22. It is staffed by intensivists, medical residents, nurses, respiratory therapists, physiotherapists, dieticians, pharmacists, social workers, service attendants, and students. The ICU does not have RBC transfusion protocols. Data-collection sheets were added to the patient’s health record for nurses to prospectively record date/time of blood collection, discard volumes, volumes of blood drawn for diagnostic purposes, and blood-transfusion events. Variables included rate of anemia (hemoglobin <12 mg/dL for women, and <14 mg/dL for men, per our laboratory), frequency of phlebotomy per patient per day, associated volume of blood loss via phlebotomy per patient per day, rate of RBC transfusions, number of RBC units (volumes ranging from 250 to 350 mL, with hematocrit levels at 50% to 80%) per transfusion, admission hemoglobin level, pretransfusion hemoglobin level (hemoglobin level within 3 hours before transfusion), and primary indication for RBC
222
www.heartandlung.org
Number of units
26 114 9 4 5 158
transfusion, as reported by the attending physician. Each patient’s age, gender, admission diagnosis, comorbidities, admission APACHE II scores, and length of ICU stay were also recorded. Approval for this study was obtained from our Health Research Ethics Board. Descriptive statistics were computed for all variables. To determine if a relationship existed between phlebotomy blood loss, anemia, and RBC transfusions, and to examine relationships between age, gender, comorbidities, illness severity (APACHE II score), and length of ICU stay in the rate of anemia and RBC transfusions, Pearson’s correlations were used.
RESULTS Patient characteristics The mean age of patients was 56.5 14.8 (SD) years (range, 18 to 85 years), with 19% (n = 19) above 70 years of age. Men accounted for 57% of this cohort. The most frequent admitting diagnosis was of complications related to the respiratory system
MAY/JUNE 2010
HEART & LUNG
Thomas et al
Anemia and blood transfusion in the critically ill
Table IV Indications for RBC transfusions Indication for transfusion
Table V Number of RBC units transfused, based on admission diagnosis
Frequency Percentage
Anemia Acute bleeding Cardiac history Hypoxemia APACHE II score
38 33 7 3 2
46 40 8 4 2
(41%). Of the comorbidities documented, a cardiac history accounted for 16% of patients. The mean admission APACHE II score was 15 7, (SD) with a range of 3 to 36. The mean ICU length of stay was 7.7 6.6 (SD) days, with a range of 1 to 28 days; 35% stayed longer than 7 days (Table I).
Admission diagnosis
Respiratory disorder Multisystem disorder Gastrointestinal disorder Cardiovascular disorder Neurologic disorder Genitourinary disorder Other
Number of RBC units Percentage
54
34
32
20.25
32
20.25
18
11.5
6
4
2
1
14
9
Course of anemia The mean admission hemoglobin level was 10.8 0.22 (SD) mg/dL (range, 6.3-18.2 mg/dL), with 71% having an admission hemoglobin level <12 mg/dL, and 39% having an admission hemoglobin level <10 mg/dL (Table I). The majority of patients (86%) were anemic on admission to the ICU. The incidence of anemia during their stay in the ICU was 98%. By ICU day 8, 97% of the patients were anemic, and by ICU day 13, 100% were anemic (Fig 1). The number and volume of blood samples drawn varied widely (Table II). The median number of phlebotomies per patient was 19.5 for total ICU stay, with a mean of 3.5 1.04 (SD) per patient per day (Fig 2). The mean blood loss per phlebotomy was 7.6 6.3 (SD) mL, with a mean of 25 10.3 (SD) mL per patient per day, and 223.6 183.4 (SD) mL per patient for total ICU stay. The mean blood loss volume was greatest on the day of admission, averaging 51.4 17.1 (SD) mL, and the lowest on the day of discharge, averaging 16.6 11.3 (SD) mL (Fig 3).
Blood transfusion practices The mean number of RBC units transfused per patient per day was 1.5 0.4 (SD), for a mean of 1.66 0.5 (SD) during the total ICU stay, peaking on ICU days 5 and 27 (Fig 4). The mean number of RBC units per transfusion order per day was 1.38 0.34 (SD), and per total ICU stay, 2 0.5 (SD). There were 83 RBC transfusions during the study period, and in total, 158 RBC units were transfused (Table III). The majority (63%) of RBC transfusion orders were for 2 units per transfusion. The mean pretransfusion
HEART & LUNG VOL. 39, NO. 3
hemoglobin level was 7.3 0.47 (SD) mg/dL for this cohort. The mean pretransfusion hemoglobin level was 7.5 4.1 (SD) mg/dL for those with acute bleeding, and 7.2 0.54 (SD) mg/dL for those without acute bleeding. Some (39%) RBC transfusions occurred in the pretransfusion hemoglobin range of 7.1 to 8.0 mg/dL, with pretransfusion hemoglobin levels exceeding 9.0 mg/dL in only 5% of cases. Patients with a cardiac history were transfused at a hemoglobin level of 8.2 0.65 (SD) mg/dL, for a transfusion rate of 8%. Table IV summarizes the indications for RBC transfusion. Patients admitted to the ICU with respiratory disorders received up to 34% of transfusions, whereas those with cardiovascular disorders received 11.5% of transfusions (Table V). The RBC transfusion rate during the total ICU stay was 40%, with a daily transfusion rate of 13.6%. Patients whose length of stay was >7 days had an overall RBC transfusion rate of 70%. Transfusion rates were consistent across age groups, with 37.5% in patients aged <50 years, and 42.5% in those aged >50 years. There was no difference in RBC transfusion rate based on gender. Over 39% of patients received their first RBC transfusion within their first 3 days in ICU. The RBC transfusion rate was 11% at admission, and peaked on ICU day 19 at 71%. The mean length of stay for those who received RBC transfusions was 11.45 days, and for those who did not receive RBC transfusions, 5.2 days. Other blood products transfused were plasma (45%), albumin (35%), platelets, and cryoprecipitate during the study period. The rate of transfusion of other blood
www.heartandlung.org
223
Anemia and blood transfusion in the critically ill
products was 18% on ICU day 1, which peaked on ICU days 18 and 19 at 43%.
DISCUSSION The prevalence of anemia in this ICU cohort was similar to rates reported in other studies.2,4 Furthermore, patients with lower admission hemoglobin levels became more anemic as their length of ICU stay progressed (r = .66, P < .01). Patients with lower admission hemoglobin levels had a longer ICU stay (r = .16, P < .01) and higher admission APACHE II scores (r = .17, P < .01). Gender, age, admission diagnoses, comorbidities, and phlebotomy practices did not influence the rate of anemia. Pearl and Pohlman15 and Parillo1 documented the impact of nutritional deficiencies, decreased renal function, sepsis, and inappropriately low erythropoietin levels in the critically ill on anemia. It is not known whether an inadequate erythropoietin response contributed to the anemia in this cohort. Only 2% of patients who were anemic received oral iron therapy, whereas none received folate, vitamin B12, or exogenous erythropoietin. Anemia in the ICU setting was also attributed to frequent phlebotomies.24 In our ICU, the mean phlebotomy blood loss per patient per day was 25 10.3 (SD) mL, and anemia increased with length of ICU stay. The RBC transfusion rate in our ICU was 40%. As length of stay in the ICU increased, the RBC transfusion rate increased (r =.78, P < .01). Patients with an ICU stay >7 days had an overall transfusion rate of 70%, with a peak on ICU day 19. This is similar to rates reported in other studies.2,26 Severity of illness on ICU admission influenced RBC transfusion rates (r =.24, P < .05), with 62% of RBC transfusions occurring in patients with APACHE II scores of 13 to 24, and 19% in patients with APACHE II scores >25. Phlebotomy practices also influenced RBC transfusion rates. As the volume of blood loss increased through phlebotomies, the total number of RBC transfusions also increased (r =.80, P < .01). Of those patients who required RBC transfusions, 70% had a phlebotomy-related blood loss of >300 mL for their total ICU stay, which is equivalent to one unit of RBCs. Patients who received >5 units of RBCs during their ICU stay had a mean phlebotomy-related blood loss of >700 mL. The mean pretransfusion hemoglobin level was 7.35 .47 (SD) mg/dL for this cohort. The pretransfusion level of hemoglobin for patients with a cardiac history was 8.2 .65 (SD) mg/dL. These levels are lower than reported in other studies.2,7,11,26 This may be attributable to the findings of the TRICC trial.9 In our ICU, the pre-RBC transfusion hemoglobin levels were similar for those with
224
www.heartandlung.org
Thomas et al
and without acute bleeding, suggesting that RBC transfusion decisions were determined by hemoglobin levels rather than clinical diagnosis per se. Gender, age, and comorbidities did not play a significant role in reported RBC transfusion decisions.
Limitations Physicians in the ICU were aware that phlebotomy and RBC transfusion practices were being recorded, which could have led to temporary modifications in behavior. There was no validation of the accuracy of the reported volume of blood drawn for phlebotomy. Moreover, blood loss via hemodialysis filters and dialyzers was not documented, and RBC transfusions in surgical patients did not include transfusion events that may have occurred in the operating room. Clinical outcomes of patients transfused at lower pretransfusion levels of hemoglobin were not assessed.
CONCLUSION Anemia occurred in 98% of critically ill patients, and hemoglobin levels continued to drop with ICU stay. Despite phlebotomy-related blood loss, there was no correlation between anemia and phlebotomy practices. However, this finding should not preclude strategies to reduce phlebotomy-related blood loss. These include using multichannel microchemistry instruments for point-of-care testing, pediatric tubes for blood collection, inline blood-conservation devices to eliminate discard volumes, the timing and batching of phlebotomies, elimination of routine ICU standing orders for phlebotomies, and a daily assessment of frequency of phlebotomies. In this cohort, the pretransfusion hemoglobin level was 7.35 .47 mg/dL, a value lower than reported in other studies. However, the RBC transfusion rate was 40%, a rate similar to that in other studies.2,4,22 The management of anemia in the ICU may need to emphasize the complementary roles of decreasing phlebotomy-related blood loss, increasing RBC production with exogenous hematopoietic agents, and decreasing RBC transfusions, by using lower hemoglobin transfusion triggers. REFERENCES 1. Parillo J. Anemia management, and evidence-based critical care medicine. Crit Care Med 2001;29(Suppl):S139. 2. Vincent JL, Baron J, Reinhart K, et al. Anemia and blood transfusion in critically ill patients. JAMA 2002;288:1499-507. 3. Brown M, Whalen P. Red blood cell transfusion in critically ill patients: emerging risks and alternatives. Crit Care Nurse 2000;20(Suppl):1-14. 4. von Ahsen N, Muller C, Serke S, Frei U, Eckardt K. Important role of nondiagnostic blood loss and blunted erythropoietic
MAY/JUNE 2010
HEART & LUNG
Thomas et al
5.
6. 7. 8. 9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
response in the anemia of medical intensive care patients. Crit Care Med 2001;29:141-59. Zimmerman JE, Seneff MG, Sun X, Wagner DP, Knaus WA. Evaluating laboratory usage in the intensive care unit: patient and institutional characteristics that influence frequency of blood sampling. Crit Care Med 1997;25:737-48. Pohlman AS, Carven JH, Lindsay K. Conserving blood in the intensive care unit. Crit Care Nurse 2001;21(Suppl):1-14. Surgenor SD, Hampers MJ, Corwin HL. Is blood transfusion good for the heart? Crit Care Med 2001;29:442-4. Hebert PC, Yetisir E, Martin C, et al. Is a low transfusion threshold safe in critically ill patients with cardiovascular diseases? Crit Care Med 2001;29:227-34. Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409-17. Canadian Medical Association. Expert working group: guidelines for red blood cell and plasma transfusion for adults and children. Can Med Assoc J 1997;156(Suppl 11): S1-24. Hebert PC, Wells G, Martin C, et al. Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998;26:482-7. Boralessa H, Rao M, Morgan C, et al. A survey of physicians’ attitudes to transfusion practice in critically ill patients in the UK. Anaesthesia 2002;57:584-8. Corwin HL, Gettinger A, Pearl RG, et al. Efficacy of recombinant human erythropoietin in critically ill patients: a randomized controlled trial. JAMA 2002;288:2827-35. Carson JL. Should patients in intensive care units receive erythropoietin [editorial]? JAMA 2002;288:2884-6. Pearl RG, Pohlman A. Understanding and managing anemia in critically ill patients. Crit Care Nurse 2002;26(Suppl):1-12. Hebert PL, Fergusson DA, Stather D, et al. Revisiting transfusion practices in critically ill patients. Crit Care Med 2005;33:7-12. Turgeon AF, Fergusson DA, Doucette S, et al. Red blood cell transfusion practices amongst Canadian anesthesiologists: a survey. Can J Anesth 2006;53:344-52. Corwin HL. Anemia of the critically ill: ‘‘acute’’ anemia of chronic disease. Crit Care Med 2000;28:3098-9. Tarpey J, Lawler PG. Iatrogenic anemia: a survey of venesection in patients in the intensive therapy unit. Anaesthesia 1990;45:396-8. Dale JC, Pruett SK. Phlebotomy: a minimalist approach. Mayo Clin Proc 1993;68:249-55.
HEART & LUNG VOL. 39, NO. 3
Anemia and blood transfusion in the critically ill 21. Ba VN, Bota DP, Melot C, Vincent J. Time course of hemoglobin concentrations in nonbleeding intensive care unit patients. Crit Care Med 2003;31:406-10. 22. Vincent JL, Baron J-F, Reinhart K, et al. Anemia and blood transfusion in critically ill patients. JAMA 2002;288: 1499-507. 23. Hebert PC, Schweitzer I, Calder L, Blajchman M, Giulivi A. Review of the clinical practice literature on allogenic red blood cell transfusion. Can Med Assoc J 1997;56(Suppl):S9-26. 24. Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion in the ICU: is there a reason? Chest 1995;108:767-71. 25. Corwin HL. Anemia in the critically ill: the role of erythropoietin. Semin Hematol 2001;38(Suppl 7):24-32. 26. Corwin HL, Gettinger A, Pearl RG, et al. The CRIT Study: anemia and blood transfusion in the critically ill: current clinical practice in the United States. Crit Care Med 2004;32:39-52. 27. Sherk PA, Granton JT, Kapral MK. Red blood cell transfusion in the intensive care unit. Intensive Care Med 2000;26:344-6. 28. Hebert PC, Wells G, Tweeddale M, et al. Does transfusion practice affect mortality in critically ill patients? Am J Respir Crit Care Med 1997;155:1618-23. 29. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368-77. 30. American College of Physicians. Practice strategies for elective red blood cell transfusion. Ann Intern Med 1992;116:403-6. 31. American Society of Anesthesiologists Task Force on Blood Component Therapy. Practice guidelines for blood component therapy. Anesthesiology 1996;84:732-47. 32. Adam RC, Lundy JS. Anesthesia in cases of poor risk: some suggestions for decreasing the risk. Surg Gynecol Obstet 1942;74:1011-9. 33. Stehling LC, Esposito B. An analysis of the appropriateness of intraoperative transfusion. Anesth Analg 1989;68(Suppl): S278-9. 34. Stehling LC, Ellison N, Faust RJ, Grotta AW, Moyers JR. A survey of transfusion practices among anesthesiologists. Vox Sang 1987;52:60-2. 35. Hebert PC, Wells G, Marshall J, et al. Transfusion requirements in critical care: a pilot study. JAMA 1995;273:1439-44. 36. Vincent JL, Sakr Y, Sprung C, et al. Are blood transfusions associated with greater mortality rates? Results of the Sepsis Occurrence in Acutely Ill Patients Study. Anesthesiology 2008; 108:31-9.
www.heartandlung.org
225