This Article Abstract Résumé de cet Article Full Text (PDF) Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hutton, B. Articles by Hébert, P. C. Search for Related Content PubMed PubMed Citation Articles by Hutton, B. Articles by Hébert, P. C.

Transfusion rates vary significantly amongst Canadian medical centres

[Les taux de transfusion varient de façon significative dans les centres médicaux canadiens]

Brian Hutton, MSc^*,, Dean Fergusson, MHA PhD^*,,, Alan Tinmouth, MD FRCPC MSc^*,,,, Lauralyn McIntyre, MD FRCPC^*,,, Andrew Kmetic, MSc PhD^*, and Paul C. Hébert, MD FRCPC MHSc^*,,

^* From the Clinical Epidemiology Program, Ottawa Health Research Institute;
the University of Ottawa Centre for Transfusion Research;
the University of Ottawa, Faculty of Medicine; and
the Canadian Blood Services, Ottawa, Ontario, Canada.

Address correspondence to: Dr. Paul Hébert, Centre for Transfusion Research, Ottawa Hospital, General Campus, 501 Smyth Road, Box 201, Ottawa, Ontario K1H 8L6, Canada. Phone: 613-737-8197; Fax: 613-739-6266; E-mail: phebert{at}ohri.ca

	Abstract

TOP Abstract Introduction Methods Results Discussion References

 
Purpose: To document variation of transfusion practice following repair of hip fracture or cardiac surgery, as well as those requiring intensive care following a surgical intervention or multiple trauma (high risk patients).

Methods: We documented rates of allogeneic red cell transfusion in 41,568 patients admitted to 11 hospitals across Canada between August 1998 and August 2000 as part of a retrospective observational cohort study. In the subgroup of 7,552 patients receiving red cells, we also compared mean nadir hemoglobin concentrations from centre to centre.

Results: The overall rate of red cell transfusion was 38.7%, and ranged from 23.8% to 51.9% across centres among the 41,568 perioperative and critically ill patients. Women were more likely to be transfused (43.7% vs 35.3%, P < 0.0001), with higher rates of transfusion in eight of 11 centres. Compared to a chosen reference hospital having a crude transfusion rate near the median, the adjusted odds of transfusion ranged from 0.44 to 1.53 overall, from 0.42 to 1.22 in patients undergoing a hip fracture repair, from 0.72 to 3.17 in cardiac surgical patients undergoing cardiac surgery, and from 0.27 to 1.11 in critically ill and trauma patients. In the 7,552 transfused patients, the mean adjusted nadir hemoglobin was 74.0 ± 4.83 g•L^–1 overall, and ranged from 66.9 ± 1.7 g•L^–1 to 84.5 ± 1.6 g•L^–1 across centres. Similar differences among centres were observed amongst hip fracture patients (71.2 ± 2.9 g•L^–1 to 82.8 ± 1.7 g•L^–1), cardiac surgical patients (65.7 ± 1.1 g•L^–1 to 77.3 ± 1.0 g•L^–1) and critically ill and trauma patients (66.1 ± 3.04 g•L^–1 to 87.5 ± 2.5 g•L^–1).

Conclusion: We noted significant differences in the rates of red cell transfusion and nadir hemoglobin concentrations in various surgical and critical care settings.

	Introduction

TOP Abstract Introduction Methods Results Discussion References

 
DESPITE a number of blood conservation strategies and ongoing concerns regarding transfusion-transmitted infections, red blood cells (RBC) remain an important life-saving therapy in perioperative and critical care.¹ Indeed, a significant proportion of the entire blood supply is administered to critically ill patients, following multiple trauma and in the perioperative setting, particularly following hip fracture repair or cardiac surgery. Brien et al.² documented that 56% of all red cells were administered to surgical patients, while a study by Ghali³ documented that 69% of red cells are transfused into surgical patients.

Exploring practice variation in various clinical settings has become an important tool in the assessment of health services. If significant practice variation is documented, this should provide the impetus to explore reasons for such variation, which may include the lack of high quality evidence, or where evidence is solid, problems in the dissemination or uptake of the existing evidence. In transfusion medicine, significant practice variation has been documented within specific disease categories,^4,5 clinical settings and surgical procedures,⁶ including hip^7–10 and knee^7,9,10 arthroplasty and coronary revascularization [or coronary artery bypass grafting (CABG)].^11–13 Controlling for population differences, blood loss and pump time, a prospective audit conducted on CABG patients¹³ identified that transfusion factors such as the nadir and discharge hematocrits accounted for significant variation in blood use among 18 tertiary care hospitals. In similar patients, Surgenor and colleagues¹¹ found that there were significant differences in the percentage of patients transfused between hospitals. Hébert et al. found a significant variation in transfusion practice (in terms of nadir hemoglobin concentrations) among six Canadian intensive care units (ICUs) after controlling for the effects of disease severity, diagnosis, age and gender. Thus, significant differences in the approach to allogeneic red cell transfusions have been consistently observed among institutions. However, there are few recent studies exploring practice variation, despite the publication of clinical trials addressing this issue and significant improvements in blood safety.^14,15 Therefore, we examined transfusion practice across 11 Canadian academic and community hospitals caring for the critically ill, as well as patients undergoing cardiac surgical procedures and hip fracture repair.

	Methods

TOP Abstract Introduction Methods Results Discussion References

 
Study design and patient population
Clinical information was collected as part of a large retrospective observational cohort examining clinical outcomes before and after the implementation of a universal pre-storage leuko-reduction program involving Canada’s entire blood supply.¹⁶ In the leuko-reduction study, we identified all consecutive patients in three distinct high-risk categories: 1) patients requiring intraoperative repair of a hip fracture; 2) patients following cardiovascular surgery requiring cardiopulmonary bypass; and 3) postoperative and multiple trauma patients admitted to an ICU.

Patients were identified from the health records of participating hospitals as well as clinical databases at all participating institutions. We excluded patients who: were less than 16 yr of age; had been previously included in this study; died within the first 24 hr of hospital admission; did not survive 24 hr following the completion of the index surgical procedure or time of admission to the ICU. For the evaluation of nadir hemoglobin concentrations in transfused patients, we also excluded patients admitted with a diagnosis of infection prior to the first transfusion, patients who had previously been diagnosed with a hematological malignancy, or were considered brain dead.

We examined variation in allogeneic RBC transfusion practice amongst 41,568 patients seen at 11 major Canadian hospitals.

Data collection
The original study was approved by the Ethics Review Board at each participating centre. Using the eligibility criteria listed above, we identified all patients between June of 1998 and September of 2000 using either electronic hospital health records databases or information collected from the leuko-reduction study. In the original leuko-reduction study, we collected information from patient medical charts using a standardized case report form and a detailed procedures manual. All personnel performing data abstraction participated in a training session to ensure quality of the data, and were subsequently given a dummy protocol in order to mask the intent of the initial study. Patient information abstracted and relevant to the current study included patient indication for transfusion (i.e., repair of hip fracture, cardiac surgery or admission to the ICU following surgery or multiple trauma), gender, date of birth, date of hospital admission, medical centre at which the patient was seen, nadir hemoglobin level, total number of transfusions administered, leuko-reduction status of blood products administered (i.e., prior to or following the implementation of universal leuko-reduction), date of hospital discharge, vital status at discharge and length of stay in hospital. Information for patients who did not receive allogeneic red cells included the procedure or diagnosis, the centre, patient age and gender, dates of admission, surgery and hospital discharge and survival status at hospital discharge.

Statistical analysis
We compared baseline variables of transfused and non-transfused patients by evaluating measures of central tendency and dispersion. As an initial exploration of gender differences, we compared the proportion of patients receiving transfusions overall and by condition using a stratified Chi-square. To explore centre effects, we calculated rates and unadjusted odds ratios (OR; along with 95% confidence intervals) for each centre and for each major diagnostic grouping (ICU, cardiac surgery and hip fracture) within each centre. As a second step, we used logistic regression to calculate the adjusted odds of transfusion for each centre using centre J as a reference while adjusting for diagnostic grouping, age and gender. Site J was chosen because it was a large centre with a crude transfusion rate close to the median of the 11 sites, both overall and within each patient subgroup.

Nadir hemoglobin values in patients receiving allogeneic red cells were also used in the evaluation of centre effects. Adjusted mean values and 95% confidence intervals for nadir hemoglobin concentrations were calculated for all centres using analysis of covariance, in all transfused patients and within each of the three major diagnostic subgroups of interest. Explanatory variables included centre, gender, age, the presence of one of 12 major co-morbid conditions (ischemic heart disease, congestive heart failure, AIDS, severe immunosuppression, metastatic cancer, severe dementia, chronic renal failure, severe lung disease, disabling stroke, severe cirrhosis/hepatic failure, diabetes, mellitus and obesity), and severity of illness using APACHE II scores in the ICU patient subgroup.

All statistical analyses were carried out using SAS, version 8.1 (SAS Institute Inc., Cary, NC, USA).

	Results

TOP Abstract Introduction Methods Results Discussion References

 
A total of 41,736 consecutive patients from 11 centres were identified between June 1st, 1998 and September 30th 2000. One hundred and sixty eight patients (0.4%) were excluded from the analysis of transfusion rates due to incomplete or incorrect data regarding patient type (n = 159), age (n = 5) and gender (n = 4). During this time interval, we enrolled 3,945 (9.5%) hip fracture patients, 11,812 (28.4%) cardiac surgical patients and 25,811 (62.1%) postoperative and critically ill patients (Table). Average ages and the proportion of women within each diagnostic group were comparable across centres. In a subset of 7,552 transfused patients, additional information was available on nadir hemoglobin levels and the number of transfusions administered during the first 30 days of hospitalization. In this subset, 35 (0.5%) patients had missing information on the presence of co-morbid illnesses; for these cases we assumed that there were no relevant co-morbid illnesses.

View larger version (51K): [in this window] [in a new window]   FIGURE 1 Gender-specific and overall crude rates of transfusion across the 11 centres included in the study for the four groups of patients of interest: (a) overall (n = 41,568), (b) repair of hip fracture patients (3,945), (c) cardiac surgery patients (n = 11,812) and (d) patients receiving post-surgical or post-traumatic intensive care unit care (n = 25,811). The horizontal dashed lines indicate the median, first quartile and third quartile transfusion rates amongst the 11 hospitals studied. ICU = intensive care unit.

View larger version (17K): [in this window] [in a new window]   FIGURE 2 Between-centre variation in the odds of transfusion both overall, and within the three patient subgroups of interest. Site J was selected as the reference site for all comparisons, and statistical adjustments for the effects of age, gender and indication were made. ICU = intensive care unit.

View larger version (55K): [in this window] [in a new window]   FIGURE 3 Distribution of the median number of transfusions amongst transfused patients at each of the 11 participating centres, along with their associated interquartile ranges. ICU = intensive care unit.

Cardiac surgical patients
The 11,812 cardiac surgical patients were gathered from seven of 11 centres participating in this study. The proportion of cardiac surgical patients that received one or more transfusions at these sites was 43.8% overall, and ranged from 28.1% to 60.5% across centres. Transfused patients were older than non-transfused patients (68.3 ± 10.3 yr vs 62.9 ± 11.7 yr), and females were transfused more frequently than males (61.3% vs 36.6%, P < 0.0001; Figure 1). After adjustment for the effects of age and gender, the degree of variation across hospitals relative to centre J ranged from adjusted odds of 0.72 to 3.17. Centre B was associated with significantly lower odds of transfusion relative to site J, while sites A, E, G and I were associated with significantly higher adjusted odds of transfusion. The median number of transfusions per patient ranged from two units to three units across centres (Figure 3).

ICU patients
Amongst the 25,811 patients admitted to ICU for postoperative and trauma care, 36.9% received one or more red cell units, with rates varying between 20.4% and 53.4% between sites. Transfused patients were older than non-transfused patients (63.5 ± 16.7 yr vs 56.8 ± 20.9 yr), and females were transfused more frequently than males (40.1% vs 34.9%, P < 0.0001; Figure 1). After adjustment for the effects of age and gender, the adjusted odds of transfusion as compared to centres J ranged from 0.27 to 1.11. Centres A, B, C, D, E, H and I had significantly lower odds of transfusion relative to centre J, while the remaining centres had comparable odds of transfusion (Figure 2). The median number of transfusions per patient ranged between two units and six units across centres (Figure 3).

Nadir hemoglobin levels within patient groups
In the subset of 7,552 patients receiving red cell transfusions, the mean adjusted nadir hemoglobin was 74.0 ± 4.83 g•L^–1, with a range of 66.9 ± 1.7 g•L^–1 to 84.5 ± 1.6 g•L^–1 (Figure 4). Following hip fracture repair, the mean nadir hemoglobin was 77.6 ± 11.6 g•L^–1 and ranged from 71.2 ± 2.9 g•L^–1 to 82.8 ± 1.7 g•L^–1 (Figure 4). Amongst cardiac surgical patients, the mean nadir hemoglobin concentration was 72.5 ± 10.7 g•L^–1, and ranged from 65.7 ± 1.1 g•L^–1 to 77.3 ± 1.0 g•L^–1 (Figure 4). In postoperative and trauma patients admitted to an ICU, the mean nadir hemoglobin was 74.8 ± 13.5 g•L^–1 and ranged from 66.1 ± 3.04 g•L^–1 to 87.5 ± 2.5 g•L^–1 (Figure 4). Centres C and F consistently demonstrated some of the higher mean levels across diagnostic groups, while centres D and E demonstrated lower mean levels. The remaining centres fell into the middle range, with relative positions changing marginally within each diagnostic grouping.

View larger version (15K): [in this window] [in a new window]   FIGURE 4 Adjusted, group-specific mean nadir hemoglobin levels for transfused patients at each of the participating centres in this study, along with their 95% confidence intervals. Adjustments were made for the effects of age, gender, hospital, and the presence/absence of 12 different co-morbidities listed earlier. ICU = intensive care unit.

	Discussion

TOP Abstract Introduction Methods Results Discussion References

A few different patterns of use emerged from examination of hemoglobin concentrations and rates of transfusion. Three centres (D, E and G) consistently demonstrated low mean nadir hemoglobin levels and mid-level or high odds of transfusion. This pattern might suggest that the centres had higher rates of transfusion, in patients with increased severity of illness or undergoing more complex procedures resulting in greater degrees of blood loss than other centres. Conversely, centre C, associated with high mean nadir hemoglobin levels and lower odds of transfusion may reflect lower rates of transfusion in a more stable, lower risk population centre. Many of the remaining centres were associated with high mean nadir hemoglobin levels and high odds of transfusion, perhaps suggesting a very liberal use of red cells at such centres. Given that only selected patients were included in the analysis of nadir hemoglobin concentrations, the qualitative patterns are largely speculative. There are, however, consistent centre-specific variations in transfusions in major perioperative consumers of red cells.

Prior studies have documented the independent effects of both gender and age on the likelihood of transfusion.^17–19 The increased risk of transfusion in women has been shown by some to be independent of smaller blood volume as compared to men,²⁰ while others have shown that the gender effect disappeared following inclusion of blood volume as an additional covariate in a multivariate analysis.²¹ Further research regarding this matter is needed. Regarding the effect of age, older patients are thought to be sicker than younger patients and, consequently, are transfused at a higher rate. However, the effects of both age and co-morbidities on transfusion needs are confounded by indication, thus complicating observational research of this issue.

In a systematic review published in 1998 as part of background documents for the Canadian Medical Association guidelines on red cells and blood product usage, Hébert and colleagues identified 189 articles that addressed questions related to clinical transfusion practice.^22,23 They only identified only a few studies that described overall utilization of red cells between and within different disciplines as well as patterns of use over time.^2,3,24–27 There were also a significant number of studies describing red cell utilization in selected patient populations. In a 1992 survey conducted in 61 Toronto area hospitals, 65% of the allogeneic red cells used were administered to patients undergoing operative procedures categorized as digestive and abdominal, cardiovascular and musculoskeletal.²⁴ Brien et al.² documented that 56% of all red cells were administered to surgical patients, while a study by Ghali³ documented that 69% of red cells are transfused into surgical patients. In cardiac surgery, the proportion of patients receiving red cells ranges from 50% to 80%.^8,24,28

In addition, there were few recent studies which systematically attempted to identify and characterize the source of practice variation. Several investigators identified practice variation as being an inter-institutional phenomenon. The first such study undertook a secondary analysis of a large database conducted in 1978²⁹ and observed a striking variation among hospitals in Connecticut. The authors inferred that physician habits and personal preferences determined institutional variation in blood utilization. However, others criticized the study for failing to control for the effects of case mix.³⁰ Subsequently, significant practice variation was also documented within specific disease categories,^4,5 clinical settings and surgical procedures,^6,31 including hip^7–10 and knee^7,9,10 arthroplasty and coronary revascularization (or CABG).^11–13 Controlling for population differences, blood loss and pump time, a prospective audit conducted on CABG patients¹³ identified transfusion factors such as the nadir and discharge hematocrits as accounting for significant variation in blood use among 18 tertiary care hospitals. In similar patients, Surgenor and colleagues¹¹ found that there were significant differences in the percentage of patients transfused between hospitals. Hébert et al. found a significant variation in transfusion practice (in terms of nadir hemoglobin concentrations) among six Canadian ICUs after controlling for the effects of disease severity, diagnosis, age and gender. Thus, significant differences in the approach to allogeneic red cell transfusions have been consistently observed among institutions.

Retrospective chart reviews and self-administered surveys have also attempted to determine if physicians account for significant variations. In the SANGUIS study, transfusion rates were found to depend more on physicians than the patient population or type of procedure or hospital.⁶ Wide variation was found among 43 hospitals in ten European countries³¹ and between hospitals within the same country. Some factors found to influence this variation were age, gender, preoperative hematocrit and blood loss, all factors related to patients. In observational studies in cardiac surgery, patient characteristics associated with transfusions of red cells include a long list of demographic factors such as age and gender, diagnostic characteristics such as procedure and co-morbid illnesses, factors related to blood volume such as preoperative hematocrit and body mass index, as well as other factors such as the use of antifibrinolytics and preoperative use of antiplatelet agents.³² Surveys of physicians have also explored factors that might influence practice.³³ There are, however, few studies that attempt to dissect physician-based factors and institutional factors that might explain practice variation while considering the confounding influence of all other characteristics.

This study documented practice variation in three major diagnostic groups among 11 Canadian centres while controlling for a few important factors that might influence the administration of red cells. However, like many other studies, we were unable to secure information on a number of important variables such as specific procedures, co-morbid illnesses and indicators of severity of illnesses. In addition, we were unable to identify physician-based factors that might have affected transfusion practice within the same institution.

In conclusion, we documented inter-institutional practice variation in three major populations of patients who frequently receive red cells. We suggest that future studies should attempt to characterize the source of variation. In addition, if variation truly exists, future studies should define optimal transfusion practice in a variety of clinical settings by conducting randomized controlled trials.

	Acknowledgments

 
The authors thank Nadine Shehata for her expert opinion during the preparation of this manuscript, as well as Nancy Cleary for her superb secretarial assistance.

Listing of leuko-reduction investigators
Regina Qu’Appelle Health Region, Regina, Saskatchewan: E. C. Alport; Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia: D. Anderson; McMaster University/Hamilton Health Sciences, Hamilton, Ontario: M. A. Blajchman; Sunnybrook and Women’s College, Toronto, Ontario: J. Callum; London Health Sciences Centre, London, Ontario: I. Chin-Yee; Foothills Hospital, Calgary, Alberta: D. Easton; Mount Sinai Hospital, Toronto: B. Fernandes; St Michael’s Hospital, Toronto: J. Freedman; Toronto East General Hospital, Toronto: J. Lentz; Institut de Cardiologie de Montréal, Montréal, Québec: R. Martineau; St John Regional Hospital, St John, New Brunswick: C. Norman; St Paul’s Hospital, Vancouver, British Columbia: D. Pi; Lakeridge Health Corporation, Oshawa, Ontario: M. Quantz; Jewish General Hospital, Montréal: S. Caplan; University of Ottawa Heart Institute, Ottawa, Ontario: J. Robblee; The Ottawa Hospital-Civic Campus, Ottawa: G. Rock; Hôpital Enfant-Jésus, Québec City, Québec: C. Shields; Hôpital Laval, Ste-Foy, Québec: B. Villeneuve; General Hospital, St John’s, Newfoundland: L. Whitman.

	Footnotes

 
This study was in part sponsored by a grant from the Canadian Institutes of Health Research and an unrestricted grant from Ortho-Biotech, Canada.

Accepted for publication October 4, 2004. Revision accepted February 15, 2005.

	References

TOP Abstract Introduction Methods Results Discussion References

 
1 Carson JL, Duff A, Poses RM, et al. Effects of anaemia and cardiovascular disease on surgical mortality and morbidity. Lancet 1996; 348: 1055–60.[Medline]

2 Brien WF, Butler RJ, Inwood MJ. An audit of blood component therapy in a Canadian general teaching hospital. CMAJ 1989; 140: 812–5.[Abstract]

3 Ghali WA, Palepu A, Paterson WG. Evaluation of red blood cell transfusion practices with the use of preset criteria. CMAJ 1994; 150: 1449–54.[Abstract]

4 Hasley PB, Lave JR, Hanusa BH, et al. Variation in the use of red blood cell transfusions. A study of four common medical and surgical conditions. Med Care 1995; 33: 1145–60.[Medline]

5 Surgenor DM, Wallace EL, Churchill WH, Hao S, Hale WB, Schnitzer J. Utility of DRG and ICD-9-CM classification codes for study of transfusion issues. Transfusions in patients with digestive diseases. Transfusion 1989; 29: 761–7.[Medline]

6 Baele PL, De Bruyere M, Deneys V, et al. Results of the SANGUIS study in Belgium. A concerted action of the Commission of the European Communities IVth Medical and Health Research Programme. The Belgium SANGUIS Study Group. Safe AND Good Use of blood In Surgery. Acta Chir Belg 1994; 94(Suppl): 1–61.

7 Mintz PD, Nordine RB, Henry JB, Webb WR. Expected hemotherapy in elective surgery. N Y State J Med 1976; 76: 532–7.[Medline]

8 Friedman BA. An analysis of surgical blood use in United States hospitals with application to the maximum surgical blood order schedule. Transfusion 1979; 19: 268–78.[Medline]

9 Surgenor DM, Wallace EL, Churchill WH, Hao SH, Chapman RH, Poss R. Red cell transfusions in total knee and total hip replacement surgery. Transfusion 1991; 31: 531–7.[Medline]

10 Pinkerton PH, Coovadia AS, Seigel C. Audit of the use of packed red blood cells in association with seven common surgical procedures. Transfus Med 1992; 2: 231–4.[Medline]

11 Surgenor DM, Wallace EL, Churchill WH, Hao SH, Chapman RH, Collins JJ Jr. Red cell transfusions in coronary artery bypass surgery (DRGs 106 and 107). Transfusion 1992; 32: 458–64.[Medline]

12 Goodnough LT, Johnston MF, Shah T, Chernosky A. A two-institution study of transfusion practice in 78 consecutive adult elective open-heart procedures. Am J Clin Pathol 1989; 91: 468–72.[Medline]

13 Goodnough LT, Johnston MF, Toy PT; The Transfusion Medicine Academic Award Group. The variability of transfusion practice in coronary artery bypass surgery. JAMA 1991; 265: 86–90.[Abstract]

14 Hebert PC, Wells G, Blajchman MA, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group. N Engl J Med 1999; 340: 409–17.[Abstract/Free Full Text]

15 Kleinman S. Hepatitis G virus biology, epidemiology, and clinical manifestations: implications for blood safety. Transfus Med Rev 2001; 15: 201–12.[Medline]

16 Hebert PC, Fergusson D, Blajchman MA, et al. Clinical outcomes following institution of the Canadian universal leukoreduction program for red blood cell transfusions. JAMA 2003; 289: 1941–9.[Abstract/Free Full Text]

17 Karkouti K, Cohen MM, McCluskey SA, Sher GD. A multivariable model for predicting the need for blood transfusion in patients undergoing first-time elective coronary bypass graft surgery. Transfusion 2001; 41: 1193–1203.[Medline]

18 Covin R, O’Brien M, Grunwald G, et al. Factors affecting transfusion of fresh frozen plasma, platelets, and red blood cells during elective coronary artery bypass graft surgery. Arch Pathol Lab Med 2003; 127: 415–23.[Medline]

19 Surgenor DM, Churchill WH, Wallace EL, et al. Determinants of red cell, platelet, plasma, and cryoprecipitate transfusions during coronary artery bypass graft surgery: the Collaborative Hospital Transfusion Study. Transfusion 1996; 36: 521–32.[Medline]

20 Magovern JA, Sakert T, Benckart DH, et al. A model for predicting transfusion after coronary artery bypass grafting. Ann Thorac Surg 1996; 61: 27–32.[Abstract/Free Full Text]

21 Cosgrove DM, Loop FD, Lytle BE, et al. Determinants of blood utilization during myocardial revascularization. Ann Thorac Surg 1985; 40: 380–4.[Abstract]

22 Hebert PC, Schweitzer I, Calder L, Blajchman M, Giulivi A. Review of the clinical practice literature on allogeneic red blood cell transfusion. CMAJ 1997; 156(11 suppl): S9–26.

23 Calder L, Hebert PC, Carter AO, Graham ID. Review of published recommendations and guidelines for the transfusion of allogeneic red blood cells and plasma. CMAJ 1997; 156(11 suppl): S1–8.

24 Chiavetta JA, Herst R, Freedman J, Axcell TJ, Wall AJ, Van Rooy SC. A survey of red cell use in 45 hospitals in central Ontario, Canada. Transfusion 1996; 36: 699–706.[Medline]

25 Baele PL, De Bruyèee M, Deneys V, et al. The SANGUIS study in Belgium: an overview of methods and results. The Belgium SANGUIS Study Group. Acta Chir Belg 1994; 94: 69–74.[Medline]

26 Kuriyan M, Kim DU, Wake E, Kress S, Pachter I, Nayak S. Analysis of surgical blood use in New Jersey. N J Med 1987; 84: 251–5.[Medline]

27 Schots J, Steenssens L. Blood usage review in a Belgian university hospital. Int J Qual Health Care 1994; 6: 41–5.[Abstract/Free Full Text]

28 Pinkerton PH, Coovadia AS, Downie H. Transfusion practice in support of surgery during introduction of a hospital-based autologous presurgical blood donor program. Can J Surg 1995; 38: 154–60.[Medline]

29 Palermo G, Bove J, Katz AJ. Patterns of blood use in Connecticut. Transfusion 1980; 20: 704–10.[Medline]

30 Welch HG, Meehan KR, Goodnough LT. Prudent strategies for elective red blood cell transfusion. Ann Intern Med 1992; 116: 393–402.

31 Sirchia G, Giovanetti AM, McClelland DB, Fracchia GN. Safe and Good Use of Blood in Surgery (SAN-GUIS): Use of blood and artificial colloids in 43 European hospitals. Luxembourg: European Comission; 1994.

32 Khanna MP, Hebert PC, Fergusson DA. Review of the clinical practice literature on patient characteristics associated with perioperative allogeneic red blood cell transfusion. Transfus Med Rev 2003; 17: 110–9.[Medline]

33 Hebert PC, Wells G, Martin C, et al. A Canadian survey of transfusion practices in critically ill patients. Crit Care Med 1998; 26: 482–7.[Medline]

This article has been cited by other articles:

				N. Shehata, K. Wilson, C. D. Mazer, G. Tomlinson, D. Streiner, P. Hebert, and G. Naglie The proportion of variation in perioperative transfusion decisions in Canada attributable to the hospital: [La proportion de variation dans les pratiques transfusionnelles perioperatoires au Canada imputable aux hopitaux] Can J Anesth, November 1, 2007; 54(11): 902 - 907. [Abstract] [Full Text] [PDF]

				A. Shander and D. Moskowitz Hydroxyethyl starch and risk of bleeding: the missing links. Anesth. Analg., March 1, 2006; 102(3): 968 - 968. [Full Text] [PDF]

This Article Abstract Résumé de cet Article Full Text (PDF) Submit a scholarly reply Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Hutton, B. Articles by Hébert, P. C. Search for Related Content PubMed PubMed Citation Articles by Hutton, B. Articles by Hébert, P. C.