TRANSFUSION PRACTICE

Cryoprecipitate use in 25 Canadian hospitals: commonly used outside of the published guidelines
Edward C. Alport, Jeannie L. Callum, Susan Nahirniak, Bernie Eurich, and Heather A. Hume

BACKGROUND: Canadian Blood Services disposition reports suggested considerable variation in cryoprecipitate use and prompted this national audit. STUDY DESIGN AND METHODS: Thirty-one institutions were invited to participate in a 2-month audit. Patient information and relevant laboratory and transfusion data were collected. Cryoprecipitate transfusions were categorized as appropriate if a brinogen level (taken 6 hr before/after transfusion) was not more than 1.0 g per L and inappropriate if the pretransfusion brinogen level was more than 1.0 g per L and posttransfusion brinogen level was more than 1.0 g per L or not performed. Appropriateness was categorized as undetermined if the pretransfusion brinogen level was not performed and the posttransfusion brinogen level was more than 1.0 g per L or not performed. RESULTS: Overall, 25 of 31 invited hospitals agreed to participate. A total of 4370 units of cryoprecipitate were transfused in 603 events to 453 patients representing 62 percent of cryoprecipitate issued to hospitals during the time period. Comparison of the number of units of cryoprecipitate per 100 units of red blood cells (RBCs) transfused by each institution showed signicant variation in practice (mean, 9 per 100 RBCs; range, 2 to 27 units). The single most common indication for cryoprecipitate was cardiac surgery (45.4% of events). Overall, 24 percent of cryoprecipitate transfusions were considered to be appropriate (pretransfusion brinogen level 1 g/L in 19% and posttransfusion brinogen level 1.0 g/L in another 5%), 34 percent were inappropriate, and in 42 percent appropriateness could not be determined. CONCLUSION: A 2-month audit of cryoprecipitate use in Canada revealed that the majority of cryoprecipitate use in Canada is not in accordance with published guidelines.

ver the past few decades, the principle of evidence-based decision making has become widely accepted. Levels of evidence have become better dened, and when there is good evidence, care maps and practice guidelines are relatively easy to implement. However, for therapies that have been available for many years, randomized controlled clinical trials may never have been performed before the therapy coming into routine use. Cryoprecipitate transfusion is a good example of the latter situation. The combination of a biologically complex product, with current use bearing little resemblance to its originally intended use, changing safety prole due to advances in transmissible disease testing, evolution of alternative products, and use in complex conditions where it is difcult to perform randomized trials, has led to a wide variation in clinical practice. Previous studies suggest 24 to 61 percent of cryoprecipitate use is inappropriate,1,2 as dened by current practice guidelines. Monthly blood component disposition reports submitted to Canadian Blood Services (the blood supplier for all provinces and territories in Canada except for Quebec) by Canadian hospitals suggest that there is also considerable interhospital variation in cryoprecipitate use in Canada. For example, historical data from a representative 4-month period (April 1, 2005, to July 31, 2005) indicated that approximately 11,500 units of cryoprecipitate and 109,000 units of red blood cells (RBCs) were issued by Canadian Blood Services to hospitals giving an overall ratio

From the Canadian Blood Services, Regina, Saskatchewan; the Sunnybrook Health Sciences Centre and The University Health Network, Toronto, Ontario; Capital Health, Edmonton, Alberta; and the Canadian Blood Services, Head Ofce, Ottawa, Ontario, Canada. Address reprint requests to: Edward C. Alport, MD, FRCPC, Canadian Blood Services, 2571 Broad Street, PO Box 1185, Regina, SK, Canada S4P 3B4; e-mail: jeannie.callum@ sunnybrook.ca. Received for publication December 21, 2007; revision received April 26, 2008; and accepted April 26, 2008. doi: 10.1111/j.1537-2995.2008.01826.x TRANSFUSION 2008;48:2122-2127.

2122 TRANSFUSION

Volume 48, October 2008

CRYOPRECIPITATE UTILIZATION

of cryoprecipitate to RBC units issued of 11:100, but with a wide variation from 3:100 to 30:100 among individual institutions. This observation prompted a joint effort of Canadian Blood Services and selected Canadian hospitals to conduct a prospective audit of cryoprecipitate use.

MATERIALS AND METHODS

The hospitals receiving 80 percent of the issues of cryoprecipitate from Canadian Blood Services were identied.This group consisted of 31 major hospitals across Canada. The blood bank medical directors in these hospitals were invited to participate in a prospective 2-month audit of cryoprecipitate use in their institutions. Hospital blood banks agreeing to participate were asked to provide basic hospital demographic information and for each cryoprecipitate transfusion event (i.e., each cryoprecipitate pool administered) between March 6, 2006, and May 5, 2006, to complete a standardized cryoprecipitate audit form developed jointly by Canadian Blood Services and hospital participants. Hospital demographic information included the number of acute care beds, the number of RBC units transfused in each month covered by the audit (March and April 2006), and the number of major cardiac surgical procedures (dened as cardiac procedures with greater than 3 hr scheduled operating room time) performed per year. For the purposes of the calculations used in this report, the total number of cardiac surgeries for the 2-month study period was assumed to be one-sixth of the institutions annual total. The information requested for each cryoprecipitate transfusion included: 1) patient demographic and clinical information (age, sex, primary diagnosis, and clinical service); 2) laboratory information (hemoglobin [Hb] level, platelet count, prothrombin time/partial thromboplastin time, creatinine, and pre- and posttransfusion brinogen levels); 3) transfusion data (number of cryoprecipitate units transfused, number of cryoprecipitate transfusion events, and the total number of blood components used in the 12 hr before and after cryoprecipitate infusion). With the use of hospital month-end reports, the number of cryoprecipitate units transfused per 100 RBC units transfused was calculated for each hospital. For each patient receiving cryoprecipitate during the audit period, the rst cryoprecipitate transfusion event (i.e., cryoprecipitate pool) was evaluated for appropriateness using criteria developed by the investigators (ECA, JLC, HAH) in accordance with Canadian Blood Services publication Clinical Guide to Transfusion3 and other published guidelines.4-6 Cryoprecipitate transfusions were categorized as likely appropriate if a brinogen level, taken within 6 hours before or 6 hours after transfusion was not more than 1.0 g per L. A cryoprecipitate transfusion was categorized as likely inappropriate if the brinogen level within the 6 hours before transfusion was more than 1.0 g per L and posttransfusion brinogen level was either more

than 1.0 g per L or not performed. Appropriateness was categorized as undetermined if no pretransfusion brinogen level was performed and posttransfusion brinogen level was more than 1.0 g per L or not performed. The exact time of laboratory tests and infusion times were not collected. Differences in the appropriateness between hospitals with and without institutional guidelines were compared utilizing the chi-square test.

RESULTS
Twenty-ve of the 31 identied hospitals agreed to participate. Six hospitals in one province declined to participate due to their institutional requirements for ethics board approval and their inability to complete the necessary submission in the time required. The 25 institutions were designated by letters A to Y. Of the 25 participating institutions, 23 were tertiary care, university-afliated institutions. Hospital size ranged from 177 to 2089 acute care beds with a median of 452 acute care beds per hospital. Transfusion guidelines for cryoprecipitate were in place at 16 of the 25 hospitals. In addition, 3 additional sites required physician approval for release of the product. Intraoperative point of care testing for brinogen level was available in only 3 participating sites. Overall, 4370 units of cryoprecipitate were transfused in the 25 institutions. During the same time frame 7036 units of cryoprecipitate were issued to hospitals by Canadian Blood Services. Thus, this audit represents approximately 62 percent of the cryoprecipitate units transfused in hospitals supplied by Canadian Blood Services during the audit period. The 4370 units transfused represented 603 transfusion events in 453 different patients. Thirty-two percent (192/603) of cryoprecipitate transfusion events were given to pediatric patients (i.e., patients less than 16 years old). The clinical settings in which cryoprecipitate was administered are summarized in Table 1 and the laboratory data are summarized in Table 2. With the use of the assumption that there is less intrainstitutional variation in use of RBCs than cryoprecipitate, the number of cryoprecipitate units transfused per 100 RBC units transfused was calculated for each hospital as an approximate comparison of cryoprecipitate use among the hospitals. The mean number was 9 per 100 RBC units, with a range from 2 to 27 (in all hospitals except one, the ratio was 16 or less, with one outlier at 27). Using the established criteria for the rst (or only) cryoprecipitate transfusion event for each patient, 24 percent (108/453) of cryoprecipitate transfusion events were considered to be likely appropriate with pretransfusion brinogen levels of not more than 1.0 g per L in 19 percent (84/453) and posttransfusion brinogen levels of less than 1.0 g per L in another 5 percent (24/453); 34 percent (154/453) were considered to be likely inappropriate; and in the remaining 42 percent (191/453), appropriateness could not be
Volume 48, October 2008 TRANSFUSION 2123

ALPORT ET AL.

we found that only 24 percent of requests were clearly within the scope of Percentage Percentage of these recommendations,4-7 whereas 76 of events cryoprecipitate units percent of cryoprecipitate transfusions Clinical setting (n = 603) (n = 4370) were likely inappropriate or of undeterCardiac surgery 45.4 37.7 mined appropriateness. This may even Noncardiac surgery 16.9 19.2 Trauma 12.3 17.4 be a conservative estimate of appropriCancer 5.0 5.4 ateness as some patients in this report Neonatal hemorrhage 2.8 0.5 with brinogen levels at or below 1.0 g Hepatic failure 3.2 3.5 Obstetrical hemorrhage 2.2 2.7 per L may not have had active bleeding, Disseminated intravascular coagulopathy 4.0 4.1 a factor that we did not include in our Other 8.3 9.6 evaluation of appropriateness. In addiTotal 100 100 tion, a large variation in practice was observed, with appropriate use per hospital ranging from 0 to 100 percent of transfusions, despite the fact that the TABLE 2. Laboratory results of 453 patients receiving cryoprecipitate transfusions vast majority of participating hospitals Percentage (n = 453) were large, urban, academic institumissing tions. The single most common indicaVariable Median Range (incomplete) tion for administration was cardiac Precryoprecipitate Hb (g/L)* 88 32-202 19.6 surgery, accounting for 45 percent (274/ Precryoprecipitate brinogen (g/L) 1.3 0.14-4.3 46.8 Precryoprecipitate creatinine (mmol/L)* 87 11-599 39.3 603) of all infusions. Between 0.3 and Precryoprecipitate INR* 1.6 0.9-34.7 36.6 10.1 percent of patients having cardiac Precryoprecipitate PTT (sec)* 45 2-300 27.6 surgery at 13 institutions were adminisPostcryoprecipitate brinogen (g/L) 1.9 0.12-5.1 47.7 tered cryoprecipitate, of which only 6 * Hb, creatinine, INR, and PTT values may have been performed longer than 6 hr before transfusion. percent (6/104) had hypobrinogen Only brinogen levels (before and after) performed within 6 hr of the transfusion were emia (brinogen 1.0 g/L) documented used. before or after the administration of INR = international normalized ratio; PTT = partial thromboplastin time cryoprecipitate. Our results are comparable to a large audit of cryoprecipitate use in Australia reported in determined. Interhospital variation of appropriate use 2003.1 In that report, Schoeld and colleagues docuwas large, ranging from 0 to 100 percent. These results are summarized in Table 3. Hospitals with transfusion guidemented that only 38 percent of 64 cryoprecipitate transfulines or physician approval process had higher rates of sions were appropriate when the same criteria (brinogen appropriateness, when compared to institutions without <1.0 g/L) was applied to their cryoprecipitate transfuguidelines (26.1% vs. 8.6%; c2 = 8.486, p < 0.01). sions. A report from the United States, also published in 2003, reviewed the use of cryoprecipitate in 51 patients Cardiac surgery was the single most common indicaand found 76 percent to be appropriate when different tion for cryoprecipitate transfusion and represented 45 appropriateness criteria were applied.2 This group considpercent (274/603) of all cryoprecipitate transfusions. The percentage of cardiac patients receiving cryoprecipitate in ered cryoprecipitate appropriate for a brinogen level of each institution and percentage appropriateness data are less than 1.0 g per L, tissue plasminogen activatorrelated shown in Table 4. hemorrhage, transfusion of 10 or more RBC units irrespective of the brinogen level, uremic bleeding, topical use, and von Willebrand disease (VWD). In our audit, the latter DISCUSSION ve indications were considered inappropriate. The use of cryoprecipitate for VWD has been replaced with the use of Although recommendations for the appropriateness of desmopressin or virally inactivated human-derived factor cryoprecipitate transfusion are based solely on the VIII/von Willebrand factor concentrates.8 The use of cryoopinion of medical experts in the eld of transfusion medicine because no randomized controlled clinical trials precipitate as a brin glue has largely been superseded evaluating the efcacy of cryoprecipitate in this setting by the use of virally inactivated human-derived brin sealhave ever been performed, cryoprecipitate is generally ants.9 Data to support the use of cryoprecipitate for the considered appropriate for patients with bleeding and sigprevention and treatment of uremic bleeding are limited nicant hypobrinogenemia (brinogen 0.8-1.0 g/L).4-7 to small case series.10,11 Only half of uremic patients have an improvement in bleeding time after cryoprecipitate In a prospective 2-month audit of consecutive cryoadministration.12 Because cryoprecipitate carries the risk precipitate transfusions at 25 large institutions in Canada,
TABLE 1. Clinical use of cryoprecipitate (all events)
2124 TRANSFUSION Volume 48, October 2008

CRYOPRECIPITATE UTILIZATION

TABLE 3. Interhospital comparison of cryoprecipitate use and appropriateness of cryoprecipitate transfusions

Number of acute care beds 327 416 2089 410 177 317 1275 524 710 371 452 624 723 365 710 450 332 775 519 800 228 1380 440 546 280 Ratio cryoprecipitate units: 100 RBC units transfused* 8.9 7.4 15.8 3.3 10.5 27.4 7.5 1.7 5.2 2.3 6.4 6.1 11.1 7.9 3.1 11.3 9.5 8.1 13.0 6.0 3.2 7.4 3.8 7.7 12.7 8.8 Cryoprecipitate transfusion events evaluated (total 453) 3 16 48 6 2 32 11 3 14 4 13 17 19 8 2 14 6 11 23 9 4 96 8 12 72 Cryoprecipitate transfusion appropriateness Appropriate Likely inappropriate Undetermined 0/3 (0) 0/3 (0) 3/3 (100) 1/16 (6) 5/16 (31) 10/16 (63) 11/48 (23) 17/48 (35) 20/48 (42) 1/6 (17) 2/6 (33) 3/6 (50) 1/2 (50) 1/2 (50) 0/2 (0) 5/32 (16) 23/32 (72) 4/32 (13) 6/11 (55) 4/11 (36) 1/11 (9) 1/3 (33) 2/3 (67) 0/3 (0) 6/14 (43) 6/14 (43) 2/14 (14 2/4 (50) 1/4 (25) 1/4 (25) 4/13 (31) 4/13 (31) 5/13 (38) 2/17 (12) 13/17 (76) 2/17 (12) 0/19 (0) 3/19 (16) 16/19 (84) 4/8 (50) 4/8 (50) 0/8 (0) 2/2 (100) 0/2 (0) 0/2 (0) 0/14 (0) 0/14 (0) 14/14 (100) 2/6 (33) 0/6 (0) 4/6 (67) 0/11 (0) 0/11 (0) 11/11 (100) 5/23 (22) 14/23 (61) 4/23 (17) 2/9 (22) 2/9 (22) 5/9 (56) 1/4 (25) 2/4 (50) 1/4 (25) 19/96 (20) 25/96 (26) 52/96 (54) 2/8 (25) 3/8 (38) 3/8 (38) 0/12 (0) 5/12 (42) 7/12 (58) 31/72 (43) 18/72 (25) 23/72 (32) 108/453 (24) 154/453 (34) 191/453 (42)

Hospital A B C D E F G H I J K L M N O P Q R S T U V W X Y Mean

* Includes units from all cryoprecipitate transfusion events administered during the audit period. Includes only the rst cryoprecipitate transfusion event administered per patient. Data are reported as number (%).

TABLE 4. Cryoprecipitate use in cardiac surgery*

Number of adult cardiovascular surgical procedures (2-month period) 250 189 245 200 350 250 100 184 83 160 183 242 117 2553 Percentage (number) appropriateness undetermined 20 (1/5) 42 (8/19) 9 (1/11) 33 (1/3) 0 (0/1) 50 (1/2) 0 (0/3) 100 (12/12) 100 (8/8) 100 (5/5) 13 (2/15) 83 (10/12) 63 (5/8) 52 (54/104)

Hospital B C F G I J L M P R S V X Total

Number (%) patients receiving cryoprecipitate 5 (2.0) 19 (10.1) 11 (4.5) 3 (1.5) 1 (0.3) 2 (0.8) 3 (3.0) 12 (6.5) 8 (9.6) 5 (3.1) 15 (8.2) 12 (5.0) 8 (6.9) 104 (4.1)

Percentage (number) appropriate 20 (1/5) 5 (1/19) 18 (2/11) 33 (1/3) 100 (1/1) 0 (0/2) 0 (0/3) 0 (0/12) 0 (0/8) 0 (0/5) 0 (0/15) 0 (0/12) 0 (0/8) 6 (6/104)

Percentage (number) likely inappropriate 60 (3/5) 53 (10/19) 73 (8/11) 33 (1/3) 0 (0/1) 50 (1/2) 100 (3/3) 0 (0/12) 0 (0/8) 0 (0/5) 87 (13/15) 17 (2/12) 38 (3/8) 42 (44/104)

* Fifteen patients received more than one cryoprecipitate transfusion. Analysis of appropriateness was limited to the initial cryoprecipitate transfusion event for each patient.

of viral transmission and is not universally effective, it has been largely replaced by other modalities.13 Applying our criteria to the US report would change the appropriate rate to 52 percent of all cryoprecipitate transfusions in that report. The use of cryoprecipitate should be balanced with the potential adverse effects. Cryoprecipitate carries the

same risk of viral transmission as other components, unit for unit. However, cryoprecipitate is administered in a dosage of 1 unit per 5 to 10 kg of body weight and therefore a typical adult dose results in an exposure to 8 to 12 donors.3 Thus, physicians should be made aware of the potential risks of this product and they should be informed that the product has not undergone any viral
Volume 48, October 2008 TRANSFUSION 2125

ALPORT ET AL.

inactivation strategies. During the period of our audit, up to 72 percent of patient exposures likely only carried risk, with no perceivable benet. Practice guidelines for cryoprecipitate have been available in the medical literature since the mid-1990s.5 Despite their reiteration in subsequent published guidelines, inappropriate use continues. In our report, we found an association between the use of guidelines or an approval process and better compliance with published guidelines. Obstacles to appropriate use include inadequate laboratory resources to allow for rapid measurement of the brinogen level, unstable clinical situations precluding measurement of the brinogen level, and lack of physician knowledge regarding this component. It is also difcult to convince many clinicians that cryoprecipitate is unnecessary in different clinical situations because supporting scientic literature is lacking. Clearly we need additional descriptive studies to determine the reasons why clinicians are deviating from the published literature and high-quality studies evaluating the efcacy of this blood component in different clinical settings. Given that low brinogen levels occur in wellrecognized clinical settings (disseminated intravascular coagulation, massive transfusion, cardiac surgery), that serum brinogen is a routine test and should be available with a short turnaround time, and that recipients will be exposed to the risks of receiving 8 to 12 allogeneic blood products, it is recommended that except in truly urgent situations, cryoprecipitate should only be used as a replacement product for documented brinogen deciency. This approach is consistent with the expectation that whenever readily available, objective measurements should be used to support clinical decisions. In situations where brinogen is not preordered, consideration should be given, in dened clinical settings, to implementation of a thaw and hold policy until the brinogen result is available (thawing, reconstitution with saline, and pooling requires 25-30 min). Although this will result in wastage of product, it will prevent the unnecessary exposure to blood, while preventing delay in the administration of the cryoprecipitate in institutions with slow turnaround time for brinogen levels. Each unit of cryoprecipitate contains 0.25 g of brinogen in 10 to 15 mL of plasma; thus 8 units of cryoprecipitate contains 2.0 g of brinogen in 80 to 120 mL of plasma. In comparison, 4 units of FFP contains approximately 2.0 g of brinogen. Thus, in a bleeding patient with a signicant coagulopathy, the administration of fresh-frozen plasma will not only replace the coagulation factors, but will provide a brinogen dosage approximately equivalent to that of 8 units of cryoprecipitate. Thus, it seems reasonable in a bleeding situation to consider the administration of cryoprecipitate in a patient who has hypobrinogenemia, only in the absence of signicant deciencies of other coagulation factors or if the hypo2126 TRANSFUSION Volume 48, October 2008

brinogenemia is excessive compared to deciencies of other coagulation factors. The limitations of this report include the limited clinical data collected (preventing in depth analysis of clinical outcomes) and inadequate number of pediatric sites to allow for presentation of the data in the pediatric population without compromising anonymity of the participating sites. Also, the data were collected at participating sites by the transfusion service, and therefore any recording of clinical outcomes or bleeding status could not be reliably included in our evaluations, thus preventing a combined analysis of hypobrinogenemic in the presence or absence of bleeding. In addition, in 42 percent of infusions there was no monitoring of brinogen levels, and therefore we do not know which of these were appropriate. However, when a physician administers blood products he or she should be monitoring for evidence of successful repletion of the missing component, and hence the lack of monitoring is in itself inappropriate. The strengths of this report include the large number of hospitals surveyed and the large number of cryoprecipitate transfusion events evaluated. To our knowledge, this is the largest cross-sectional analysis of the use of this blood component that has been reported. Future studies should attempt to determine the impact of cryoprecipitate on important clinical outcomes (bleeding and mortality), and the impact of turnaround time of brinogen results and transfusion guidelines on appropriateness of use. There is a distinct lack of reports on the use of cryoprecipitate and we hope this report will prompt other centers to perform additional research in this area. With respect to the institutions in our jurisdiction that participated in the audit, site-specic as well as aggregate results along with published clinical transfusion guidelines were provided after the audit period with the intent of encouraging optimal practices with respect to cryoprecipitate transfusion.

ACKNOWLEDGMENT The authors acknowledge the participation of the 25 Canadian Hospitals that took part in this study. To preserve the anonymity of the participating institutions, their institutional names are not listed in this report.

REFERENCES
1. Schoeld WN, Rubin GL, Dean MG. Appropriateness of platelet, fresh frozen plasma and cryoprecipitate transfusion in New South Wales public hospitals. Med J Aust 2003; 178:117-21. 2. Pantanowitz L, Kruskall MS, Uhl L. Cryoprecipitate: patterns of use. Am J Clin Pathol 2003;119:874-81.

CRYOPRECIPITATE UTILIZATION

3. Canadian Blood Services/Socit canadienne du sang. Cryoprecipitated AHF, LR. In: Clinical Guide to Transfusion, 4th ed. Toronto: Savattuq Inc, 2006, 28-31. 4. OShaughnessy DF, Atterbury C, Bolton Maggs P, Murphy M, Thomas D, Yates S, Williamson LM; British Committee for Standards in Haematology, Blood Transfusion Task Force. Guidelines for the use of fresh-frozen plasma, cryoprecipitate and cryosupernatant. Br J Haematol 2004;126: 11-28. 5. Development Task Force of the College of American Pathologists. Practice parameter for the use of fresh-frozen plasma, cryoprecipitate, and platelets. Fresh-Frozen Plasma, Cryoprecipitate, and Platelets Administration Practice Guidelines Development Task Force of the College of American Pathologists. JAMA 1994;271:777-81. 6. American Society of Anesthesiologists Task Force on Blood Component Therapy. Practice guidelines for blood component therapy. Anesthesiology 1996;84:732-47.

7. British Committee for Standards in Haematology. Guidelines for the management of massive blood loss. Br J Haematol 2006;135:634-41. 8. Mannucci PM. Treatment of von Willebrands disease. N Engl J Med 2004;351:683-94. 9. Fattahi T, Mohan M, Caldwell GT. Clinical applications of brin sealants. J Oral Maxillofac Surg 2004;62:218-24. 10. Janson PA, Jubelirer SJ, Weinstein MJ, Deykin D. Treatment of the bleeding tendency in uremia with cryoprecipitate. N Engl J Med 1980;303:1318-22. 11. Maierhoter W, Adams MB, Kleinman JG, Roth DA. Treatment of the bleeding tendency in uremia with cryoprecipitate [letter]. N Engl J Med 1981;305:645. 12. Triulzi DJ, Blumberg N. Variability in response to cryoprecipitate treatment for hemostatic defects in uremia. Yale J Biol Med 1990;63:1-7. 13. Weigert AL, Schafer AI. Uremic bleeding: pathogenesis and therapy. Am J Med Sci 1998;316:94-104.

Volume 48, October 2008

TRANSFUSION 2127