G Model
ACCPM-48; No. of Pages 5 Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Original Article
Hospital audit of delayed transfusion after orthopaedic surgery§ Cle´ment Dubost a, Sylvain Ausset b, Ce´sar Vincent a, Charles Gozlan a, Yves Auroy c, Loan N’Guyen d, Andre´ Lienhart e, Dan Benhamou a,*,1 a
De´partement d’anesthe´sie-re´animation, hoˆpital Biceˆtre, AP–HP, 78, rue du Ge´ne´ral-Leclerc, 94275 Le Kremlin-Biceˆtre cedex, France De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es Percy, 101, avenue Henri-Barbusse, BP 406, 92141 Clamart cedex, France c De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es du Val-de-Grace, 74, boulevard de Port-Royal, 75230 Paris, France d Service d’he´movigilance, hoˆpital de la Pitie´-Salpe´trie`re, 47-81, boulevard de l’Hoˆpital, 75651 Paris cedex 13, France e De´partement d’anesthe´sie-re´animation, hoˆpital Saint-Antoine, AP–HP, 184, faubourg Saint-Antoine, 75571 Paris, France b
A R T I C L E I N F O
A B S T R A C T
Article history: Available online xxx
Background and objectives: To understand the mechanisms related to both the onset and correction of severe anaemia after orthopaedic surgery, we analysed all the full blood counts (FBCs) for patients on one orthopaedic ward during a one-year period in an academic hospital. Methods: FBCs were screened and the medical records of those patients for whom a postoperative haemoglobin (Hb) concentration below 8 g/dL was recorded at least once were reviewed. The onset of postoperative anaemia was determined by calculating the various time intervals delineated by surgery, the time at which the transfusion threshold was reached and the time at which the lowest Hb level (nadir) and transfusion (if any) occurred. Results: A total of 6573 FBCs drawn from 1255 patients were screened. The medical records of 74 consecutive patients with at least one Hb value < 8 g/dL were analysed. The postoperative Hb nadir was 7.4 ( 0.6) g/dL (mean – SD). The medians (IQR 25–75) of the calculated intervals were: (surgery – nadir): 72 (48–144) h, (nadir – transfusion): 7 (5–21) h and (transfusion threshold – transfusion): 26 (11– 51) h. Conclusions: Delayed transfusion (defined as > 12 hours between the time at which the transfusion threshold was reached and actual transfusion) was observed in 57% of severely anaemic patients after orthopaedic surgery. ß 2015 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Keywords: Anaemia Postoperative period Transfusion Risk
1. Introduction Postoperative anaemia is a major challenge for patient safety. Indeed, a survey of anaesthesia-related mortality in 2006 in France showed that nearly one quarter of deaths partially or totally related to anaesthesia were the result of inadequate management of postoperative anaemia [1]. A thorough investigation of these deaths revealed that they did not usually occur during surgery but rather during the third or fourth postoperative day, related to a slowly occurring anaemia with major end-organ
§ Preliminary data were presented in part at the annual meeting of the European Society of Anaesthesiology in Amsterdam, June 11–14, 2011. * Corresponding author. Tel.: +33 1 45 21 39 35; fax: +33 1 45 21 28 75. E-mail address: [email protected] (D. Benhamou). 1 De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es Be´gin, 69, avenue de Paris, 94160 Saint-Mande´, France.
dysfunction, such as myocardial infarction [2]. The impact of preoperative anaemia on patient outcomes has been well studied [3–6] but data concerning the postoperative period are more limited. Acute postoperative anaemia has been found in many institutions to be most common following orthopaedic surgery and thus is the main indication for blood transfusion in surgical patients [2,7]. Post-operative anaemia may worsen post-operative outcomes and increase both mortality and morbidity [4]. Although transfusion thresholds have been defined by recommendations in various countries [8,9], risk might also be related to the duration of postoperative anaemia. Therefore, delayed transfusion might also increase risk [10]. In order to establish the timing of severe postoperative anaemia (i.e. which reached the transfusion threshold), we analysed all full blood counts (FBC) drawn from elective and emergency patients on one orthopaedic ward during a one-year period in a major teaching and trauma hospital.
http://dx.doi.org/10.1016/j.accpm.2015.02.006 2352-5568/ß 2015 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
2
2. Methods
medians and 25th–75th percentiles or means standard deviation (SD) (according to the observed distribution).
Database analysis was performed in three steps. First, the FBC results obtained from patients in the orthopaedic ward were extracted from the Haematology Laboratory database. Then, data from patients with at least one postoperative haemoglobin (Hb) level < 8 g/dL were extracted and merged with that of the transfusion centre. Finally, the medical records of these patients were thoroughly peer-reviewed by a pair of experienced physicians. Data were recorded in a preformed Excel1 table (Microsoft Excel1 2008 for Mac) after group validation. According to French recommendations for red blood cell (RBC) transfusion, the desirable transfusion threshold has been defined as 8 g/dL in patients without a history of acute coronary artery disease [8]. The timing of postoperative blood checks was decided by the physician in charge, following national and institutional recommendations [11]. For the medical records analysed, the data recorded were: the date and type of surgery, age, sex and American Society of Anesthesiologists (ASA) status, data from the patient’s medical history pertinent to anaemia and transfusion – especially cardiovascular history. The Revised Cardiac Risk Index (RCRI) was determined for each patient [12]. The occurrence of any perioperative event that might be linked to anaemia (cause or consequence) was searched for by carefully reading medical and nursing files. The lowest postoperative Hb level (nadir) and the time at which it occurred were recorded as well as the amount and timing of intra- or postoperative transfusion. The patient files, containing all the results of biological tests, were retrospectively analysed. Only the validated, paper versions of results were analysed. The time at which blood transfusion was prescribed, the time of delivery of RBCs by the blood bank and the time of transfusion onset were also recorded. For each patient, a graphical representation was used to estimate the time at which the Hb level reached the transfusion threshold (Fig. 1). When no blood result was available on the day of threshold crossing, intervals were estimated based on the hypothesis of a linear decrease in Hb. From the above data, the following time intervals were calculated: (surgery – threshold), (threshold – nadir), (nadir – transfusion [if any]), (nadir – prescription), (prescription – delivery) and (delivery – transfusion). Transfusion was considered to have been delayed when more than 12 hours had elapsed between reaching the transfusion threshold and the actual transfusion performance. Statistical analyses were performed using Prism1 4.00 for Mac (Inc GraphPad Software, San Diego California USA, http://www. graphpad.com). Data are presented as percentages or using
A total of 6573 FBCs drawn from 1255 patients from a single centre were screened (Fig. 2). Anaemia with at least one Hb value lower than 10 g/dL was found in 456 patients (36%), while 162 (13%) exhibited at least one Hb value < 8 g/dL, among whom 88 were excluded for the following reasons: no surgery (n = 33), admission to the ICU (n = 20) and missing data (n = 25). The medical records of the remaining 74 patients were analysed. Patient characteristics are presented in Table 1. In 14 patients (19%), an intra-operative event related to haemorrhage (such as a drop in blood pressure or high blood volume collected in the suction drains) was observed. Nineteen patients (26%) received RBCs during surgery. A Hb measurement was performed during surgery or in the Post-Anaesthesia Care Unit (PACU) in 50 patients (68%). The median intraoperative Hb was 9.2 g/dL. Fifty-five patients did not receive RBCs during surgery (74%), 17 patients (23%) received 1–4 RBCs and 2 received more than 4 RBCs (3%). The mean postoperative Hb nadir was 7.4 (SD 0.6) g/dL. In 15 patients (20%), a postoperative event that could be related to anaemia (haematoma [6], delirium [4], chest pain [1], and low blood pressure [4]) was recorded. Among the 74 patients, 64 received a blood transfusion at least once during the postoperative period. The median number of RBCs transfused during the postoperative period was 2 (IQR 2–2). Reasons for the absence of transfusion included patient refusal, good clinical tolerance and lowering of the transfusion threshold by the physician in charge. No adverse effect was identified in the patients not receiving transfusion. Among them, the mean value of nadir Hb was 7.9 g/dL (SD 1). Delayed transfusion was observed in 42 (57%) patients. Time intervals were calculated in 64 patients. The median value (IQR) of the calculated intervals in hours is presented in Table 2. The median time between crossing the transfusion threshold and transfusion was 26 hours (11–51). The median time interval between prescription and delivery was 1 hour (0.7–2) and the median interval between delivery and transfusion was 0.8 hours (0.6–1). The time intervals (transfusion threshold – transfusion) and (nadir – transfusion) are represented in Figs. 3A and B with respect to the RCRI score. Shorter intervals were not consistently observed in patients with a high RCRI score. Thirty-two patients had an RCRI score greater than 2. Among them, postoperative complications related to anaemia were identified: haematoma (4), delirium (2) and low blood pressure (1).
Fig. 1. Graphical determination of the nadir. Graphical representation of Hb concentration from postoperative day 1 to the time of Hb nadir. This graph was used to estimate the time at which Hb transfusion threshold was reached.
Fig. 2. Flow chart of the study.
3. Results
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx Table 1 Demographic characteristics of the population studied. Age (years) ASA 1–2 ASA 3–4 Sex ratio M/F Revised Cardiac Risk Index: number of predictive factors One Two Three/Four Emergency surgery Type of surgery Hip Femur Knee Spine Others
3
Table 2 Relevant time intervals. 72,5 (44–82.5) 51 (69%) 23 (31%) 0.85 49 (66%) 21 (28%) 4 (6%) 49 (66%) 27 (37%) 9 (12%) 10 (14%) 7 (9%) 21 (28%)
Time intervals (h)
All patients (n = 64)
Transfusion > 12 h (n = 42)
Transfusion < 12 h (n = 22)
Surgery – nadir Surgery – transfusion Nadir – prescription Nadir – delivery Nadir – transfusion Threshold reached – prescription Threshold reached – delivery Threshold reached – transfusion Prescription – delivery Delivery – transfusion
72 72 5 6 7 23
72 72 6 8 10 29
72 72 3 5 6 3
(48–144) (48–120) (3–12) (4–19) (5–21) (8–48)
(48–132) (48–108) (4–26) (6–27) (7–27) (21–54)
(36–156) (36–144) (2–4) (3–7) (4–7) (2–6)
24 (10–50)
30 (23–56)
5 (4–7)
26 (11–51)
31 (23–56)
6 (4–8)
1 (0.7–2) 0.8 (0.6–1)
1 (0.7–2) 0.8 (0.6–1)
1 (0.4–2) 0.6 (0.4–0,9)
4. Discussion By analysing the medical records of 74 severely anaemic patients (i.e. Hb < 8 g/dL) after orthopaedic surgery, we found that the nadir Hb value occurred during the first three postoperative days in 61% of our patients. We also identified that transfusion was delayed in 57% of these cases. Delayed transfusion was almost always related to delayed diagnosis of anaemia and/or delayed prescription of blood transfusion rather than delayed delivery of blood from the blood bank. Delayed transfusion was not better avoided in high-risk cardiac patients. These results suggest that hospital policies for monitoring and caring for the anaemic postoperative patients are far from adequate. They are also of concern because the significant delay recorded might be associated
with an increase in the incidence of cardiac complications in highrisk patients. The time needed to obtain blood from the blood bank and for transfusion to occur was not of concern in this study, as only one hour was needed to deliver RBCs to the ward and 48 minutes to transfuse them. The local organization of the transfusion service may however be a critical point, particularly in smaller hospitals where blood supply may be an issue. Time intervals recorded were in accordance with French guidelines [8] for non-urgent situations; RBCs were easy to obtain as the hospital in which the study was performed is a tertiary level facility in which the blood bank is situated in the same building as the surgical wards. Moreover, major emergency situations are encountered on a daily basis,
Fig. 3. Intervals with respect to the RCRI score. These graphs respectively represent the following time intervals: (transfusion threshold – transfusion) and (Hb nadir – transfusion). Patients are categorised according to the RCRI score and numbers in the boxes correspond to the number of patients.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 4
C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
explaining why teams are well trained with a good knowledge of blood transfusion procedures. We observed a median of 26 hours between the moment at which the threshold was reached and the time at which transfusion was effectively performed on one side (threshold – transfusion) and a median of 7 hours between nadir and transfusion on the other side (nadir – transfusion interval). In the subgroup of patients with delayed transfusion, these median times were respectively 31 and 10 hours. In general, two main reasons accounted for delayed transfusion. The first problem clearly appears to be the diagnosis of anaemia, which was delayed by poor surgical ward organization. There are several factors which may have been responsible for this dysfunctional situation in this hospital at the time the study was performed (2005). First, blood samples prescribed on a nonemergency basis (usually on the day before or earlier) were taken in the early morning by nurses, gathered and transported together to the laboratory. This single step can take time due to the number of patients, the number of nurses available, the distance between wards and the laboratory and the shortage of personnel dedicated to transportation of blood samples. The time for blood analysis was usually short because of automated processes. Results were then validated by the laboratory technician, edited on paper sheets and then returned to the wards. There was no live display on the internal network in our institution. However, electronic information systems reduce only the time taken for transportation of documents to the wards and do not significantly modify the other steps. Results were usually available at noon. Surgeons and anaesthetists spend most of their time in the operating theatre and look at blood results at the end of their operating session, i.e. generally not earlier than 4–6 p.m. Prescription of blood products was therefore made at this time. It should be noted that when very abnormal results are observed, the laboratory technician calls the ward nurse who may alert a physician. A Hb value at 8 g/dL does not however generally lead to a specific call. An additional causal factor is that anaemia was diagnosed from blood samples that are mainly taken to monitor the risk of thrombocytopenia. As many patients are treated with low molecular weight heparin (LMWH) following orthopaedic surgery, the goal of evaluation (i.e. anaemia or thrombocytopenia) is often unclear. Anaemia occurs much earlier than the occurrence of thrombocytopenia. It can occur at any time during the first 3–4 postoperative days and it is well known that total blood loss after orthopaedic surgery is greater than that observed intra-operatively [13]. In their analysis of 4387 patients after non-cardiac surgery, Beattie et al. [14] found that the Hb nadir occurred within the first three postoperative days in 82% of patients. In a recent study including only orthopaedic patients, Irisson et al. also found that the Hb nadir was reached before the 3rd postoperative day [15]. Relying on platelet monitoring intervals, as seen here, can therefore lead to significant delays. This is especially important in an era in which LMWH is being replaced by drugs not associated with a risk for thrombocytopenia. The perceived need to monitor haematological variables may diminish and further increase the risk of delayed diagnosis of anaemia. Daily monitoring of Hb values would certainly shorten the delay in diagnosis of anaemia. To address this need, blood samples could be sent to the central laboratory using pneumatic tubes, which have been shown to decrease laboratory turn-around time [16]. An even simpler solution may be to use point-of-care devices, which are specifically dedicated to Hb measurement rather than sending samples to the central hospital laboratory. Moreover, because point-of-care devices provide results within minutes, the team could respond more rapidly to a newly discovered anaemia. Daily Hb measurements with a point-of-care device could also reduce the lag time before anaemia is diagnosed. Having a dedicated physician on the ward would probably shorten delays. The most logical strategy would be to have enough
specialists (surgeons and anaesthetists) to care for these complex and demanding patients. In a situation of shortage or economic constraints, different models could be envisioned. In the US model, hospitalists have become a widely used model of care and may play a role in improving quality of care [17]. Another way to deal with this problem is to allow nurses to prescribe blood transfusion. This solution is of interest but supporting data for implementation are missing [18]. Such a practice is currently not allowed in France but is accepted in the United Kingdom in certain situations (i.e. nurse practitioners in intensive care). An additional solution could be that nurses call the anaesthetist when Hb < 8 g/dL, for immediate blood prescription. The problem of identifying anaemic patients has been broached in a recent editorial [19]. It is probable that the future of transfusion will have to include individual transfusion thresholds on top of a patient-specific monitoring, thus allowing a tailored transfusion for each single clinical case. In other words, following the presently accepted low transfusion thresholds makes sense only if rapid response transfusion is possible, thus avoiding the occurrence of the side effects associated with profound and long lasting anaemia. To our knowledge, this is the first study to focus on time intervals and delayed transfusion in the post-operative setting. We were able to estimate the time interval at which the threshold was reached and the nadir occurred. The calculation of these intervals helps improve our understanding of the medical and practical reasons leading to profound anaemia; additionally we demonstrated that the main problem was diagnosis of the onset of anaemia. By understanding the various steps associated with delays, changes in hospital management can be suggested. A repeat audit would be necessary to show if real improvement had occurred as a consequence of the changes in the organization. Our study is subject to several limitations. First of all, the number of medical records reviewed was limited to 74. However, this apparently small sample represents all the severely anaemic patients who had undergone orthopaedic surgery during the period under review. Moreover, by choosing a threshold at 8 g/dL, according to French guidelines on blood transfusion [8], the number of anaemic patients was significantly fewer than if using a threshold based on the WHO definition of anaemia: 12 g/dL for women and 13 g/dL for men [20]. Secondly, our study was retrospective and limited to a single, albeit large, centre. Although our data will need to be confirmed in other settings, similar results have already been found in another French institution [10]. Thirdly, we were unable to measure the consequences of delayed transfusion. Measuring the incidence of patients with severe anaemia can be viewed as a surrogate marker of cardiovascular complications only. Numerous data exist however in the medical literature showing the clinical impact of post-operative anaemia on patient outcomes [4,14,21]. Disappointingly, Figs. 3A and B show that the risk of delayed transfusion is independent of cardiac risk. That means that the Hb levels of high-risk patients were not followed more carefully than those of low-risk patients. Even if we were unable to assess the clinical impact of those delayed transfusions, this is a concern due to the epidemiology of the population. We believe that the surrogate end-point that we used in this survey is a good start to improve hospital policy. Because the number of clinical events is relatively low [22], this criterion (i.e. specific follow up of high risk patients) could be used in quality improvement programmes and represent a useful surrogate marker of cardiac complications after surgery. Fourthly, the determination of a transfusion threshold remains a ‘‘hot topic’’. Although targeting 10 g/dL in patients with an acute episode of coronary artery disease is widely accepted [23], the question as to threshold levels in other categories of patients remains less clear. In the present study, we chose to follow the
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
French guidelines [8] in which a threshold Hb value of 8 g/dL for the average patient is recommended. Moreover, patient files were reviewed by a pair of experienced physicians and individual situations were analysed carefully. Fifthly, the time at which the ‘‘threshold’’ was reached was calculated and inferred from the time profile of Hb concentrations for each patient. We hypothesized that a simple linear decline would describe the change in Hb concentrations over time (see Fig. 1); this might be too simplistic and not reflect the individual situation of each patient. This choice is however reinforced by data from Irisson et al. [15], which suggest a rather linear decrease in Hb concentration during the first three days. Finally, we arbitrarily defined delayed transfusion as a time interval (threshold – transfusion) > 12 hours and acceptance of such a time interval could be viewed as a lax definition for an acute care setting. By modifying this threshold, the incidence can be increased significantly. Nevertheless, even with this non-stringent definition, the number of cases associated with delayed transfusion was already high and could in itself be a starting point for improvement. 5. Conclusion Identification of the time at which the transfusion threshold is reached after orthopaedic surgery is shown to be critical. In many patients from this unit, Hb remained under the desirable level for a too long period of time. Solutions to improve these delays are suggested and include, in particular, the daily use of point-of-care devices measuring haemoglobin concentration at the bedside. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgments This work was supported by the Department of Anaesthesiology and Intensive Care, University Hospital of Biceˆtre, Le KremlinBiceˆtre, France. The authors would like to thank Dr Amanda Bull for carefully reviewing the article and making changes to improve the English language. References [1] Lienhart A, Auroy Y, Pequignot F, Benhamou D, Warszawski J, Bovet M, et al. Survey of anesthesia-related mortality in France. Anesthesiology 2006;105: 1087–97. [2] Auroy Y, Lienhart A, Pequignot F, Benhamou D. Complications related to blood transfusion in surgical patients: data from the French national survey on anesthesia-related deaths. Transfusion 2007;47:184S–99S. discussion 201S.
5
[3] Wu WC, Schifftner TL, Henderson WG, Eaton CB, Poses RM, Uttley G, et al. Preoperative hematocrit levels and postoperative outcomes in older patients undergoing noncardiac surgery. JAMA 2007;297:2481–8. [4] Spahn DR. Anemia and patient blood management in hip and knee surgery: a systematic review of the literature. Anesthesiology 2010;113:482–95. [5] Beattie WS, Karkouti K, Wijeysundera DN, Tait G. Risk associated with preoperative anemia in noncardiac surgery: a single-center cohort study. Anesthesiology 2009;110:574–81. [6] Mantilla CB, Wass CT, Goodrich KA, Johanns CJ, Kool ML, Zhu X, et al. Risk for perioperative myocardial infarction and mortality in patients undergoing hip or knee arthroplasty: the role of anemia. Transfusion 2011;51: 82–91. [7] Wells AW, Mounter PJ, Chapman CE, Stainsby D, Wallis JP. Where does blood go? Prospective observational study of red cell transfusion in north England. BMJ 2002;325:803. [8] Afssaps. Recommandations. Transfusion de globules rouges homologues: produits, indications, alternatives. Ann Fr Anesth Reanim 2003;22:67–81. [9] American Society of Anesthesiologists Task Force on Perioperative Blood Transfusion and Adjuvant Therapies. Practice guidelines for perioperative blood transfusion and adjuvant therapies: an updated report by the American Society of Anesthesiologists Task Force on Perioperative Blood Transfusion and Adjuvant Therapies. Anesthesiology 2006;105:198–208. [10] Ausset S, Donat N, Foisseau V, Cirodde A, Clavier B, Auroy Y. Risk analysis of postoperative anemia through biological databases querying. Transfus Clin Biol 2008;15:168–73. [11] Afssaps. Modification des recommandations sur la surveillance plaquettaire d’un traitement par He´parine de Bas Poids Mole´culaire; 2011, http://ansm. sante.fr/var/ansm_site/storage/original/application/ 58af9a851799004cfc1317baf34a70c9.pdf. [12] Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999;100:1043–9. [13] Foss NB, Kehlet H. Hidden blood loss after surgery for hip fracture. J Bone Joint Surg 2006;88:1053–9. [14] Beattie WS, Wijeysundera DN, Karkouti K, McCluskey S, Tait G, Mitsakakis N, et al. Acute surgical anemia influences the cardioprotective effects of betablockade: a single-center, propensity-matched cohort study. Anesthesiology 2010;112:25–33. [15] Irisson E, Kerbaul F, Parratte S, He´mon Y, Argenson JN, Rosencher N, et al. Perioperative management based on kinetics of bleeding during total primary arthroplasty. Ann Fr Anesth Reanim 2013;32:170–4. [16] Guss DA, Chan TC, Killeen JP. The impact of a pneumatic tube and computerized physician order management on laboratory turnaround time. Ann Emerg Med 2008;51:181–5. [17] White HL, Glazier RH. Do hospitalist physicians improve the quality of inpatient care delivery? A systematic review of process, efficiency and outcome measures. BMC Med 2011;9:58. [18] Pirie E, Green J. Should nurses prescribe blood components? Nurs Stand 2007;21:35–41. [19] Rosencher N, Ozier Y, Souied F, Lienhart A, Samama CM. How can we explain the gap between randomised studies and ‘real life’ practice in postoperative transfusion triggers? Do we need to change recommended thresholds for transfusion? Eur J Anaesthesiol 2012;29:460–1. [20] WHO/UNICEF/UNU. Iron deficiency anaemia: assessment, prevention and .control. p. 33. A guide for programme managers. World Health Organization; 2001 [21] van Klei WA, Bryson GL, Yang H, Forster AJ. Effect of beta-blocker prescription on the incidence of postoperative myocardial infarction after hip and knee arthroplasty. Anesthesiology 2009;111:717–24. [22] Mantilla CB, Horlocker TT, Schroeder DR, Berry DJ, Brown DL. Frequency of myocardial infarction, pulmonary embolism, deep venous thrombosis, and death following primary hip or knee arthroplasty. Anesthesiology 2002;96: 1140–6. [23] He´bert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409–17.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
ACCPM-48; No. of Pages 5 Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Original Article
Hospital audit of delayed transfusion after orthopaedic surgery§ Cle´ment Dubost a, Sylvain Ausset b, Ce´sar Vincent a, Charles Gozlan a, Yves Auroy c, Loan N’Guyen d, Andre´ Lienhart e, Dan Benhamou a,*,1 a
De´partement d’anesthe´sie-re´animation, hoˆpital Biceˆtre, AP–HP, 78, rue du Ge´ne´ral-Leclerc, 94275 Le Kremlin-Biceˆtre cedex, France De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es Percy, 101, avenue Henri-Barbusse, BP 406, 92141 Clamart cedex, France c De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es du Val-de-Grace, 74, boulevard de Port-Royal, 75230 Paris, France d Service d’he´movigilance, hoˆpital de la Pitie´-Salpe´trie`re, 47-81, boulevard de l’Hoˆpital, 75651 Paris cedex 13, France e De´partement d’anesthe´sie-re´animation, hoˆpital Saint-Antoine, AP–HP, 184, faubourg Saint-Antoine, 75571 Paris, France b
A R T I C L E I N F O
A B S T R A C T
Article history: Available online xxx
Background and objectives: To understand the mechanisms related to both the onset and correction of severe anaemia after orthopaedic surgery, we analysed all the full blood counts (FBCs) for patients on one orthopaedic ward during a one-year period in an academic hospital. Methods: FBCs were screened and the medical records of those patients for whom a postoperative haemoglobin (Hb) concentration below 8 g/dL was recorded at least once were reviewed. The onset of postoperative anaemia was determined by calculating the various time intervals delineated by surgery, the time at which the transfusion threshold was reached and the time at which the lowest Hb level (nadir) and transfusion (if any) occurred. Results: A total of 6573 FBCs drawn from 1255 patients were screened. The medical records of 74 consecutive patients with at least one Hb value < 8 g/dL were analysed. The postoperative Hb nadir was 7.4 ( 0.6) g/dL (mean – SD). The medians (IQR 25–75) of the calculated intervals were: (surgery – nadir): 72 (48–144) h, (nadir – transfusion): 7 (5–21) h and (transfusion threshold – transfusion): 26 (11– 51) h. Conclusions: Delayed transfusion (defined as > 12 hours between the time at which the transfusion threshold was reached and actual transfusion) was observed in 57% of severely anaemic patients after orthopaedic surgery. ß 2015 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Keywords: Anaemia Postoperative period Transfusion Risk
1. Introduction Postoperative anaemia is a major challenge for patient safety. Indeed, a survey of anaesthesia-related mortality in 2006 in France showed that nearly one quarter of deaths partially or totally related to anaesthesia were the result of inadequate management of postoperative anaemia [1]. A thorough investigation of these deaths revealed that they did not usually occur during surgery but rather during the third or fourth postoperative day, related to a slowly occurring anaemia with major end-organ
§ Preliminary data were presented in part at the annual meeting of the European Society of Anaesthesiology in Amsterdam, June 11–14, 2011. * Corresponding author. Tel.: +33 1 45 21 39 35; fax: +33 1 45 21 28 75. E-mail address: [email protected] (D. Benhamou). 1 De´partement d’anesthe´sie-re´animation, hoˆpital d’instruction des Arme´es Be´gin, 69, avenue de Paris, 94160 Saint-Mande´, France.
dysfunction, such as myocardial infarction [2]. The impact of preoperative anaemia on patient outcomes has been well studied [3–6] but data concerning the postoperative period are more limited. Acute postoperative anaemia has been found in many institutions to be most common following orthopaedic surgery and thus is the main indication for blood transfusion in surgical patients [2,7]. Post-operative anaemia may worsen post-operative outcomes and increase both mortality and morbidity [4]. Although transfusion thresholds have been defined by recommendations in various countries [8,9], risk might also be related to the duration of postoperative anaemia. Therefore, delayed transfusion might also increase risk [10]. In order to establish the timing of severe postoperative anaemia (i.e. which reached the transfusion threshold), we analysed all full blood counts (FBC) drawn from elective and emergency patients on one orthopaedic ward during a one-year period in a major teaching and trauma hospital.
http://dx.doi.org/10.1016/j.accpm.2015.02.006 2352-5568/ß 2015 Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar). Published by Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
2
2. Methods
medians and 25th–75th percentiles or means standard deviation (SD) (according to the observed distribution).
Database analysis was performed in three steps. First, the FBC results obtained from patients in the orthopaedic ward were extracted from the Haematology Laboratory database. Then, data from patients with at least one postoperative haemoglobin (Hb) level < 8 g/dL were extracted and merged with that of the transfusion centre. Finally, the medical records of these patients were thoroughly peer-reviewed by a pair of experienced physicians. Data were recorded in a preformed Excel1 table (Microsoft Excel1 2008 for Mac) after group validation. According to French recommendations for red blood cell (RBC) transfusion, the desirable transfusion threshold has been defined as 8 g/dL in patients without a history of acute coronary artery disease [8]. The timing of postoperative blood checks was decided by the physician in charge, following national and institutional recommendations [11]. For the medical records analysed, the data recorded were: the date and type of surgery, age, sex and American Society of Anesthesiologists (ASA) status, data from the patient’s medical history pertinent to anaemia and transfusion – especially cardiovascular history. The Revised Cardiac Risk Index (RCRI) was determined for each patient [12]. The occurrence of any perioperative event that might be linked to anaemia (cause or consequence) was searched for by carefully reading medical and nursing files. The lowest postoperative Hb level (nadir) and the time at which it occurred were recorded as well as the amount and timing of intra- or postoperative transfusion. The patient files, containing all the results of biological tests, were retrospectively analysed. Only the validated, paper versions of results were analysed. The time at which blood transfusion was prescribed, the time of delivery of RBCs by the blood bank and the time of transfusion onset were also recorded. For each patient, a graphical representation was used to estimate the time at which the Hb level reached the transfusion threshold (Fig. 1). When no blood result was available on the day of threshold crossing, intervals were estimated based on the hypothesis of a linear decrease in Hb. From the above data, the following time intervals were calculated: (surgery – threshold), (threshold – nadir), (nadir – transfusion [if any]), (nadir – prescription), (prescription – delivery) and (delivery – transfusion). Transfusion was considered to have been delayed when more than 12 hours had elapsed between reaching the transfusion threshold and the actual transfusion performance. Statistical analyses were performed using Prism1 4.00 for Mac (Inc GraphPad Software, San Diego California USA, http://www. graphpad.com). Data are presented as percentages or using
A total of 6573 FBCs drawn from 1255 patients from a single centre were screened (Fig. 2). Anaemia with at least one Hb value lower than 10 g/dL was found in 456 patients (36%), while 162 (13%) exhibited at least one Hb value < 8 g/dL, among whom 88 were excluded for the following reasons: no surgery (n = 33), admission to the ICU (n = 20) and missing data (n = 25). The medical records of the remaining 74 patients were analysed. Patient characteristics are presented in Table 1. In 14 patients (19%), an intra-operative event related to haemorrhage (such as a drop in blood pressure or high blood volume collected in the suction drains) was observed. Nineteen patients (26%) received RBCs during surgery. A Hb measurement was performed during surgery or in the Post-Anaesthesia Care Unit (PACU) in 50 patients (68%). The median intraoperative Hb was 9.2 g/dL. Fifty-five patients did not receive RBCs during surgery (74%), 17 patients (23%) received 1–4 RBCs and 2 received more than 4 RBCs (3%). The mean postoperative Hb nadir was 7.4 (SD 0.6) g/dL. In 15 patients (20%), a postoperative event that could be related to anaemia (haematoma [6], delirium [4], chest pain [1], and low blood pressure [4]) was recorded. Among the 74 patients, 64 received a blood transfusion at least once during the postoperative period. The median number of RBCs transfused during the postoperative period was 2 (IQR 2–2). Reasons for the absence of transfusion included patient refusal, good clinical tolerance and lowering of the transfusion threshold by the physician in charge. No adverse effect was identified in the patients not receiving transfusion. Among them, the mean value of nadir Hb was 7.9 g/dL (SD 1). Delayed transfusion was observed in 42 (57%) patients. Time intervals were calculated in 64 patients. The median value (IQR) of the calculated intervals in hours is presented in Table 2. The median time between crossing the transfusion threshold and transfusion was 26 hours (11–51). The median time interval between prescription and delivery was 1 hour (0.7–2) and the median interval between delivery and transfusion was 0.8 hours (0.6–1). The time intervals (transfusion threshold – transfusion) and (nadir – transfusion) are represented in Figs. 3A and B with respect to the RCRI score. Shorter intervals were not consistently observed in patients with a high RCRI score. Thirty-two patients had an RCRI score greater than 2. Among them, postoperative complications related to anaemia were identified: haematoma (4), delirium (2) and low blood pressure (1).
Fig. 1. Graphical determination of the nadir. Graphical representation of Hb concentration from postoperative day 1 to the time of Hb nadir. This graph was used to estimate the time at which Hb transfusion threshold was reached.
Fig. 2. Flow chart of the study.
3. Results
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx Table 1 Demographic characteristics of the population studied. Age (years) ASA 1–2 ASA 3–4 Sex ratio M/F Revised Cardiac Risk Index: number of predictive factors One Two Three/Four Emergency surgery Type of surgery Hip Femur Knee Spine Others
3
Table 2 Relevant time intervals. 72,5 (44–82.5) 51 (69%) 23 (31%) 0.85 49 (66%) 21 (28%) 4 (6%) 49 (66%) 27 (37%) 9 (12%) 10 (14%) 7 (9%) 21 (28%)
Time intervals (h)
All patients (n = 64)
Transfusion > 12 h (n = 42)
Transfusion < 12 h (n = 22)
Surgery – nadir Surgery – transfusion Nadir – prescription Nadir – delivery Nadir – transfusion Threshold reached – prescription Threshold reached – delivery Threshold reached – transfusion Prescription – delivery Delivery – transfusion
72 72 5 6 7 23
72 72 6 8 10 29
72 72 3 5 6 3
(48–144) (48–120) (3–12) (4–19) (5–21) (8–48)
(48–132) (48–108) (4–26) (6–27) (7–27) (21–54)
(36–156) (36–144) (2–4) (3–7) (4–7) (2–6)
24 (10–50)
30 (23–56)
5 (4–7)
26 (11–51)
31 (23–56)
6 (4–8)
1 (0.7–2) 0.8 (0.6–1)
1 (0.7–2) 0.8 (0.6–1)
1 (0.4–2) 0.6 (0.4–0,9)
4. Discussion By analysing the medical records of 74 severely anaemic patients (i.e. Hb < 8 g/dL) after orthopaedic surgery, we found that the nadir Hb value occurred during the first three postoperative days in 61% of our patients. We also identified that transfusion was delayed in 57% of these cases. Delayed transfusion was almost always related to delayed diagnosis of anaemia and/or delayed prescription of blood transfusion rather than delayed delivery of blood from the blood bank. Delayed transfusion was not better avoided in high-risk cardiac patients. These results suggest that hospital policies for monitoring and caring for the anaemic postoperative patients are far from adequate. They are also of concern because the significant delay recorded might be associated
with an increase in the incidence of cardiac complications in highrisk patients. The time needed to obtain blood from the blood bank and for transfusion to occur was not of concern in this study, as only one hour was needed to deliver RBCs to the ward and 48 minutes to transfuse them. The local organization of the transfusion service may however be a critical point, particularly in smaller hospitals where blood supply may be an issue. Time intervals recorded were in accordance with French guidelines [8] for non-urgent situations; RBCs were easy to obtain as the hospital in which the study was performed is a tertiary level facility in which the blood bank is situated in the same building as the surgical wards. Moreover, major emergency situations are encountered on a daily basis,
Fig. 3. Intervals with respect to the RCRI score. These graphs respectively represent the following time intervals: (transfusion threshold – transfusion) and (Hb nadir – transfusion). Patients are categorised according to the RCRI score and numbers in the boxes correspond to the number of patients.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 4
C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
explaining why teams are well trained with a good knowledge of blood transfusion procedures. We observed a median of 26 hours between the moment at which the threshold was reached and the time at which transfusion was effectively performed on one side (threshold – transfusion) and a median of 7 hours between nadir and transfusion on the other side (nadir – transfusion interval). In the subgroup of patients with delayed transfusion, these median times were respectively 31 and 10 hours. In general, two main reasons accounted for delayed transfusion. The first problem clearly appears to be the diagnosis of anaemia, which was delayed by poor surgical ward organization. There are several factors which may have been responsible for this dysfunctional situation in this hospital at the time the study was performed (2005). First, blood samples prescribed on a nonemergency basis (usually on the day before or earlier) were taken in the early morning by nurses, gathered and transported together to the laboratory. This single step can take time due to the number of patients, the number of nurses available, the distance between wards and the laboratory and the shortage of personnel dedicated to transportation of blood samples. The time for blood analysis was usually short because of automated processes. Results were then validated by the laboratory technician, edited on paper sheets and then returned to the wards. There was no live display on the internal network in our institution. However, electronic information systems reduce only the time taken for transportation of documents to the wards and do not significantly modify the other steps. Results were usually available at noon. Surgeons and anaesthetists spend most of their time in the operating theatre and look at blood results at the end of their operating session, i.e. generally not earlier than 4–6 p.m. Prescription of blood products was therefore made at this time. It should be noted that when very abnormal results are observed, the laboratory technician calls the ward nurse who may alert a physician. A Hb value at 8 g/dL does not however generally lead to a specific call. An additional causal factor is that anaemia was diagnosed from blood samples that are mainly taken to monitor the risk of thrombocytopenia. As many patients are treated with low molecular weight heparin (LMWH) following orthopaedic surgery, the goal of evaluation (i.e. anaemia or thrombocytopenia) is often unclear. Anaemia occurs much earlier than the occurrence of thrombocytopenia. It can occur at any time during the first 3–4 postoperative days and it is well known that total blood loss after orthopaedic surgery is greater than that observed intra-operatively [13]. In their analysis of 4387 patients after non-cardiac surgery, Beattie et al. [14] found that the Hb nadir occurred within the first three postoperative days in 82% of patients. In a recent study including only orthopaedic patients, Irisson et al. also found that the Hb nadir was reached before the 3rd postoperative day [15]. Relying on platelet monitoring intervals, as seen here, can therefore lead to significant delays. This is especially important in an era in which LMWH is being replaced by drugs not associated with a risk for thrombocytopenia. The perceived need to monitor haematological variables may diminish and further increase the risk of delayed diagnosis of anaemia. Daily monitoring of Hb values would certainly shorten the delay in diagnosis of anaemia. To address this need, blood samples could be sent to the central laboratory using pneumatic tubes, which have been shown to decrease laboratory turn-around time [16]. An even simpler solution may be to use point-of-care devices, which are specifically dedicated to Hb measurement rather than sending samples to the central hospital laboratory. Moreover, because point-of-care devices provide results within minutes, the team could respond more rapidly to a newly discovered anaemia. Daily Hb measurements with a point-of-care device could also reduce the lag time before anaemia is diagnosed. Having a dedicated physician on the ward would probably shorten delays. The most logical strategy would be to have enough
specialists (surgeons and anaesthetists) to care for these complex and demanding patients. In a situation of shortage or economic constraints, different models could be envisioned. In the US model, hospitalists have become a widely used model of care and may play a role in improving quality of care [17]. Another way to deal with this problem is to allow nurses to prescribe blood transfusion. This solution is of interest but supporting data for implementation are missing [18]. Such a practice is currently not allowed in France but is accepted in the United Kingdom in certain situations (i.e. nurse practitioners in intensive care). An additional solution could be that nurses call the anaesthetist when Hb < 8 g/dL, for immediate blood prescription. The problem of identifying anaemic patients has been broached in a recent editorial [19]. It is probable that the future of transfusion will have to include individual transfusion thresholds on top of a patient-specific monitoring, thus allowing a tailored transfusion for each single clinical case. In other words, following the presently accepted low transfusion thresholds makes sense only if rapid response transfusion is possible, thus avoiding the occurrence of the side effects associated with profound and long lasting anaemia. To our knowledge, this is the first study to focus on time intervals and delayed transfusion in the post-operative setting. We were able to estimate the time interval at which the threshold was reached and the nadir occurred. The calculation of these intervals helps improve our understanding of the medical and practical reasons leading to profound anaemia; additionally we demonstrated that the main problem was diagnosis of the onset of anaemia. By understanding the various steps associated with delays, changes in hospital management can be suggested. A repeat audit would be necessary to show if real improvement had occurred as a consequence of the changes in the organization. Our study is subject to several limitations. First of all, the number of medical records reviewed was limited to 74. However, this apparently small sample represents all the severely anaemic patients who had undergone orthopaedic surgery during the period under review. Moreover, by choosing a threshold at 8 g/dL, according to French guidelines on blood transfusion [8], the number of anaemic patients was significantly fewer than if using a threshold based on the WHO definition of anaemia: 12 g/dL for women and 13 g/dL for men [20]. Secondly, our study was retrospective and limited to a single, albeit large, centre. Although our data will need to be confirmed in other settings, similar results have already been found in another French institution [10]. Thirdly, we were unable to measure the consequences of delayed transfusion. Measuring the incidence of patients with severe anaemia can be viewed as a surrogate marker of cardiovascular complications only. Numerous data exist however in the medical literature showing the clinical impact of post-operative anaemia on patient outcomes [4,14,21]. Disappointingly, Figs. 3A and B show that the risk of delayed transfusion is independent of cardiac risk. That means that the Hb levels of high-risk patients were not followed more carefully than those of low-risk patients. Even if we were unable to assess the clinical impact of those delayed transfusions, this is a concern due to the epidemiology of the population. We believe that the surrogate end-point that we used in this survey is a good start to improve hospital policy. Because the number of clinical events is relatively low [22], this criterion (i.e. specific follow up of high risk patients) could be used in quality improvement programmes and represent a useful surrogate marker of cardiac complications after surgery. Fourthly, the determination of a transfusion threshold remains a ‘‘hot topic’’. Although targeting 10 g/dL in patients with an acute episode of coronary artery disease is widely accepted [23], the question as to threshold levels in other categories of patients remains less clear. In the present study, we chose to follow the
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
G Model
ACCPM-48; No. of Pages 5 C. Dubost et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
French guidelines [8] in which a threshold Hb value of 8 g/dL for the average patient is recommended. Moreover, patient files were reviewed by a pair of experienced physicians and individual situations were analysed carefully. Fifthly, the time at which the ‘‘threshold’’ was reached was calculated and inferred from the time profile of Hb concentrations for each patient. We hypothesized that a simple linear decline would describe the change in Hb concentrations over time (see Fig. 1); this might be too simplistic and not reflect the individual situation of each patient. This choice is however reinforced by data from Irisson et al. [15], which suggest a rather linear decrease in Hb concentration during the first three days. Finally, we arbitrarily defined delayed transfusion as a time interval (threshold – transfusion) > 12 hours and acceptance of such a time interval could be viewed as a lax definition for an acute care setting. By modifying this threshold, the incidence can be increased significantly. Nevertheless, even with this non-stringent definition, the number of cases associated with delayed transfusion was already high and could in itself be a starting point for improvement. 5. Conclusion Identification of the time at which the transfusion threshold is reached after orthopaedic surgery is shown to be critical. In many patients from this unit, Hb remained under the desirable level for a too long period of time. Solutions to improve these delays are suggested and include, in particular, the daily use of point-of-care devices measuring haemoglobin concentration at the bedside. Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgments This work was supported by the Department of Anaesthesiology and Intensive Care, University Hospital of Biceˆtre, Le KremlinBiceˆtre, France. The authors would like to thank Dr Amanda Bull for carefully reviewing the article and making changes to improve the English language. References [1] Lienhart A, Auroy Y, Pequignot F, Benhamou D, Warszawski J, Bovet M, et al. Survey of anesthesia-related mortality in France. Anesthesiology 2006;105: 1087–97. [2] Auroy Y, Lienhart A, Pequignot F, Benhamou D. Complications related to blood transfusion in surgical patients: data from the French national survey on anesthesia-related deaths. Transfusion 2007;47:184S–99S. discussion 201S.
5
[3] Wu WC, Schifftner TL, Henderson WG, Eaton CB, Poses RM, Uttley G, et al. Preoperative hematocrit levels and postoperative outcomes in older patients undergoing noncardiac surgery. JAMA 2007;297:2481–8. [4] Spahn DR. Anemia and patient blood management in hip and knee surgery: a systematic review of the literature. Anesthesiology 2010;113:482–95. [5] Beattie WS, Karkouti K, Wijeysundera DN, Tait G. Risk associated with preoperative anemia in noncardiac surgery: a single-center cohort study. Anesthesiology 2009;110:574–81. [6] Mantilla CB, Wass CT, Goodrich KA, Johanns CJ, Kool ML, Zhu X, et al. Risk for perioperative myocardial infarction and mortality in patients undergoing hip or knee arthroplasty: the role of anemia. Transfusion 2011;51: 82–91. [7] Wells AW, Mounter PJ, Chapman CE, Stainsby D, Wallis JP. Where does blood go? Prospective observational study of red cell transfusion in north England. BMJ 2002;325:803. [8] Afssaps. Recommandations. Transfusion de globules rouges homologues: produits, indications, alternatives. Ann Fr Anesth Reanim 2003;22:67–81. [9] American Society of Anesthesiologists Task Force on Perioperative Blood Transfusion and Adjuvant Therapies. Practice guidelines for perioperative blood transfusion and adjuvant therapies: an updated report by the American Society of Anesthesiologists Task Force on Perioperative Blood Transfusion and Adjuvant Therapies. Anesthesiology 2006;105:198–208. [10] Ausset S, Donat N, Foisseau V, Cirodde A, Clavier B, Auroy Y. Risk analysis of postoperative anemia through biological databases querying. Transfus Clin Biol 2008;15:168–73. [11] Afssaps. Modification des recommandations sur la surveillance plaquettaire d’un traitement par He´parine de Bas Poids Mole´culaire; 2011, http://ansm. sante.fr/var/ansm_site/storage/original/application/ 58af9a851799004cfc1317baf34a70c9.pdf. [12] Lee TH, Marcantonio ER, Mangione CM, et al. Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery. Circulation 1999;100:1043–9. [13] Foss NB, Kehlet H. Hidden blood loss after surgery for hip fracture. J Bone Joint Surg 2006;88:1053–9. [14] Beattie WS, Wijeysundera DN, Karkouti K, McCluskey S, Tait G, Mitsakakis N, et al. Acute surgical anemia influences the cardioprotective effects of betablockade: a single-center, propensity-matched cohort study. Anesthesiology 2010;112:25–33. [15] Irisson E, Kerbaul F, Parratte S, He´mon Y, Argenson JN, Rosencher N, et al. Perioperative management based on kinetics of bleeding during total primary arthroplasty. Ann Fr Anesth Reanim 2013;32:170–4. [16] Guss DA, Chan TC, Killeen JP. The impact of a pneumatic tube and computerized physician order management on laboratory turnaround time. Ann Emerg Med 2008;51:181–5. [17] White HL, Glazier RH. Do hospitalist physicians improve the quality of inpatient care delivery? A systematic review of process, efficiency and outcome measures. BMC Med 2011;9:58. [18] Pirie E, Green J. Should nurses prescribe blood components? Nurs Stand 2007;21:35–41. [19] Rosencher N, Ozier Y, Souied F, Lienhart A, Samama CM. How can we explain the gap between randomised studies and ‘real life’ practice in postoperative transfusion triggers? Do we need to change recommended thresholds for transfusion? Eur J Anaesthesiol 2012;29:460–1. [20] WHO/UNICEF/UNU. Iron deficiency anaemia: assessment, prevention and .control. p. 33. A guide for programme managers. World Health Organization; 2001 [21] van Klei WA, Bryson GL, Yang H, Forster AJ. Effect of beta-blocker prescription on the incidence of postoperative myocardial infarction after hip and knee arthroplasty. Anesthesiology 2009;111:717–24. [22] Mantilla CB, Horlocker TT, Schroeder DR, Berry DJ, Brown DL. Frequency of myocardial infarction, pulmonary embolism, deep venous thrombosis, and death following primary hip or knee arthroplasty. Anesthesiology 2002;96: 1140–6. [23] He´bert PC, Wells G, Blajchman MA, Marshall J, Martin C, Pagliarello G, et al. A multicenter, randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409–17.
Please cite this article in press as: Dubost C, et al. Hospital audit of delayed transfusion after orthopaedic surgery. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2015.02.006
