- Email: [email protected]
Impact of minimizing diagnostic blood loss in the critically ill
Impact of minimizing diagnostic blood loss in the critically ill Heather S. Dolman, MD,a,b Kelly Evans, C-NP,c Lisa Hall Zimmerman, PharmD,d Todd Lavery, MD,a,b Alfred E. Baylor, MD,a,b Robert F. Wilson, MD,a,b and James G. Tyburski, MD,a,b Detroit, MI
Background. The use of a small-volume phlebotomy tube (SVPT) versus conventional-volume phlebotomy tube (CVPT) has led to a decrease in daily blood loss. Blood loss due to phlebotomy can lead ultimately to decreased rates of anemia and blood transfusions, which can be important in the critically ill patient. Methods. We compared SVPT vs CVPT retrospectively in critically ill adult patients age $18 years admitted to a surgical intensive care unit for $48 hours. CVPT were evaluated from January 2011 to May 2011 and SVPT from June 2012 to October 2012. Results. Amount of blood drawn for laboratory tests and transfusions were evaluated in 248 patients (116 SVPT vs 132 CVPT). When compared with CVPT, total blood volume removed (mean ± SD) with SVPT was less overall, 174 ± 182 mL vs 299 ± 355 mL, P = .001. Daily blood draws also were less, 22.5 ± 17.3 mL vs 31.7 ± 15.5 mL, P < .001. The units of packed red blood cells given were not significant, 4.4 ± 3.6 units vs 6.0 ± 8.2 units, P = .16. Conclusion. The use of SVPT blood sampling led to a decreased amount of blood drawn. Strategies that use SVPT in a larger cohort also may decrease the number of transfusions in selected patients. Every effort should be made to use SVPT. (Surgery 2015;158:1083-8.) From the Department of Surgery,a Detroit Receiving Hospital; Wayne State Universityb; and Departments of Nursingc and Pharmacy Services,d Detroit Receiving Hospital, Detroit, MI
ANEMIA IS A COMMON PROBLEM IN CRITICALLY ILL PA1 TIENTS. Studies have examined the relationships between blood draws and outcomes in this population. Some studies report phlebotomized volumes in excess of 600 mL for long-term, ventilated, critically ill patients during their hospital stay.1 Shaffer2 found a significant correlation between severity of illness, number of blood draws, and total amount of blood drawn. Anemia secondary to phlebotomy accounted for 40% of packed red blood cells (pRBC) transfusion requirements.3 In addition, anemia and resultant blood transfusions may lead to greater rates of sepsis, transfusion-related reactions, greater severity of illness, and prolonged durations of hospital stay.3 The use of small-volume phlebotomy tubes (SVPTs) is a blood conservation strategy to minimize blood volumes lost with laboratory testing.4 Accepted for publication May 22, 2015. Reprint requests: Heather S. Dolman, MD, Assistant Professor of Surgery, Wayne State School of Medicine/Detroit Medical Center, Michael and Marian Ilitch Department of Surgery, Detroit Receiving Hospital, 4201 St. Antoine Blvd, Suite 4S-13, Detroit, MI 48201. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Published by Elsevier Inc. http://dx.doi.org/10.1016/j.surg.2015.05.018
The use of SVPT compared with conventionalvolume phlebotomy tubes (CVPTs) has resulted in a 46% decrease in blood loss.4 The purpose of this study was to evaluate SVPTs versus CVPTs in critically ill adult patients to determine whether a significant decrease in blood loss from phlebotomy would occur and if the use of SVPTs decreased the need for blood transfusion. MATERIALS AND METHODS After approval from the Wayne State University Human Investigation Committee, this retrospective cohort evaluated patients admitted for $48 hours to the surgical intensive care unit (ICU) with an open admission policy and age $18 years. Patients in whom CVPTs were used were evaluated from January 2011 to May 2011 and compared with patients whom SVPTs were used from June 2012 to October 2012. After education by a dedicated clinical nurse practitioner, all patients admitted to the surgical intensive care were switched to phlebotomy with SVPT on June 1, 2012. On the basis of the practice model our institution, 95% of our patients are admitted through the emergency department. In patients presenting with active bleeding from trauma, gastrointestinal bleeding, or bleeding from any other cause, homeostasis was obtained before evaluation in this SURGERY 1083
1084 Dolman et al
study, and, thus, units of blood transfused were not evaluated in this population. Baseline demographics, Acute Physiology and Chronic Health Evaluation (APACHE) II, admitting service, and laboratory data, including hemoglobin (Hgb), were evaluated. Analysis of phlebotomy included the number of blood studies performed, the blood volume removed per ICU day, and the total blood volume for the entire ICU stay. pRBC transfusions, ICU and hospital durations of stay, and in-hospital mortality were compared between groups. BD Vacutainer phlebotomy tubes (Becton, Dickinson, and Company, Franklin Lakes, NJ) were used for both groups during the study periods. In the CVPT group, 8.5 mL was used for a basic metabolic panel (serum electrolytes and creatine and blood urea nitrogen) and therapeutic drug levels. For complete blood count (CBC) and crossmatch, 6.0 mL was used in a K2 EDTA tube (Becton, Dickinson, and Company). A buffered sodium citrate 0.109 M, 3.2% 2.7 mL was used for prothrombin time, international normalized ratio, and partial thromboplastin times. In the SVPT group, 5.0 mL was used for basic metabolic panel and therapeutic drug levels. The K2 EDTA 2.0 mL tube was used for CBC and crossmatch. A buffered sodium citrate 0.109 M, 3.2% 1.8 mL was used for prothrombin time, international normalized ratio, and partial thromboplastin times. For both groups, arterial blood gas analysis used 3 mL. Blood cultures were drawn using BD Bactec bottles, one aerobic and one anaerobic, using 10 mL each. For CVPT and SVPT, the core laboratory at the Detroit Medical Center used the Dimension Vista for electrolytes and SYSMEX for CBC and other analyzes. When laboratory studies were ordered, phlebotomy was obtained routinely from a triple lumen, central venous catheter or peripherally inserted central catheter (ie, PICC) line. Arterial blood gas analysis was performed via an arterial catheter. Per our critical care practice policy, a venous arterial blood management protection system was used to minimize blood loss with phlebotomy. Waste occurred during the initial phase of phlebotomy when the blood sample was first removed. Waste was defined as two times the volume of the catheter. The central venous catheter used during the study period was the ARROWgard Blue PLUS (Arrow International, Reading, PA) with the maximum priming volume of 0.42 mL with the distal lumen. Hence, the maximum waste would be ;1 mL per phlebotomy episode.
Surgery October 2015
No restriction was placed on the number of phlebotomies performed. Additionally, no change in transfusion thresholds was made during the study periods. On the basis of the Hgb, anemia was defined based on the following severity: mild Hgb 9.0–11.0 g/dL, moderate for a Hgb 7.0–8.9 g/dL, and severe <7.0 g/dL. The decision to transfuse was at the discretion of the primary team with a restrictive transfusion policy of a Hgb <7.0 g/dL unless hemodynamic instability or active bleeding was present.5 Statistical analysis was performed using SPSS v 21 (IBM, Armonk, NY). Univariate analyses evaluated baseline differences between groups. Categorical variables were compared using Pearson’s v2 analysis. Continuous variables were analyzed with the Student t test or Mann–Whitney U as appropriate. Continuous variables are presented as the mean ± SD. Power analysis revealed 62 patients were needed in each group to see a decrease in pRBC transfusions by one unit. RESULTS The study evaluated 248 patients: 116 SVPT versus 132 CVPT. The mean age was 56 ± 19 years, and 62% of the patients were male. When we compared patients with SVPT vs CVPT, the APACHE II was not different between groups, 14.1 ± 8.6 vs 12.7 ± 6.9; P = .17 (Table I). Baseline Hgb levels, 11.7 ± 2.6 g/dL vs 11.6 ± 2.5 g/dL; P = .70, and the number of phlebotomy studies per ICU days were not different between groups, 5.5 ± 2.8 studies/ICU day SVPT vs 5.6 ± 2.4 studies/ICU day CVPT; P = .74. The total blood volume removed in these critically ill patients with SVPT was less, 174 ± 182 mL vs 299 ± 355 mL; P = .001. In addition, overall ICU daily volume was less with SVPT, 22.5 ± 17.3 mL vs 31.7 ± 15.5 mL; P < .001 (Table II). Analyzing transfusion practices, patients were of similar severity of illness (APACHE II), 14.1 ± 7.7 with transfusion versus 12.6 ± 7.8 without transfusion (P = .13). The units of pRBCs transfused were not different with SVPT, 4.4 ± 3.6 units vs 6.0 ± 8.2 units; P = .16 with no difference in rates of administration been groups. At least one episode of severe anemia (Hgb <7.0 g/dL) was seen less frequently in the SVPT group, 12/116 (10%) vs 29/132 (22%); P = .01. The surviving patients who did not receive a transfusion had a lesser duration of ICU stay (6 ± 5 days vs 12 ± 15 days for patients receiving transfusion; P < .001) and a lesser hospital duration of stay (11 ± 13 days vs 19 ± 16 days for patients receiving transfusion, P < .001).
Dolman et al 1085
Surgery Volume 158, Number 4
Table I. Patient demographics Characteristics
SVPT (N = 116)
Table II. Overall blood loss with each phlebotomy CVPT (N = 132)
Age, y, mean ± SD 55 ± 20 57 ± 19 Sex, male, n (%) 73 (63) 81 (61) Admitting service, n (%) Emergency Surgery/ 73 (63) 85 (64) Trauma Medicine 25 (22) 23 (18) Other 18 (15) 24 (18) APACHE II score 14.1 ± 8.6 12.7 ± 6.9
P value
Variable
.63 .80
Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL pRBC, units transfused
.81 .41 .61 .17
SVPT (N = 116)
CVPT (N = 132)
P value
174 ± 182
299 ± 355
<.001
22.5 ± 17.3
31.7 ± 15.5
<.001
4.4 ± 3.6
6.0 ± 8.2
.16
CVPT, Conventional-volume phlebotomy tube; pRBC, packed red blood cells; SVPT, small-volume phlebotomy tube.
APACHE II, Acute Physiology and Chronic Health Evaluation; CVPT, conventional-volume phlebotomy tube; SVPT, small-volume phlebotomy tube.
Regarding severely ill patients with an APACHE II score greater than 20, Hgb decreased by 2.2 ± 1.8 g/dL from ICU admission to ICU day 3, and our mean daily blood volume for studies was 110 ± 78 mL. Evaluating patients alive at day 3, no difference was seen in transfusion practices in SVPT and CVPT in severely ill patients with an APACHE score greater than 20. In surviving patients, the ICU and hospital durations of stay were not different in the 2 groups (SVPT vs CVPT). Of the 39 patients who died, the median day of death was 10 days, range 2–50 days. In a subgroup of patients admitted for emergency operation/trauma, the total blood volume removed with phlebotomy was less with SVPT, 156 ± 167 mL vs 288 ± 365 mL; P = .005 (Table III). In addition, the daily volume was lower with SVPT, 20.9 ± 19.7 mL vs 29.5 ± 11.8 mL; P = .001. The amount of pRBC administered also tended to be lower, 4.1 ± 3.0 units vs 6.6 ± 9.1 units; P = .06. Durations of ICU or hospital stay were not different between groups in this population. For a subgroup of medical patients, no differences were seen for total blood volume, the amount of pRBC transfused and durations of stay between groups. However, a difference was seen in total daily blood volumes using SVPT (P = .003). The effects of severe anemia (Hgb < 7.0 g/dL) impacted in-hospital mortality and durations of stay. In-hospital mortality was less likely in patients with severe anemia with SVPT phlebotomy practices (12% vs 45%; P = .03). Hospital stay was greater in patients with at least one episode of severe anemia (23.3 ± 25.5 days vs 13.6 ± 11.8 days without severe anemia, P < .001). Stay in the ICU was greater in patients with a Hgb <7.0 g/dL (16.0 ± 19.1 days vs 7.7 ± 9.4 days; P < .001). In patients with moderate anemia (Hgb 7.0–9.0 g/dL) hospital duration of stay was greater (19.3 ± 18.9 days vs 11.0 ± 8.3 days; P < .001) and ICU duration of stay greater
(11.9 ± 15.1 days vs 6.2 ± 6.0 days; P < .001) were greater. DISCUSSION Decreased iatrogenic blood loss is one method to minimize anemia and the subsequent need for blood transfusions in the critically ill patient. The use of SVPT is a blood-conservation strategy that has been shown to decrease the risk of anemia secondary to phlebotomy.5 Although the use of SVPT can decrease blood loss, many centers still use the greater-volume CVPTs in their critically ill patients. Our study and prior authors have demonstrated a decrease in iatrogenic blood loss and episodes of anemia due to phlebotomy with the use of SVPT.3,4 The emergency general surgery/trauma patients in our study appeared to have fewer blood transfusions, but the difference was not significant (P = .16). A recent review of blood-conservation strategies showed volume of blood loss and transfusion requirements were decreased even in the setting of a restrictive transfusion strategy.6 We propose that the rates of transfusions or the number of units transfused was not different between groups, because transfusions occurred at times when Hgb was greater than 7.0 g/dL or if the patient was hemodynamically unstable. Although not the primary objective, we did not find differences in durations of stay in the ICU or hospital or in-hospital mortality between groups. The triggers for transfusion generally followed established guidelines for our patients.5 In addition, care was not altered during the study periods; only the tube size was changed. Our definition of anemia was stricter than the World Health Organization who defines severe anemia as a Hgb less than 8.0 g/dL.7 Using the WHO definition, we would have had more patients identified with severe anemia. It has been reported that for every 125 mL of blood drawn, there is a decrease in Hgb by
1086 Dolman et al
Surgery October 2015
Table III. Blood loss with phlebotomy based on admitting service Emergency surgery/trauma Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL Duration of ICU stay, d* Hgb <7 g/dL, n (%) pRBC, units transfused
SVPT (N = 73) 156 20.9 9.2 6 4.1
± 167 ± 19.7 ± 10.1 (8) ± 3.0
CVPT (N = 85) 288 29.5 10.6 20 6.6
± 365 ± 11.8 ± 13.8 (24) ± 9.1
P value .005 .001 .46 .01 .06
Medicine
SVPT (N = 23)
CVPT (N = 25)
P value
Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL Duration of ICU stay, d* Hgb <7 g/dL, n (%) pRBC, units transfused
218 ± 205 25.3 ± 12.8 9.7 ± 8.8 3 (13) 4.8 ± 3.5
235 ± 137 42.6 ± 24.6 6.6 ± 4.0 4 (16) 3.1 ± 1.4
.72 .003 .12 .99 .15
*Duration of stay in surviving patients. CVPT, Conventional-volume phlebotomy tube; Hgb, hemoglobin; ICU, intensive care unit; pRBC, packed red blood cells; SVPT, small-volume phlebotomy tube.
1.0 g/dL.8 Our severely ill patients had at least a 2.0 g/dL decrease in Hgb from ICU admission to ICU day 3 with only 110 mL removed for studies. This difference may be secondary to resuscitation in the early phases of critical illness or unrecognized waste from phlebotomy draws not returned to the patient. We looked at patients who were alive at day 3 or longer and did not find a difference is transfusion practices between groups. Although we have a restrictive transfusion policy, we learned with this study that transfusions occurred even when the patient’s Hgb was greater than 7 g/dL perhaps secondary to hemodynamic instability. We also found no differences inhospital mortality in patients. Mathematical modeling reveals weeks of repetitive phlebotomy is additive to the bone marrow suppression for production of blood cells in critically ill patients.9 If the amount of blood phlebotomized can be decreased, a resultant decrease in anemia and a subsequent decrease in need for transfusion should occur. The SVPT patients who were transfused also needed 2 units of blood transfused less than the CVPT patients, especially in the surgery/trauma cohort. Although the majority of our ICU admissions are emergency general surgery and trauma patients, our surgical ICU has an open admission policy. We evaluated these patients separately, because they represent different populations regarding baseline co-morbid states and clinical management. Comparing our surgical and medical patients, the durations of stay were similar which was not expected. We anticipated the medical critically ill patients to have a greater duration of stay based on their co-morbid states.
Regarding the financial impact of converting from CVPT to SVPT, no cost difference occurred for the tubes or processing at our institution. Although we did not perform an economic analysis for this comparison, we project a cost savings if less blood is transfused. One recent study suggested that the cost to transfuse one unit of pRBC ranges from $522 to $1,183 per-unit.10 Cost variables include the initial blood collection, donor recruitment, screening, blood processing, donor notification, transport from the collection facility to the transfusion center, and costs related to inventory. The costs involved in the transfusion of one unit include testing of blood compatibility, administration, and monitoring. Limitations exists requiring further discussion. Confounders such as the use of medications, eg, anticoagulants, nonsteroidal anti-inflammatory drugs, or blood loss from operative procedures were not evaluated. In addition, operative blood loss was not reviewed, because this study population was analyzed after operative or endoscopic homeostasis was obtained. Also, blood drawn with each phlebotomy but not returned to the patient was considered to be waste and not analyzed. The waste could have been as great as 20 mL based on survey despite critical care practice policy. In conclusion, in our open ICU model with a restrictive transfusion practice, SVPT results in less blood drawn from patients both overall and on a daily basis without significant differences in transfusion rates in a mixed medical and surgical cohort matched for APACHE II scores. A causal relationship exists between anemia and increased mortality and morbidity. We found the use of SVPT decreased episodes of anemia. As such, anemia is
Surgery Volume 158, Number 4
an important problem in the critically ill patient that correlates with adverse events. The presence of anemia can lead to blood transfusions. These transfusions impart negative sequela such as increase risk of infection, circulatory overload, transfusion reactions, and mortality. Considering these variables, multiple methods of blood conservation should be used. Every effort should be made to use SVPTs for laboratory analysis. REFERENCES 1. Dale JC, Pruett SK. Phlebotomy---a minimalist approach. Mayo Clin Proc 1993;68:249-55. 2. Shaffer C. Diagnostic blood loss in mechanically ventilated patients. Heart Lung 2007;36:217-22. 3. Chant C, Wilson G, Friedrich JO. Anemia, transfusion, and phlebotomy practices in critically ill patients with prolonged ICU length of stay: a cohort study. Crit Care 2006; 10:R140. 4. Smoller BR, Kruskall MS, Horowitz GL. Reducing adult phlebotomy blood loss with the use of pediatric-sized blood collection tubes. Am J Clin Pathol 1989;91:701-3. 5. Napolitano LM, Kurek S, Luchette FA, Corwin HL, Barie PS, Tisherman SA, et al. Clinical practice guideline: red blood cell transfusion in adult trauma and critical care. Crit Care Med 2009;37:3124-57. 6. Page C, Retter A, Wyncoll D. Blood conservation devices in critical care: a narrative review. Ann Intensive Care 2013;3:14. 7. World Health Organization. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System. Geneva: World Health Organization; 2011.(WHO/NMH/NHD/ MNM/11.1). Available from: http://www.who.int/vmnis/ indicators/haemoglobin.pdf. Accessed April 20, 2015. 8. Thavendiranathan P, Bagai A, Ebidia A, Detsky AS, Choudhry NK. Do blood tests cause anemia in hospitalized patients? the effect of diagnostic phlebotomy on hemoglobin and hematocrit levels. J Gen Intern Med 2005;20:520-4. 9. Lyon AW, Chin AC, Slotsve GA, Lyon ME. Simulation of repetitive diagnostic blood loss and onset of iatrogenic anemia in critical care patients with a mathematical model. Comput Biol Med 2013;43:84-90. 10. Shander A, Hofmann A, Ozawa S, Theusinger OM, Gombotz H, Spahn DR. Activity-based costs of blood transfusions in surgical patients at four hospitals. Transfusion 2010;50:753-65.
DISCUSSION Dr William Cirocco (Columbus, OH): The authors have addressed the problem of blood draws in critically ill patients which, during a substantial stay in the ICU, may contribute to anemia and associated morbidity, including the potential need for blood transfusion. Anemia secondary to phlebotomy may account for up to 40% of blood transfusions, with potential associated complications including transfusion reaction and sepsis. The proposed correction is simply to substitute small-volume pediatric phlebotomy tubes for
Dolman et al 1087
conventional large-volume phlebotomy tubes used for adult patients who are critically ill. The goal of decreasing blood loss from phlebotomy is commendable and would seem to be a no brainer. The excess average 10 mL saved on a daily basis will be substantial for a given patient during a prolonged ICU stay, adding up to less occurrences of severe anemia (Hgb <7.0 gm/dL) despite the fact that statistical significance was not achieved in this study. Previous studies which have established that for every 125 mL of blood drawn, there is a decrease in hemoglobin by 1.0 gm/dL projects that every 12 days, the ICU patient will have a corresponding decrease of Hgb by 1.0 g/dL because of phlebotomy alone. However, the authors report at least a 2.0 g/dL decrease by ICU day 3, with only 110 mL removed for studies, which may reflect blood wasted from phlebotomy draws not recorded and returned to the patient. This ‘‘waste’’ blood was not analyzed in this study but was estimated to be as high as 20 mL, depending on the type of intravenous catheter and is most certainly an added factor in ICU anemia. The authors have demonstrated that minimizing the blood volume drawn for daily labs in the ICU setting is not only possible but practical by simply switching small-volume phlebotomy tubes for the conventional larger volume phlebotomy tubes. As a medical student, I was impressed that the blood from a needle prick in the heel of a neonate could produce daily labs at the Children’s Hospital of Michigan (CHM) but at the adult hospitals on the same campus (Harper, Hutzel, and Receiving Hospitals) a 5- to 10-mL blood draw was required. The machines spitting out the laboratory data do not discriminate whether the analyzed blood has come from a patient who is 60 years old or 60 minutes old. The technology exists to make these simple changes. Why hasn’t this already been implemented at the DMC and other institutions? Are there barriers, such as contractual issues, for instance, standing in the way of replacing technology at the adult hospitals with the technology in use at CHM? Second, you mention ICU protocol regarding blood transfusion. Do you envision other potential protocols that also may have an impact on iatrogenic blood loss, such as a policy on blood waste or limiting daily labs (especially for patients with a prolonged ICU stay)? Finally, are there further initiatives to counter ICU anemia that we should be considering in this
1088 Dolman et al
critically ill patient population? Is there a role for commercially available products such as Epogen or intravenous iron and other products which are currently part of protocols to correct severe anemia in the Jehovah’s Witness population of patients, for instance? Dr Heather Dolman: To answer the first question why it hasn’t been implemented, as you allude, there are many barriers to change. This study was discussed for a year and a half before the lab group would change over, after being certain that there would be no difference in cost, which is what everyone is concerned with. You are correct. Why aren’t we using pediatric blood draws? The lab machine, actually, only needs 0.1 mL to run the labs. Why on earth are we drawing so much blood from these patients? As
Surgery October 2015
you know, it’s culture change. How do you implement that? Obviously, ICU protocols are very important. Other protocols to implement change would be things like minimizing your daily lab draws, particularly in those people who don’t need labs. People like data to treat patients. It makes you feel better, like you’re creating change. We need to learn to be different, to stop and think: do we really need the labs to be drawn? Further things to consider, as you mention, are Epogen and IV iron. We actually use pediatric blood draws in Jehovah’s Witness patients. We didn’t include them in this study. Dr William Cirocco: One last comment. This should be reproduced across the country and across North America. It’s awesome. Thank you.
Background. The use of a small-volume phlebotomy tube (SVPT) versus conventional-volume phlebotomy tube (CVPT) has led to a decrease in daily blood loss. Blood loss due to phlebotomy can lead ultimately to decreased rates of anemia and blood transfusions, which can be important in the critically ill patient. Methods. We compared SVPT vs CVPT retrospectively in critically ill adult patients age $18 years admitted to a surgical intensive care unit for $48 hours. CVPT were evaluated from January 2011 to May 2011 and SVPT from June 2012 to October 2012. Results. Amount of blood drawn for laboratory tests and transfusions were evaluated in 248 patients (116 SVPT vs 132 CVPT). When compared with CVPT, total blood volume removed (mean ± SD) with SVPT was less overall, 174 ± 182 mL vs 299 ± 355 mL, P = .001. Daily blood draws also were less, 22.5 ± 17.3 mL vs 31.7 ± 15.5 mL, P < .001. The units of packed red blood cells given were not significant, 4.4 ± 3.6 units vs 6.0 ± 8.2 units, P = .16. Conclusion. The use of SVPT blood sampling led to a decreased amount of blood drawn. Strategies that use SVPT in a larger cohort also may decrease the number of transfusions in selected patients. Every effort should be made to use SVPT. (Surgery 2015;158:1083-8.) From the Department of Surgery,a Detroit Receiving Hospital; Wayne State Universityb; and Departments of Nursingc and Pharmacy Services,d Detroit Receiving Hospital, Detroit, MI
ANEMIA IS A COMMON PROBLEM IN CRITICALLY ILL PA1 TIENTS. Studies have examined the relationships between blood draws and outcomes in this population. Some studies report phlebotomized volumes in excess of 600 mL for long-term, ventilated, critically ill patients during their hospital stay.1 Shaffer2 found a significant correlation between severity of illness, number of blood draws, and total amount of blood drawn. Anemia secondary to phlebotomy accounted for 40% of packed red blood cells (pRBC) transfusion requirements.3 In addition, anemia and resultant blood transfusions may lead to greater rates of sepsis, transfusion-related reactions, greater severity of illness, and prolonged durations of hospital stay.3 The use of small-volume phlebotomy tubes (SVPTs) is a blood conservation strategy to minimize blood volumes lost with laboratory testing.4 Accepted for publication May 22, 2015. Reprint requests: Heather S. Dolman, MD, Assistant Professor of Surgery, Wayne State School of Medicine/Detroit Medical Center, Michael and Marian Ilitch Department of Surgery, Detroit Receiving Hospital, 4201 St. Antoine Blvd, Suite 4S-13, Detroit, MI 48201. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Published by Elsevier Inc. http://dx.doi.org/10.1016/j.surg.2015.05.018
The use of SVPT compared with conventionalvolume phlebotomy tubes (CVPTs) has resulted in a 46% decrease in blood loss.4 The purpose of this study was to evaluate SVPTs versus CVPTs in critically ill adult patients to determine whether a significant decrease in blood loss from phlebotomy would occur and if the use of SVPTs decreased the need for blood transfusion. MATERIALS AND METHODS After approval from the Wayne State University Human Investigation Committee, this retrospective cohort evaluated patients admitted for $48 hours to the surgical intensive care unit (ICU) with an open admission policy and age $18 years. Patients in whom CVPTs were used were evaluated from January 2011 to May 2011 and compared with patients whom SVPTs were used from June 2012 to October 2012. After education by a dedicated clinical nurse practitioner, all patients admitted to the surgical intensive care were switched to phlebotomy with SVPT on June 1, 2012. On the basis of the practice model our institution, 95% of our patients are admitted through the emergency department. In patients presenting with active bleeding from trauma, gastrointestinal bleeding, or bleeding from any other cause, homeostasis was obtained before evaluation in this SURGERY 1083
1084 Dolman et al
study, and, thus, units of blood transfused were not evaluated in this population. Baseline demographics, Acute Physiology and Chronic Health Evaluation (APACHE) II, admitting service, and laboratory data, including hemoglobin (Hgb), were evaluated. Analysis of phlebotomy included the number of blood studies performed, the blood volume removed per ICU day, and the total blood volume for the entire ICU stay. pRBC transfusions, ICU and hospital durations of stay, and in-hospital mortality were compared between groups. BD Vacutainer phlebotomy tubes (Becton, Dickinson, and Company, Franklin Lakes, NJ) were used for both groups during the study periods. In the CVPT group, 8.5 mL was used for a basic metabolic panel (serum electrolytes and creatine and blood urea nitrogen) and therapeutic drug levels. For complete blood count (CBC) and crossmatch, 6.0 mL was used in a K2 EDTA tube (Becton, Dickinson, and Company). A buffered sodium citrate 0.109 M, 3.2% 2.7 mL was used for prothrombin time, international normalized ratio, and partial thromboplastin times. In the SVPT group, 5.0 mL was used for basic metabolic panel and therapeutic drug levels. The K2 EDTA 2.0 mL tube was used for CBC and crossmatch. A buffered sodium citrate 0.109 M, 3.2% 1.8 mL was used for prothrombin time, international normalized ratio, and partial thromboplastin times. For both groups, arterial blood gas analysis used 3 mL. Blood cultures were drawn using BD Bactec bottles, one aerobic and one anaerobic, using 10 mL each. For CVPT and SVPT, the core laboratory at the Detroit Medical Center used the Dimension Vista for electrolytes and SYSMEX for CBC and other analyzes. When laboratory studies were ordered, phlebotomy was obtained routinely from a triple lumen, central venous catheter or peripherally inserted central catheter (ie, PICC) line. Arterial blood gas analysis was performed via an arterial catheter. Per our critical care practice policy, a venous arterial blood management protection system was used to minimize blood loss with phlebotomy. Waste occurred during the initial phase of phlebotomy when the blood sample was first removed. Waste was defined as two times the volume of the catheter. The central venous catheter used during the study period was the ARROWgard Blue PLUS (Arrow International, Reading, PA) with the maximum priming volume of 0.42 mL with the distal lumen. Hence, the maximum waste would be ;1 mL per phlebotomy episode.
Surgery October 2015
No restriction was placed on the number of phlebotomies performed. Additionally, no change in transfusion thresholds was made during the study periods. On the basis of the Hgb, anemia was defined based on the following severity: mild Hgb 9.0–11.0 g/dL, moderate for a Hgb 7.0–8.9 g/dL, and severe <7.0 g/dL. The decision to transfuse was at the discretion of the primary team with a restrictive transfusion policy of a Hgb <7.0 g/dL unless hemodynamic instability or active bleeding was present.5 Statistical analysis was performed using SPSS v 21 (IBM, Armonk, NY). Univariate analyses evaluated baseline differences between groups. Categorical variables were compared using Pearson’s v2 analysis. Continuous variables were analyzed with the Student t test or Mann–Whitney U as appropriate. Continuous variables are presented as the mean ± SD. Power analysis revealed 62 patients were needed in each group to see a decrease in pRBC transfusions by one unit. RESULTS The study evaluated 248 patients: 116 SVPT versus 132 CVPT. The mean age was 56 ± 19 years, and 62% of the patients were male. When we compared patients with SVPT vs CVPT, the APACHE II was not different between groups, 14.1 ± 8.6 vs 12.7 ± 6.9; P = .17 (Table I). Baseline Hgb levels, 11.7 ± 2.6 g/dL vs 11.6 ± 2.5 g/dL; P = .70, and the number of phlebotomy studies per ICU days were not different between groups, 5.5 ± 2.8 studies/ICU day SVPT vs 5.6 ± 2.4 studies/ICU day CVPT; P = .74. The total blood volume removed in these critically ill patients with SVPT was less, 174 ± 182 mL vs 299 ± 355 mL; P = .001. In addition, overall ICU daily volume was less with SVPT, 22.5 ± 17.3 mL vs 31.7 ± 15.5 mL; P < .001 (Table II). Analyzing transfusion practices, patients were of similar severity of illness (APACHE II), 14.1 ± 7.7 with transfusion versus 12.6 ± 7.8 without transfusion (P = .13). The units of pRBCs transfused were not different with SVPT, 4.4 ± 3.6 units vs 6.0 ± 8.2 units; P = .16 with no difference in rates of administration been groups. At least one episode of severe anemia (Hgb <7.0 g/dL) was seen less frequently in the SVPT group, 12/116 (10%) vs 29/132 (22%); P = .01. The surviving patients who did not receive a transfusion had a lesser duration of ICU stay (6 ± 5 days vs 12 ± 15 days for patients receiving transfusion; P < .001) and a lesser hospital duration of stay (11 ± 13 days vs 19 ± 16 days for patients receiving transfusion, P < .001).
Dolman et al 1085
Surgery Volume 158, Number 4
Table I. Patient demographics Characteristics
SVPT (N = 116)
Table II. Overall blood loss with each phlebotomy CVPT (N = 132)
Age, y, mean ± SD 55 ± 20 57 ± 19 Sex, male, n (%) 73 (63) 81 (61) Admitting service, n (%) Emergency Surgery/ 73 (63) 85 (64) Trauma Medicine 25 (22) 23 (18) Other 18 (15) 24 (18) APACHE II score 14.1 ± 8.6 12.7 ± 6.9
P value
Variable
.63 .80
Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL pRBC, units transfused
.81 .41 .61 .17
SVPT (N = 116)
CVPT (N = 132)
P value
174 ± 182
299 ± 355
<.001
22.5 ± 17.3
31.7 ± 15.5
<.001
4.4 ± 3.6
6.0 ± 8.2
.16
CVPT, Conventional-volume phlebotomy tube; pRBC, packed red blood cells; SVPT, small-volume phlebotomy tube.
APACHE II, Acute Physiology and Chronic Health Evaluation; CVPT, conventional-volume phlebotomy tube; SVPT, small-volume phlebotomy tube.
Regarding severely ill patients with an APACHE II score greater than 20, Hgb decreased by 2.2 ± 1.8 g/dL from ICU admission to ICU day 3, and our mean daily blood volume for studies was 110 ± 78 mL. Evaluating patients alive at day 3, no difference was seen in transfusion practices in SVPT and CVPT in severely ill patients with an APACHE score greater than 20. In surviving patients, the ICU and hospital durations of stay were not different in the 2 groups (SVPT vs CVPT). Of the 39 patients who died, the median day of death was 10 days, range 2–50 days. In a subgroup of patients admitted for emergency operation/trauma, the total blood volume removed with phlebotomy was less with SVPT, 156 ± 167 mL vs 288 ± 365 mL; P = .005 (Table III). In addition, the daily volume was lower with SVPT, 20.9 ± 19.7 mL vs 29.5 ± 11.8 mL; P = .001. The amount of pRBC administered also tended to be lower, 4.1 ± 3.0 units vs 6.6 ± 9.1 units; P = .06. Durations of ICU or hospital stay were not different between groups in this population. For a subgroup of medical patients, no differences were seen for total blood volume, the amount of pRBC transfused and durations of stay between groups. However, a difference was seen in total daily blood volumes using SVPT (P = .003). The effects of severe anemia (Hgb < 7.0 g/dL) impacted in-hospital mortality and durations of stay. In-hospital mortality was less likely in patients with severe anemia with SVPT phlebotomy practices (12% vs 45%; P = .03). Hospital stay was greater in patients with at least one episode of severe anemia (23.3 ± 25.5 days vs 13.6 ± 11.8 days without severe anemia, P < .001). Stay in the ICU was greater in patients with a Hgb <7.0 g/dL (16.0 ± 19.1 days vs 7.7 ± 9.4 days; P < .001). In patients with moderate anemia (Hgb 7.0–9.0 g/dL) hospital duration of stay was greater (19.3 ± 18.9 days vs 11.0 ± 8.3 days; P < .001) and ICU duration of stay greater
(11.9 ± 15.1 days vs 6.2 ± 6.0 days; P < .001) were greater. DISCUSSION Decreased iatrogenic blood loss is one method to minimize anemia and the subsequent need for blood transfusions in the critically ill patient. The use of SVPT is a blood-conservation strategy that has been shown to decrease the risk of anemia secondary to phlebotomy.5 Although the use of SVPT can decrease blood loss, many centers still use the greater-volume CVPTs in their critically ill patients. Our study and prior authors have demonstrated a decrease in iatrogenic blood loss and episodes of anemia due to phlebotomy with the use of SVPT.3,4 The emergency general surgery/trauma patients in our study appeared to have fewer blood transfusions, but the difference was not significant (P = .16). A recent review of blood-conservation strategies showed volume of blood loss and transfusion requirements were decreased even in the setting of a restrictive transfusion strategy.6 We propose that the rates of transfusions or the number of units transfused was not different between groups, because transfusions occurred at times when Hgb was greater than 7.0 g/dL or if the patient was hemodynamically unstable. Although not the primary objective, we did not find differences in durations of stay in the ICU or hospital or in-hospital mortality between groups. The triggers for transfusion generally followed established guidelines for our patients.5 In addition, care was not altered during the study periods; only the tube size was changed. Our definition of anemia was stricter than the World Health Organization who defines severe anemia as a Hgb less than 8.0 g/dL.7 Using the WHO definition, we would have had more patients identified with severe anemia. It has been reported that for every 125 mL of blood drawn, there is a decrease in Hgb by
1086 Dolman et al
Surgery October 2015
Table III. Blood loss with phlebotomy based on admitting service Emergency surgery/trauma Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL Duration of ICU stay, d* Hgb <7 g/dL, n (%) pRBC, units transfused
SVPT (N = 73) 156 20.9 9.2 6 4.1
± 167 ± 19.7 ± 10.1 (8) ± 3.0
CVPT (N = 85) 288 29.5 10.6 20 6.6
± 365 ± 11.8 ± 13.8 (24) ± 9.1
P value .005 .001 .46 .01 .06
Medicine
SVPT (N = 23)
CVPT (N = 25)
P value
Phlebotomy blood loss, mL Daily phlebotomy blood loss, mL Duration of ICU stay, d* Hgb <7 g/dL, n (%) pRBC, units transfused
218 ± 205 25.3 ± 12.8 9.7 ± 8.8 3 (13) 4.8 ± 3.5
235 ± 137 42.6 ± 24.6 6.6 ± 4.0 4 (16) 3.1 ± 1.4
.72 .003 .12 .99 .15
*Duration of stay in surviving patients. CVPT, Conventional-volume phlebotomy tube; Hgb, hemoglobin; ICU, intensive care unit; pRBC, packed red blood cells; SVPT, small-volume phlebotomy tube.
1.0 g/dL.8 Our severely ill patients had at least a 2.0 g/dL decrease in Hgb from ICU admission to ICU day 3 with only 110 mL removed for studies. This difference may be secondary to resuscitation in the early phases of critical illness or unrecognized waste from phlebotomy draws not returned to the patient. We looked at patients who were alive at day 3 or longer and did not find a difference is transfusion practices between groups. Although we have a restrictive transfusion policy, we learned with this study that transfusions occurred even when the patient’s Hgb was greater than 7 g/dL perhaps secondary to hemodynamic instability. We also found no differences inhospital mortality in patients. Mathematical modeling reveals weeks of repetitive phlebotomy is additive to the bone marrow suppression for production of blood cells in critically ill patients.9 If the amount of blood phlebotomized can be decreased, a resultant decrease in anemia and a subsequent decrease in need for transfusion should occur. The SVPT patients who were transfused also needed 2 units of blood transfused less than the CVPT patients, especially in the surgery/trauma cohort. Although the majority of our ICU admissions are emergency general surgery and trauma patients, our surgical ICU has an open admission policy. We evaluated these patients separately, because they represent different populations regarding baseline co-morbid states and clinical management. Comparing our surgical and medical patients, the durations of stay were similar which was not expected. We anticipated the medical critically ill patients to have a greater duration of stay based on their co-morbid states.
Regarding the financial impact of converting from CVPT to SVPT, no cost difference occurred for the tubes or processing at our institution. Although we did not perform an economic analysis for this comparison, we project a cost savings if less blood is transfused. One recent study suggested that the cost to transfuse one unit of pRBC ranges from $522 to $1,183 per-unit.10 Cost variables include the initial blood collection, donor recruitment, screening, blood processing, donor notification, transport from the collection facility to the transfusion center, and costs related to inventory. The costs involved in the transfusion of one unit include testing of blood compatibility, administration, and monitoring. Limitations exists requiring further discussion. Confounders such as the use of medications, eg, anticoagulants, nonsteroidal anti-inflammatory drugs, or blood loss from operative procedures were not evaluated. In addition, operative blood loss was not reviewed, because this study population was analyzed after operative or endoscopic homeostasis was obtained. Also, blood drawn with each phlebotomy but not returned to the patient was considered to be waste and not analyzed. The waste could have been as great as 20 mL based on survey despite critical care practice policy. In conclusion, in our open ICU model with a restrictive transfusion practice, SVPT results in less blood drawn from patients both overall and on a daily basis without significant differences in transfusion rates in a mixed medical and surgical cohort matched for APACHE II scores. A causal relationship exists between anemia and increased mortality and morbidity. We found the use of SVPT decreased episodes of anemia. As such, anemia is
Surgery Volume 158, Number 4
an important problem in the critically ill patient that correlates with adverse events. The presence of anemia can lead to blood transfusions. These transfusions impart negative sequela such as increase risk of infection, circulatory overload, transfusion reactions, and mortality. Considering these variables, multiple methods of blood conservation should be used. Every effort should be made to use SVPTs for laboratory analysis. REFERENCES 1. Dale JC, Pruett SK. Phlebotomy---a minimalist approach. Mayo Clin Proc 1993;68:249-55. 2. Shaffer C. Diagnostic blood loss in mechanically ventilated patients. Heart Lung 2007;36:217-22. 3. Chant C, Wilson G, Friedrich JO. Anemia, transfusion, and phlebotomy practices in critically ill patients with prolonged ICU length of stay: a cohort study. Crit Care 2006; 10:R140. 4. Smoller BR, Kruskall MS, Horowitz GL. Reducing adult phlebotomy blood loss with the use of pediatric-sized blood collection tubes. Am J Clin Pathol 1989;91:701-3. 5. Napolitano LM, Kurek S, Luchette FA, Corwin HL, Barie PS, Tisherman SA, et al. Clinical practice guideline: red blood cell transfusion in adult trauma and critical care. Crit Care Med 2009;37:3124-57. 6. Page C, Retter A, Wyncoll D. Blood conservation devices in critical care: a narrative review. Ann Intensive Care 2013;3:14. 7. World Health Organization. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System. Geneva: World Health Organization; 2011.(WHO/NMH/NHD/ MNM/11.1). Available from: http://www.who.int/vmnis/ indicators/haemoglobin.pdf. Accessed April 20, 2015. 8. Thavendiranathan P, Bagai A, Ebidia A, Detsky AS, Choudhry NK. Do blood tests cause anemia in hospitalized patients? the effect of diagnostic phlebotomy on hemoglobin and hematocrit levels. J Gen Intern Med 2005;20:520-4. 9. Lyon AW, Chin AC, Slotsve GA, Lyon ME. Simulation of repetitive diagnostic blood loss and onset of iatrogenic anemia in critical care patients with a mathematical model. Comput Biol Med 2013;43:84-90. 10. Shander A, Hofmann A, Ozawa S, Theusinger OM, Gombotz H, Spahn DR. Activity-based costs of blood transfusions in surgical patients at four hospitals. Transfusion 2010;50:753-65.
DISCUSSION Dr William Cirocco (Columbus, OH): The authors have addressed the problem of blood draws in critically ill patients which, during a substantial stay in the ICU, may contribute to anemia and associated morbidity, including the potential need for blood transfusion. Anemia secondary to phlebotomy may account for up to 40% of blood transfusions, with potential associated complications including transfusion reaction and sepsis. The proposed correction is simply to substitute small-volume pediatric phlebotomy tubes for
Dolman et al 1087
conventional large-volume phlebotomy tubes used for adult patients who are critically ill. The goal of decreasing blood loss from phlebotomy is commendable and would seem to be a no brainer. The excess average 10 mL saved on a daily basis will be substantial for a given patient during a prolonged ICU stay, adding up to less occurrences of severe anemia (Hgb <7.0 gm/dL) despite the fact that statistical significance was not achieved in this study. Previous studies which have established that for every 125 mL of blood drawn, there is a decrease in hemoglobin by 1.0 gm/dL projects that every 12 days, the ICU patient will have a corresponding decrease of Hgb by 1.0 g/dL because of phlebotomy alone. However, the authors report at least a 2.0 g/dL decrease by ICU day 3, with only 110 mL removed for studies, which may reflect blood wasted from phlebotomy draws not recorded and returned to the patient. This ‘‘waste’’ blood was not analyzed in this study but was estimated to be as high as 20 mL, depending on the type of intravenous catheter and is most certainly an added factor in ICU anemia. The authors have demonstrated that minimizing the blood volume drawn for daily labs in the ICU setting is not only possible but practical by simply switching small-volume phlebotomy tubes for the conventional larger volume phlebotomy tubes. As a medical student, I was impressed that the blood from a needle prick in the heel of a neonate could produce daily labs at the Children’s Hospital of Michigan (CHM) but at the adult hospitals on the same campus (Harper, Hutzel, and Receiving Hospitals) a 5- to 10-mL blood draw was required. The machines spitting out the laboratory data do not discriminate whether the analyzed blood has come from a patient who is 60 years old or 60 minutes old. The technology exists to make these simple changes. Why hasn’t this already been implemented at the DMC and other institutions? Are there barriers, such as contractual issues, for instance, standing in the way of replacing technology at the adult hospitals with the technology in use at CHM? Second, you mention ICU protocol regarding blood transfusion. Do you envision other potential protocols that also may have an impact on iatrogenic blood loss, such as a policy on blood waste or limiting daily labs (especially for patients with a prolonged ICU stay)? Finally, are there further initiatives to counter ICU anemia that we should be considering in this
1088 Dolman et al
critically ill patient population? Is there a role for commercially available products such as Epogen or intravenous iron and other products which are currently part of protocols to correct severe anemia in the Jehovah’s Witness population of patients, for instance? Dr Heather Dolman: To answer the first question why it hasn’t been implemented, as you allude, there are many barriers to change. This study was discussed for a year and a half before the lab group would change over, after being certain that there would be no difference in cost, which is what everyone is concerned with. You are correct. Why aren’t we using pediatric blood draws? The lab machine, actually, only needs 0.1 mL to run the labs. Why on earth are we drawing so much blood from these patients? As
Surgery October 2015
you know, it’s culture change. How do you implement that? Obviously, ICU protocols are very important. Other protocols to implement change would be things like minimizing your daily lab draws, particularly in those people who don’t need labs. People like data to treat patients. It makes you feel better, like you’re creating change. We need to learn to be different, to stop and think: do we really need the labs to be drawn? Further things to consider, as you mention, are Epogen and IV iron. We actually use pediatric blood draws in Jehovah’s Witness patients. We didn’t include them in this study. Dr William Cirocco: One last comment. This should be reproduced across the country and across North America. It’s awesome. Thank you.