Cell salvage for minimising perioperative allogeneic blood transfusion

Paul A Carless; David A Henry; Annette J Moxey; Dianne O'Connell; Tamara Brown; Dean A Fergusson

doi:10.1002/14651858.CD001888.pub4

. 2010 Apr 14;2010(4):CD001888. doi: 10.1002/14651858.CD001888.pub4

Cell salvage for minimising perioperative allogeneic blood transfusion

Paul A Carless ^1,^✉, David A Henry ², Annette J Moxey ³, Dianne O'Connell ⁴, Tamara Brown ⁵, Dean A Fergusson ⁶

Editor: Cochrane Injuries Group

PMCID: PMC4163967 EMSID: EMS57130 PMID: 20393932

Abstract

Background

Concerns regarding the safety of transfused blood have prompted reconsideration of the use of allogeneic (from an unrelated donor) red blood cell (RBC) transfusion, and a range of techniques to minimise transfusion requirements.

Objectives

To examine the evidence for the efficacy of cell salvage in reducing allogeneic blood transfusion and the evidence for any effect on clinical outcomes.

Search methods

We identified studies by searching CENTRAL (The Cochrane Library 2009, Issue 2), MEDLINE (1950 to June 2009), EMBASE (1980 to June 2009), the internet (to August 2009) and bibliographies of published articles.

Selection criteria

Randomised controlled trials with a concurrent control group in which adult patients, scheduled for non‐urgent surgery, were randomised to cell salvage (autotransfusion) or to a control group who did not receive the intervention.

Data collection and analysis

Data were independently extracted and the risk of bias assessed. Relative risks (RR) and weighted mean differences (WMD) with 95% confidence intervals (CIs) were calculated. Data were pooled using a random‐effects model. The primary outcomes were the number of patients exposed to allogeneic red cell transfusion and the amount of blood transfused. Other clinical outcomes are detailed in the review.

Main results

A total of 75 trials were included. Overall, the use of cell salvage reduced the rate of exposure to allogeneic RBC transfusion by a relative 38% (RR 0.62; 95% CI 0.55 to 0.70). The absolute reduction in risk (ARR) of receiving an allogeneic RBC transfusion was 21% (95% CI 15% to 26%). In orthopaedic procedures the RR of exposure to RBC transfusion was 0.46 (95% CI 0.37 to 0.57) compared to 0.77 (95% CI 0.69 to 0.86) for cardiac procedures. The use of cell salvage resulted in an average saving of 0.68 units of allogeneic RBC per patient (WMD ‐0.68; 95% CI ‐0.88 to ‐0.49). Cell salvage did not appear to impact adversely on clinical outcomes.

Authors' conclusions

The results suggest cell salvage is efficacious in reducing the need for allogeneic red cell transfusion in adult elective cardiac and orthopaedic surgery. The use of cell salvage did not appear to impact adversely on clinical outcomes. However, the methodological quality of trials was poor. As the trials were unblinded and lacked adequate concealment of treatment allocation, transfusion practices may have been influenced by knowledge of the patients' treatment status potentially biasing the results in favour of cell salvage.

Plain language summary

Cell salvage (collecting a patient's own blood during surgery) for reducing transfusions with donated blood

Some patients who undergo surgery require blood transfusions to compensate for the blood loss that occurs during the procedure. Often the blood used for the transfusion has been donated by a volunteer. The risks associated with receiving volunteer donor blood that has been screened by a competently managed modern laboratory are considered minimal, with the risk of contracting diseases such as HIV and hepatitis C being extremely low. However there is concern in many developing countries, where there is a high prevalence of such infections and transfusion services are inadequately equipped to screen donor blood as thoroughly. Although in developed countries the risks of acquiring a disease from transfused blood are low, the financial costs associated with providing a safe and reliable blood product are escalating. Therefore there is much attention being placed on alternative strategies to minimise the need for transfusions of donor blood.

'Cell salvage' or 'autotransfusion' is one technique designed to reduce the use of such transfusions. It involves the collection of a patient's own blood from surgical sites which can be transfused back into the same person during or after surgery, as required.

The authors undertook this systematic review to examine the evidence for the effectiveness of cell salvage in reducing the need for blood transfusions of donor blood in adults (over 18 years) undergoing surgery.

The authors found 75 studies investigating the effectiveness of cell salvage in orthopaedic (36 studies), cardiac (33 studies), and vascular (6 studies) surgery. Overall, the findings show that cell salvage reduces the need for transfusions of donated blood. The authors conclude that there appears to be sufficient evidence to support the use of cell salvage in cardiac and orthopaedic surgery. Cell salvage does not appear to cause any adverse clinical outcomes.

As the methodological quality of the trials was poor, the findings may be biased in favour of cell salvage. Large trials of high methodological quality that assess the relative effectiveness, safety, and cost‐effectiveness of cell salvage in different surgical procedures should be the focus of future research in this area.

Background

Concerns regarding the safety of transfused blood, have prompted reconsideration of the use of allogeneic (blood from an unrelated donor) red cell transfusion. The risks associated with volunteer donor blood transfusion (allogeneic blood) that has been screened by a competently managed modern laboratory are generally considered minimal, with the risks of human immunodeficiency virus (HIV) and hepatitis C (HCV) being extremely low (Glynn 2000; Whyte 1997). However, of great concern is that in many developing countries there is a high prevalence of such infections and transfusion services are inadequately equipped to conduct universal antibody screening (Lackritz 1998; McFarland 1997). Meanwhile, in developed countries, although the risks of acquiring a transfusion‐transmitted disease (TTD) are considered low, the costs associated with providing a safe and reliable blood product are escalating (Hadjianastassiou 2002).

Recent concerns that variant Creutzfeldt‐Jakob disease (vCJD) could be transmitted by blood transfusion (Brown 2000; Houston 2000) have prompted blood transfusion services worldwide to adopt more stringent donor selection procedures and the deferral of current donors who may have been exposed to vCJD (Oliver 2002). The ramifications of such actions have been, in some cases, the elimination of a sizable proportion of blood donors from an already declining volunteer donor pool. Blood is now, more than ever, an incredibly scarce resource.

Concerns regarding blood safety, continual blood shortages, and spiraling health costs associated with blood bank operations, have collectively generated considerable enthusiasm for the use of technologies intended to reduce the use of allogeneic blood (Forgie 1998). However, some of the alternatives to allogeneic blood have their own risks and are highly expensive (Coyle 1999; Fergusson 1999a).

With early reports suggesting that between 60% and 70% of all red blood cell units are transfused in the surgical setting (Cook 1991; Lenfant 1992; Surgenor 1990; Wallace 1993) and more recently, that half of all the blood transfused in the United Kingdom is to surgical patients (Regan 2002), it is of no surprise that considerable interest has been shown in a range of interventions designed to reduce perioperative allogeneic red cell transfusion. Generally, such interventions fall into three groups: (1) the administration of agents to diminish blood loss (e.g. aprotinin, tranexamic acid, epsilon aminocaproic acid, fibrin sealant), (2) agents that promote red blood cell production (e.g. erythropoietin), and (3) techniques for re‐infusing a patient's own blood (e.g. pre‐operative autologous donation, acute normovolaemic haemodilution, cell salvage).

Cell salvage is one technique that has been used extensively in the surgical setting. Cell salvage (CS), alternatively known as 'auto‐transfusion', is a term that covers a range of techniques that scavenge blood from operative fields or wound sites, and re‐infuse the blood back into the patient. Cell salvage can be performed during the intra‐ and/or postoperative periods. To remove non‐cellular matter prior to reinfusion, some of the devices use centrifugal washing of the salvaged blood (Huet 1999).

This review builds on the systematic review published by Huet et al (Huet 1999). It examines the evidence for the efficacy of cell salvage in reducing the need for perioperative allogeneic red blood cell transfusion in adult elective surgery and whether there is a greater reduction in allogeneic blood transfusion demonstrated in identifiable patient sub‐groups. This review employs methods developed by the International Study of Perioperative Transfusion (ISPOT) study group (a ten country study of evidence, attitudes and practices relating to the use of alternatives to allogeneic blood transfusion) (Fergusson 1999b).

Objectives

To examine the effects of cell salvage in minimising perioperative allogeneic red blood cell transfusion and on other clinical outcomes.

Methods

Criteria for considering studies for this review

Types of studies

Randomised controlled trials with a concurrent control group.

Types of participants

The study participants were adults (over 18 years). The surgery being conducted was elective or non‐urgent.

Types of interventions

The intervention considered was cell salvage (CS). Studies with a combination of active comparisons were included if both the intervention and control groups were equally exposed to the active treatment (active plus cell salvage versus active comparisons).

Types of outcome measures

Primary outcomes

The number of patients who were transfused with allogeneic or autologous blood, or both
The amounts of allogeneic and/or autologous blood transfused

Secondary outcomes

Re‐operation for bleeding
Post‐operative complications (thrombosis, infection, renal failure, non‐fatal myocardial infarction)
Mortality
Length of hospital stay (LOS)

Search methods for identification of studies

Electronic searches

This review drew on the literature searches that were constructed as part of the International Study of Perioperative Transfusion (ISPOT) (Huet 1999). The searches were last updated in June 2009.

We searched the following electronic databases:

Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2009, Issue 2);
MEDLINE (1950 to June (Week 3) 2009);
EMBASE (1980 to Week 26, 2009);
Current Contents (ISI Web of Knowledge) (to June 2009).

The searches were based on the MEDLINE strategy shown in Appendix 1, and were modified as appropriate to the specifications of each database.

In MEDLINE and EMBASE two search filters were used to restrict electronic searches and improve the specificity of the searches. Firstly, the ISPOT filter (Laupacis 1997), which identifies blood transfusion trials, and secondly, a modified version of the Cochrane Collaboration filter (Dickersin 1996) which identifies randomised controlled trials. These search filters were coupled with the MeSH headings and relevant text‐word terms for cell salvage.

Searching other resources

The internet site of the International Network of Agencies of Health Technology Assessment (INAHTA) was searched to June 2009. Reference lists of relevant reviews and identified articles were searched for additional studies. Contact was made with experts in the field to identify reports or projects in progress, relevant to the review. In addition, authors were contacted to identify any additional published or unpublished data.

Data collection and analysis

Selection of studies

The titles and/or abstracts of identified studies were screened by two independent authors. To be eligible for inclusion, studies had to include adult patients, scheduled for elective surgery, who were randomised to cell salvage or to a control group that did not receive cell salvage. Study reports had to provide data on the number of patients transfused with red cells, or the volume of blood transfused. Two authors independently selected trials that met the defined inclusion criteria with disagreements resolved by consensus.

Data extraction and management

Two authors independently extracted study characteristics and outcomes using an article extraction form. The extraction form was used to record information regarding; randomisation criteria, trial methodology, the presence of a transfusion protocol, the type of surgery involved, treatment outcomes, and general comments. Information regarding demographics (age, sex), the type of surgery, the presence or absence of a transfusion protocol, the timing of cell salvage, and the type of cell salvage (washed or unwashed) was also recorded. Data were extracted for allogeneic blood transfusion if it were expressed as whole blood or packed red cells. Transfusion data expressed in millilitres were converted to units by dividing by 300.

Data on the following outcomes were recorded:

the number of patients transfused allogeneic and/or autologous blood;
the volume of red cells transfused;
post‐operative complications (thrombosis, infection, haemorrhage, non‐fatal myocardial infarction, renal failure);
mortality;
length of hospital stay (LOS).

Data were also recorded for blood loss and the number of patients requiring re‐operation for bleeding.

Authors were contacted to provide missing data where possible.

Assessment of risk of bias in included studies

Studies were assessed for methodological quality using the Cochrane Collaboration's tool for assessing risk of bias presented in Higgins 2009. Disagreements were resolved by consensus.

The following domains were assessed for each study:

sequence generation,

allocation concealment,

blinding.

We completed a risk of bias table for each study, incorporating a description of the study's performance against each of the above domains and our overall judgement of the risk of bias for each entry as follows; 'Yes' indicates a low risk of bias, 'Unclear' indicates unclear or unknown risk of bias, and 'No' indicates a high risk of bias.

Assessment of reporting biases

Funnel plots were inspected for evidence of publication bias.

Data synthesis

Data were extracted and then entered into Review Manager by one author. Articles identified as duplicate publications were combined to obtain one set of data. Dichotomous and continuous data were pooled across trials using a random effects model. If the standard deviation (SD) or the standard error of the mean (SEM) were not reported for continuous data the study was not included in the meta‐analysis.

Subgroup analysis and investigation of heterogeneity

Statistical heterogeneity was examined by both the I‐squared and chi‐squared tests. The I‐squared test describes the percentage of total variation across studies due to heterogeneity rather than chance. A value of 0% indicates no observed heterogeneity, whereas values >50% indicates substantial heterogeneity (Higgins 2009). The Q statistic, which has an approximate Chi² distribution with degrees of freedom equal to the number of studies minus one, was also used to assess heterogeneity of treatment effect (Der Simonian 1986). A P value less than or equal to 0.1 was used to define statistically significant heterogeneity.

Analysis of a priori subgroups was performed to determine whether effect sizes varied according to factors such as:

the type of surgery;
the use of transfusion protocols;
the type of salvaged blood retransfused (washed or unwashed);
the timing of cell salvage (intra‐ or post‐operative, or both);
trial methodological quality.

Results

Description of studies

Seventy‐five studies fulfilled the inclusion criteria. The majority of trials included in the analysis were conducted in the United Kingdom (n = 23) and the United States (n = 14). The remaining trials were conducted across a range of countries: the Netherlands (n = 6), Sweden (n = 5), China (n = 3), France (n = 3), Germany (n = 3), India (n = 3), Australia (n = 2), Canada (n = 2), Denmark (n = 2), Greece (n = 2), Croatia (n = 1), Hong Kong (n = 1), Italy (n = 1), Lithuania (n = 1), Norway (n = 1), Russia (n = 1) and Turkey (n = 1). Studies were published between 1978 and 2008. Five studies (Fragnito 1995; Lorentz 1991; Menges 1992; Rosencher 1994; Zhang 2008) were published in languages other than English. These studies were translated before being included in the analysis. The majority of trials were small with less than 100 patients in each arm of the trial. Only four trials included more than 100 patients in each trial arm (Gannon 1991; Klein 2008; McGill 2002; Ritter 1994). Of the 75 trials included in the analysis, 36 trials involved patients undergoing orthopaedic procedures, 33 involved patients undergoing cardiac procedures, and six involved vascular surgery.

Methods of cell savage

Of the 75 included trials, 48 studied cell salvage during the post‐operative period, 16 studied intra‐operative cell salvage, and 11 studied both intra‐operative and post‐operative cell salvage. One trial (Sait 1999) failed to describe the timing of cell salvage. Twenty‐six trials studied cell salvage systems that reinfused washed salvaged blood, and 44 trials studied cell salvage systems that reinfused unwashed filtered salvaged blood. One trial (Rollo 1995) studied both washed and unwashed cell salvage (four‐arm trial) and provided two comparisons of cell salvage (Rollo 1995a; Rollo 1995b). One trial (Klein 2008) studied intra‐operative washed and post‐operative unwashed cell salvage. For three trials (Mercer 2004; Sait 1999; Zhang 2008) the method used to process salvaged autologous blood prior to re‐transfusion was unclear.

Types of cell salvage devices

Various types of cell salvage (autotransfusion) systems were studied. The following is a list of those systems used.

ABTrans autologous re‐transfusion system
Atrium 2050
Atrium 2550 in‐line autotransfusion drainage system
Autovac postoperative orthopaedic autotransfusion canister
Bard cardiotomy reservoir
Bellovac ABT autotransfusion system
Beijing PerMed Biomedical Engineering Company
Bentley Catr hard shell cardiotomy reservoir
BIODREN autotransfusion system
BRAT‐2 Cell Saver
CATR 3500 cardiotomy reservoir
Cell Trans system (Summit Medical)
ConstaVac CBC system
ConstaVac CBCII system
COBE Bayler rapid autotransfusion system
Dideco Compact
Dideco Electra system
Dideco 742 cardiotomy reservoir
Dideco 797 reinfusion system (Sorin Biomedical)
DONOR system (Van Straten Medical)
Electromedic Autotrans AT‐100
Electromedics BT‐795
Flow‐Gard 6200 (Baxter)
Frensenius continuous autotransfusion system (C.A.T.S)
Gish Orthofuser Biomedical autotransfusion system
Haemonetics Cell Saver
Haemonetics Cell Saver 3
Haemonetics Cell Saver 3 Plus
Haemonetics Cell Saver 4
Haemonetics Cell Saver 5
Haemonetics Haemolite cell washer
Haemonetics Haemolite‐2
Medtronic Autolog system
Ortho‐Evac system
Pleur‐evac autotransfusion system
Shiley hardshell cardiotomy reservoir
Solcotrans Cell Saver
Solcotrans Orthopedic Plus system
Solcotrans Orthopedic system
Sorenson ATS (autotransfusion system)
Terumo TE‐171 system (Terumo)

Transfusion 'triggers' or thresholds

Of the 75 included trials, 60 reported the use of a transfusion protocol. Of these, 59 trials included a transfusion 'trigger' value, that being the haemoglobin (Hb) or haematocrit (Hct) value at which point a transfusion of red blood cells, was considered appropriate. However, there was significant variation between trials in the transfusion 'trigger' value used. The post‐operative transfusion trigger for haemoglobin (Hb) ranged from 7.0 g/dL to 10.0 g/dL, whereas the intra‐operative Hb transfusion trigger value ranged from 5.6 to 10.0 g/dL.

Of the 55 trials that reported a post‐operative transfusion threshold,15 trials reported a Hb transfusion threshold of 10.0 g/dL, 15 trials reported a transfusion threshold of between 9.0 g/dL and 9.5 g/dL, 21 trials reported a transfusion threshold of between 8.0 g/dL and 8.9g/dL, two trials reported a transfusion threshold of 7.0 g/dL, one trial reported a transfusion threshold of 5.0 mmol/L and one trial transfused patients when the haematocrit value was less than 30%. Of the 21 trials that reported the use of an intra‐operative transfusion threshold, the haemoglobin 'trigger' value ranged from as low as 5.6 g/dL to as high as 10.0 g/dL.

Risk of bias in included studies

The performance of the included trials against each domain is summarised in the 'Risk of bias' tables (Figure 1; Figure 2).

Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.

Methodological quality summary: review authors' judgements about each methodological quality item for each included study.

Adequate sequence generation

The risk of bias for this item was judged to be low for 21 trials. For four trials the method of sequence generation was judged to be inadequate. The remaining 50 trials presented no information regarding the method of sequence generation and were rated unclear.

Allocation concealment

The risk of bias for this item was judged to be low in one trial (Pleym 2005) which used central randomisation (off‐site, computer‐generated randomisation). For 27 trials the method used to conceal treatment allocation was judged to be inadequate (for example sealed envelopes). The remaining 47 trials presented no information regarding the method of allocation concealment and were rated unclear.

Blinding

None of the 75 included trials were judged to be double‐blind. Given the nature of the intervention double‐blinding is accepted as being problematic.

Effects of interventions

Aggregated analysis

Sixty‐seven trials of cell salvage (autotransfusion) reported data on the number of subjects exposed to allogeneic blood transfusion. These trials included a total of 6025 patients of whom 3048 were randomised to cell salvage. Overall, the use of cell salvage reduced the rate of exposure to allogeneic RBC transfusion by a relative 38% (RR 0.61; 95% CI 0.55 to 0.70). Heterogeneity between these trials was statistically significant (P<0.00001, I²=81%). The absolute reduction in risk (ARR) of exposure to allogeneic blood transfusion was 21% (risk difference (RD) ‐0.21; 95% CI ‐0.26 to ‐0.15) and, on average, 4.8 patients would need to undergo cell salvage so that one would avoid an allogeneic RBC transfusion (number needed to treat (NNT)).

Transfusion protocol

Of the 67 trials that reported data on the number of subjects exposed to allogeneic blood transfusion, 52 reported the use of transfusion protocols. These trials included a total of 4755 patients, of whom 2377 were randomised to cell salvage. The relative risk of exposure to allogeneic RBC transfusion in those patients treated with cell salvage was 0.62 (95% CI 0.54 to 0.71). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 81%). For the 15 trials that did not report the use of transfusion protocols, the relative risk of exposure to allogeneic RBC transfusion was 0.58 (95% CI 0.41 to 0.82). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 77%).

Type of surgery

Of the trials that reported data on the number of patients exposed to allogeneic RBC transfusion, 32 involved orthopaedic surgery, 31 involved cardiac procedures and four involved vascular surgery. A larger relative risk reduction (RRR) was observed in orthopaedic trials (RRR 54%; 95% CI 43% to 63%) than in cardiac trials (RRR 23%; 95% CI 14% to 31%). For those four trials that involved vascular surgery, the relative risk of exposure to allogeneic blood transfusion was not statistically significant (RR 0.63; 95% CI 0.34 to 1.15). For each of these three subgroup analyses, heterogeneity was statistically significant (I² = 72%, I² = 62%, and I² = 81%; respectively).

Type of cell salvage ‐ washed versus unwashed

Twenty‐seven trials studied washed cell salvage whilst 40 trials investigated unwashed cell salvage. One trial (Rollo 1995) studied both washed and unwashed cell salvage (four‐arm trial) and provided two comparisons of cell salvage (Rollo 1995a; Rollo 1995b). One trial (Sait 1999) did not describe the method of cell salvage investigated.

Overall, when cell salvage was conducted with devices that washed salvaged blood, the relative risk of exposure to red cell transfusion was only slightly lower than that with unwashed cell salvage. For those trials that used washed cell salvage the relative risk of exposure to allogeneic RBC transfusion was 0.60 (95% CI 0.51 to 0.70) compared to 0.66 (95% CI 0.57 to 0.77) for those trials that used unwashed cell salvage. For both these subgroup analyses, heterogeneity was statistically significant (I² = 68% and I² = 81%; respectively).

Timing of cell salvage

Eleven trials reported the use of intra‐operative cell salvage. These trials included a total of 805 patients, of whom 402 were randomised to intra‐operative cell salvage. The relative risk of exposure to allogeneic blood transfusion was 0.59 (95% CI 0.46 to 0.76). Heterogeneity between these trials was statistically significant (P = 0.002, I² = 64%). Forty‐six trials reported the use of post‐operative cell salvage. These trials included a total of 4361 patients, of whom 2209 were randomised to post‐operative cell salvage. The risk of exposure to allogeneic blood transfusion was reduced on average by a relative 37% in those patients treated with post‐operative cell salvage compared to control (RR =0.63; 95% CI 0.54 to 0.74). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 83%). Nine trials reported the use of both intra‐ and post‐operative cell salvage. These trials included a total of 737 patients of whom 357 were randomised to intra‐ and post‐operative cell salvage. The use of intra‐ and post‐operative cell salvage decreased exposure to allogeneic red cell transfusion by a relative 30% (RR 0.70; 95% CI 0.54 to 0.92). Heterogeneity between these trials was statistically significant (P < 0.005, I² = 64%).

Volume of blood transfused

Thirty‐two trials provided data for the volume of allogeneic RBC transfused. These trials included a total of 2321 patients, of whom 1172 were randomised to cell salvage. On average, the use of cell salvage reduced the volume of red cells transfused by 0.68 units per patient (WMD ‐0.68 units; 95% CI ‐0.88 to ‐0.49 units). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 75%). For those 27 trials that reported the use of a transfusion protocol the use of cell salvage reduced the amount of allogeneic blood transfused by an average of 0.69 units per patient (WMD ‐0.69 units; 95% CI ‐0.90 to ‐0.49 units). For those five trials that did not report the use of a transfusion protocol the use of cell salvage did not statistically significantly reduce the amount of allogeneic blood transfused (WMD ‐0.64 units; 95% CI ‐1.30 to 0.01 units). For both these subgroup analyses, heterogeneity between trials was statistically significant (P < 0.00001 and P < 0.0001, respectively).

When data were stratified by the type of surgery performed, greater reductions in the volume of allogeneic RBC transfused per patient were observed in trials that involved orthopaedic surgery (WMD ‐0.81 units; 95% CI ‐1.22 to ‐0.39 units) compared to cardiac surgery (WMD ‐0.67 units; 95% CI ‐0.89 to ‐0.44 units). For both these subgroup analyses, heterogeneity between trials was statistically significant (P < 0.00001 and P < 0.0001, respectively). Similar statistically significant reductions in the volume of RBC transfused was not observed in trials involving vascular surgery (WMD 0.02 units; 95% CI ‐0.34 to 0.38 units).

Type of cell salvage ‐ washed versus unwashed ‐ type of surgery

When the type of cell salvage was stratified by the type of surgery, the results indicated that the use of washed cell salvage in cardiac surgery was associated with an average 34% relative risk reduction in exposure to allogeneic red cell transfusion (RR 0.66; 95% CI = 0.55 to 0.80). Heterogeneity between these trials was statistically significant (P = 0.002, I² = 61%). Reduced exposure to red cell transfusion was also observed in those trials that used unwashed cell salvage in cardiac surgery, (RR 0.85; 95% CI 0.76 to 0.95), heterogeneity between these trials was statistically significant (P = 0.0001, I² = 65%). For orthopaedic trials both types of cell salvage were associated with significant reductions in transfusion exposure rates. For washed cell salvage the relative risk of exposure to red cell transfusion was reduced on average by 52% (RR 0.48; 95% CI 0.36 to 0.64) and for unwashed cell salvage there was a relative 53% reduction in risk of exposure (RR 0.47; 95% CI 0.36 to 0.63). For these subgroup analyses, heterogeneity was statistically significant (P = 0.03 and P = 0.0001, respectively). All four trials conducted in the setting of vascular surgery used washed cell salvage. For these trials, the relative risk of exposure to allogeneic red cell transfusion was not statistically significant (RR 0.63; 95% CI 0.34 to 1.15). Heterogeneity between these trials was statistically significant (P = 0.001, I² = 81%).

Blood loss

Thirty‐two trials of cell salvage reported data for total blood loss. These trials included a total of 2311 patients of whom 1158 were randomised to cell salvage. The use of cell salvage did not appear to adversely impact on blood loss volumes (WMD ‐39.02 mls; 95% CI ‐85.10 to 7.05 mls). Heterogeneity between these trials was statistically significant (P = 0.01, I² = 41%).

Active versus control

Forty‐three trials compared cell salvage alone to a control group who did not receive cell salvage (autotransfusion) or any other form of active treatment. These trials included a total of 3666 patients, of whom 1854 were randomised to cell salvage. The relative risk of receiving an allogeneic RBC transfusion was 0.61 (95% CI 0.52 to 0.71). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 83%).

Transfusion protocol

Thirty‐four trials of cell salvage reported the use of transfusion protocols. These trials included a total of 3030 patients of whom 1525 were randomised to cell salvage. For those trials that used a transfusion protocol, the risk of receiving an allogeneic red cell transfusion was reduced on average by 39% (RR 0.61; 95% CI 0.51 to 0.73). For the nine trials that did not report the use of a transfusion protocol, the relative risk of receiving an allogeneic red cell transfusion was 0.56 (95% CI 0.35 to 0.89). For both these subgroup analyses, heterogeneity was statistically significant (P < 0.00001).

Type of surgery

Of the 43 trials that provided data for the number of patients exposed to allogeneic blood transfusion, 18 involved cardiac surgery, 21 involved orthopaedic surgery, and four involved vascular surgery. When cell salvage was used in orthopaedic surgery, the risk of exposure to red cell transfusion was reduced by a relative 55% (RR 0.45; 95%CI 0.34 to 0.60) compared to 22% (RR 0.78; 95% CI 0.68 to 0.91) in the case of cardiac surgery. For the four trials involving vascular surgery, the relative risk of receiving an allogeneic RBC transfusion was not statistically significant (RR 0.63; 95% CI 0.34 to 1.15). Heterogeneity was statistically significant for each of the subgroups analysed (P < 0.001).

Type of cell salvage ‐ washed versus unwashed

When washed cell salvage was used, the relative risk of receiving an allogeneic RBC transfusion was 0.57 (95% CI 0.45 to 0.72) compared to 0.67 (95% CI 0.56 to 0.81) for unwashed cell salvage. Heterogeneity was statistically significant in the two subgroups analysed (P < 0.00001).

Timing of cell salvage

Eight trials of intra‐operative cell salvage provided data for the number of patients exposed to allogeneic RBC transfusion. These trials included a total of 564 patients, of whom 281 were randomised to cell salvage. Intra‐operative cell salvage was associated with a relative reduction in the risk of receiving an allogeneic RBC transfusion of 36% (RR 0.64; 95% CI 0.50 to 0.83). Heterogeneity between these trials was statistically significant (P = 0.07, I² = 46%). Twenty‐nine trials of post‐operative cell salvage provided data for the number of patients exposed to allogeneic RBC transfusion. These trials included a total of 2852 patients, of whom 1452 were randomised to cell salvage. The use of post‐operative cell salvage reduced the risk of exposure to allogeneic red cell transfusion by a relative 41% (RR 0.59; 95% CI 0.48 to 0.73). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 87%). Six trials studied both intra‐ and post‐operative cell salvage. These trials included a total of 250 patients of whom 121 were randomised to cell salvage. with The combined use of intra‐ and post‐operative cell salvage reduced exposure to allogeneic RBC transfusion by a relative 42% (RR 0.58; 95% CI 0.35 to 0.95). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 84%).

Volume of blood transfused

Twenty‐three trials reported data for the volume of allogeneic RBC transfused. These trials included a total of 1608 patients, of whom 818 were randomised to cell salvage. Overall, the use of cell salvage was associated with a modest reduction in the volume of red cells transfused. In those patients treated with cell salvage, there was an average saving of 0.81 units of RBC per patient (WMD ‐0.81 units; 95% CI ‐1.08 to ‐0.54 units). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 80%).

Volume of blood transfused ‐ transfusion protocol

Stratifying the volume of blood transfused by the presence of a transfusion protocol, showed that greater reductions in the volume of red cells transfused were observed in those trials that reported the use of transfusion protocols (WMD ‐86; 95% CI ‐1.17 to ‐0.55 units) compared to those trials that did not use a transfusion protocol to guide transfusion practice (WMD ‐0.68 units; 95% CI ‐1.63 to 0.27 units). For both these subgroup analyses, heterogeneity was statistically significant (P < 0.00001).

Volume of blood transfused ‐ type of surgery

There were 13 cardiac trials, including a total of 995 patients, that reported data for the volume of RBC transfused. The use of cell salvage in cardiac surgery provided, on average, a saving of around one unit of blood per patient (WMD ‐0.93 units; 95% CI ‐1.27 to ‐0.59 units). Heterogeneity between these trials was statistically significant (P < 0.0001, I² = 71%). There were seven orthopaedic trials, including a total of 427 patients, that reported data for the volume of blood transfused. For those patients randomised to cell salvage, there was an average saving of 0.82 units of RBC per patient (WMD ‐0.82 units; 95% CI ‐1.36 to ‐0.27 units). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 86%). For the three trials involving vascular surgery, the weighted mean difference in the volume of RBC transfused was not statistically significant (WMD 0.02 units; 95% CI ‐0.34 to 0.38 units).

Blood loss

Twenty‐four trials of cell salvage reported data for total blood loss. These trials included a total of 1570 patients, of whom 789 were randomised to cell salvage. The use of cell salvage did not appear to adversely impact on total blood loss (WMD ‐38.98 mls; 95% CI ‐99.91 to 21.95 mls). Heterogeneity between these trials was statistically significant (P = 0.01, I² = 44%).

Active versus active

There were 25 trials that compared cell salvage, combined with another form of active treatment (blood conservation intervention), to a control group who received the same active treatment but did not receive cell salvage (autotransfusion). For these 25 trials the relative risk of exposure to allogeneic RBC transfusion was 0.63 (95% CI 0.53 to 0.76). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 76%).

Transfusion protocol

There were 19 trials of cell salvage that reported the use of transfusion protocols. For these trials, the relative risk of exposure to allogeneic blood transfusion was 0.64 (95% CI 0.52 to 0.77). Heterogeneity between these trials was statistically significant (P < 0.00001, I² = 79%). For those six trials that did not report the use of a transfusion protocol, the relative risk of exposure to allogeneic red cells was 0.59 (95% CI 0.35 to 1.00).

Type of surgery

Fourteen trials involving cardiac surgery provided data for the number of patients receiving allogeneic red cell transfusion. These trials included a total of 1304 patients, of whom 641 were randomised to cell salvage. For these trials, the risk of receiving an allogeneic red cell transfusion was reduced on average by a relative 27% (RR 0.73; 95% CI 0.61 to 0.87). Heterogeneity between these trials was statistically significant (P < 0.0001, I² = 72%). For the 11 trials involving orthopaedic surgery, the risk of receiving an allogeneic red cell transfusion was reduced by a relative 54% (RR 0.46; 95% CI 0.33 to 0.65). Heterogeneity between these trials was statistically significant (P = 0.03, I² = 50%).

Type of cell salvage ‐ washed versus unwashed

When washed cell salvage was used, the relative risk of receiving an allogeneic red cell transfusion was 0.61 (95% CI 0.48 to 0.78) compared to 0.68 (95% CI 0.52 to 0.89) for unwashed cell salvage. For both of these subgroup analyses, heterogeneity was statistically significant (P = 0.01 and P < 0.00001, respectively).

Timing of cell salvage

Four trials of intra‐operative cell salvage provided data on the number of patients receiving allogeneic red cell transfusion. These trials included a total of 281 patients, of whom 141 were randomised to cell salvage. Intra‐operative cell salvage reduced the rate of exposure to allogeneic red cell transfusion by a relative 46% (RR 0.54; 95% CI 0.35 to 0.84). The 18 trials of post‐operative cell salvage included a total of 1629 patients, of whom 817 were randomised to cell salvage. For these trials, the risk of receiving an allogeneic red cell transfusion was reduced on average by a relative 35% (RR 70; 95% CI 0.50 to 0.85). Heterogeneity for both these subgroup analyses was statistically significant (P = 0.005 and P < 0.00001, respectively). For the four trials involving intra‐ and post‐operative cell salvage, the relative risk of receiving an allogeneic RBC transfusion was 0.76 (95% CI 0.59 to 0.98; I² = 27%).

Volume of blood transfused

Eight trials of cell salvage reported data for the volume of allogeneic red cells transfused. On average there was a saving of 0.66 units of RBC per patient (WMD ‐0.66; 95% CI ‐1.02 to ‐0.30) when cell salvage was combined with another form of active treatment. Of these eight trials, six involved cardiac surgery and seven reported the use of transfusion protocols. For these two subgroup analyses, the volume of allogeneic red cells transfused was reduced on average by 0.39 units per patient (WMD ‐0.39; 95% CI ‐0.67 to ‐0.12) and by 0.67 units per patient (WMD ‐0.67; 95% CI ‐1.09 to ‐0.25), respectively. In the case of the two orthopaedic trials, the use of cell salvage reduced the amount of allogeneic red blood cells transfused on average by 1.1 units per patient (WMD ‐1.10; 95% CI ‐1.91 to ‐0.29; I² = 79%).

Blood loss

Eight trials of cell salvage reported data for total blood loss. These trials included a total of 741 patients of whom 369 were randomised to cell salvage. The use of cell salvage did not appear to adversely impact on total blood loss (WMD ‐48.32 mls; 95% CI ‐116.38 to 19.74 mls).

Adverse events and other outcomes

Mortality

In aggregate, 22 trials of cell salvage reported data for mortality. These trials included a total of 1788 patients, of whom 902 were randomised to cell salvage. For eight trials there were no reported deaths for either the intervention or control groups; therefore it was not possible to estimate the relative risks for these trials. Overall, the use of cell salvage did not statistically significantly impact on mortality (RR 0.96; 95% CI 0.49 to 1.88; I² = 0%). Similar non‐significant results were observed for those 16 trials that compared cell salvage alone to untreated controls (RR 1.41; 95% CI 0.66 to 3.05; I² = 0%). It should be noted that two trials provided over 22% of the information in the analysis and six of the 16 trials reported no deaths in either the intervention or control groups. There were six trials of active versus active comparisons (cell salvage combined with another form of active treatment compared to an actively treated control group) that provided mortality data. These trials included a total of 656 patients, of whom 326 were randomised to cell salvage. For these trials there were nine recorded deaths; eight of which occurred in the control arms. The use of cell salvage in this subgroup of trials did not appear to impact significantly on the rate of mortality (RR 0.27; 95% CI 0.07 to 1.08; I² = 0%).

Re‐operation for bleeding

In aggregate, 19 trials of cell salvage provided data for re‐operation due to bleeding. These trials included a total of 1683 patients, of whom 841 were randomised to cell salvage. The use of cell salvage did not statistically significantly impact on the rates of re‐operation for bleeding (RR 0.90; 95% CI 0.53 to 1.53; I² = 0%). Ten trials, comparing cell salvage to untreated controls (active versus control comparisons), reported data for re‐operation due to bleeding. These trials included a total of 688 patients, of whom 349 were randomised to cell salvage. The relative risk of requiring a re‐operation due to bleeding was not statistically significant (RR 1.00; 95% CI 0.45 to 2.24; I² = 0%). There were nine trials of active versus active comparisons (cell salvage combined with another form of active treatment compared to an actively treated control group) that provided data for re‐operation due to bleeding. Cell salvage did not appear to statistically significantly impact on the rates of re‐operation for bleeding in this subset of trials (RR 0.83; 95% CI 0.41 to 1.68; I² = 0%).

Any infection

In aggregate, 23 trials reported data for infection of any type. These trials included a total of 2892 patients, of whom 1474 were randomised to cell salvage. The use of cell salvage appeared to be associated with a slight decrease in the rate of infection compared to control (RR 0.68; 95% CI 0.46 to 0.99; I² = 0%). Sixteen trials, comparing cell salvage to untreated controls, reported data for infection of any type. These trials included a total of 1860 patients, of whom 942 were randomised to cell salvage. A statistically significant reduction in the rate of infection in those patients treated with cell salvage was observed (RR 0.63; 95% CI 0.41 to 0.96; I² = 1%). For the seven trials that investigated active versus active comparisons the relative risk of developing an infection was not statistically significant (RR 0.88; 95% CI 0.37 to 2.07; I²=0%).

Wound complications

In aggregate, 16 trials of cell salvage reported data for a wound complication (for example haematoma, infection). These trials included a total of 1962 patients, of whom 1011 were randomised to cell salvage. The use of cell salvage did not statistically significantly impact on the rates of wound complication (RR 0.94; 95% CI 0.57 to 1.55; I² = 0%). Similar results were observed for the twelve trials that compared cell salvage alone to an untreated control group (RR 0.88; 95% CI 0.49 to 1.57; I² = 0%). For the four trials of active versus active comparisons, the relative risk of developing a wound complication was 1.13 (95% CI 0.43 to 2.99; I² = 0%).

Any thrombosis

In aggregate, 11 trials reported data for thrombosis of any type. These trials included a total of 925 patients, of whom 476 were randomised to cell salvage. For five of the 11 trials, there were no reported events of thrombosis in either the intervention or control groups. The relative risk of developing any thrombosis in those patients treated with cell salvage compared to control was not statistically significant (RR 1.13; 95% CI 0.48 to 2.66; I² = 0%). Of the 11 trials that reported data for this outcome 10 involved active versus untreated control comparisons.

Stroke

In aggregate, seven trials reported data for stroke. These trials included a total of 696 patients, of whom 347 were randomised to cell salvage. The relative risk of developing a stroke in those patients treated with cell salvage compared to control was not statistically significant (RR 0.65; 95% CI 0.21 to 1.98; I²=0%).

Non‐fatal myocardial infarction

In aggregate, 11 trials reported data for non‐fatal myocardial infarction. These trials included a total of 951 patients, of whom 471 were randomised to cell salvage. The relative risk of non‐fatal myocardial infarction in those patients treated with cell salvage compared to control was not statistically significant (RR 0.80; 95% CI 0.43 to 1.46; I² = 0%). Similar results were observed for the six trials that compared cell salvage alone to an untreated control group (RR 0.65; 95% CI 0.32 to 1.31; I² = 0%). For those five trials that investigated active versus active comparisons, the relative risk of non‐fatal myocardial infarction was not statistically significant (RR 1.46; 95% CI 0.43 to 4.88; I² = 0%).

Deep vein thrombosis

Seven trials reported data for deep venous thrombosis. These trials included a total of 645 patients, of whom 320 were randomised to cell salvage. The relative risk of developing deep venous thrombosis in those subjects treated with cell salvage compared to control was not statistically significant (RR 0.83; 95% CI 0.31 to 2.23; I² = 0%).

Hospital length of stay

Ten trials reported data for hospital length of stay. These trials included a total of 772 patients, of whom 399 were randomised to cell salvage. In those patients treated with cell salvage hospital length of stay was not statistically significantly reduced compared to control (WMD ‐1.38 days; 95% CI ‐2.79 to 0.03 days; I² = 83%).

Discussion

Principal findings

We identified 75 randomised trials of cell salvage, carried out over a 29‐year period (1979 to 2008). Overall, the results of the meta‐analysis indicated that the use of cell salvage reduced peri‐operative allogeneic RBC transfusion exposure by a relative 38% (RR 0.62; 95% CI 0.55 to 0.70). The average absolute reduction in risk (ARR) of exposure to allogeneic red cell transfusion was 21% (RD ‐0.21; 95% CI ‐0.26 to ‐0.15), equating to a number needed to treat (NNT) of 4.8. The efficacy of cell salvage in reducing the need for allogeneic red cell transfusion appeared to be greatest in the setting of orthopaedic surgery. In this setting, cell salvage reduced the risk of exposure to allogeneic blood transfusion by a relative 54% compared to 23% in cardiac surgery. In orthopaedic surgery very little difference in treatment effects was observed between washed cell salvage (RR 0.48; 95% CI 0.36 to 0.64) and unwashed cell salvage (RR 0.47; 95% CI 0.36 to 0.63). This was not the case in cardiac surgery, where there were clear differences between washed cell salvage, which showed significant reductions in allogeneic red cell transfusion rates (RR 0.66; 95% CI 0.55 to 0.80), and unwashed cell salvage, which appeared to be only marginally effective (RR 0.85; 95% CI 0.76 to 0.95). However, significant variation in treatment effects was observed for all of the main study outcomes.

In the case of exposure to allogeneic blood transfusion, a relative risk reduction of 37% was observed when cell salvage was combined with another active intervention (for example pre‐operative autologous donation (PAD), acute normovolaemic haemodilution (ANH), aprotinin) and compared with that intervention on its own. When cell salvage was compared to a non‐active intervention (for example standard untreated control), a relative risk reduction of 39% was observed.

The use of cell salvage was also associated with only slight reductions in the volume of red cells transfused. Overall, in those patients treated with cell salvage, there was an average saving of 0.68 units of RBC per patient (WMD ‐0.68 units; 95% CI ‐0.88 to ‐0.49 units). When cell salvage was combined with another form of active intervention and compared with that intervention on its own, the reduction in the volume of RBC transfused was around 0.66 units of RBC per patient (WMD ‐0.66 units; 95% CI ‐1.02 to ‐0.30 units). Such a result may well have been expected as both groups were actively treated with some form of blood sparing intervention.

Sources of heterogeneity

The observed variation in treatment effects was in terms of both the size and direction of effect with relative risk point estimates for red cell transfusion exposure for the individual trials, ranging from 0.03 to 5.64. Of the 67 trials that provided data for the number of patients exposed to allogeneic red cell transfusion, more than half of these trials (n = 36) found that cell salvage did not statistically significantly reduce the risk of receiving a red cell transfusion. None of the subgroup analyses performed established a clear reason for the observed variability in treatment effect. Statistically significant heterogeneity was observed for all of the subgroup analyses performed.

The impact that trial methodological quality had on treatment effects was difficult to determine, as the majority of trials were of poor quality. The most concerning feature of the trials reviewed here is that only one trial reported a method of concealing treatment allocation that was judged to be adequate. A lack of allocation concealment has been shown to significantly influence the estimate of treatment effect (Schulz 1995). When trials that reported data for the number of patients transfused allogeneic red blood cells were stratified by methodological quality (adequate allocation concealment: yes, unclear, or no), the relative risks of exposure to allogeneic blood transfusion varied only slightly. For those trials (n = 24) that were assessed as providing inadequate concealment of treatment allocation the relative risk of exposure to allogeneic red cell transfusion was 0.62 (95% CI 0.51 to 0.75), whereas for those trials (n=42) that did not report the method used to conceal treatment allocation or it was unclear what method was used to conceal treatment allocation, the relative risk was 0.62 (95% CI 0.53 to 0.72).

Sources of bias

The majority of trials reviewed here were small with less than 60 participants in each trial arm. Reliance on small trials raises concerns about the effects of publication bias. Funnel plot assessment revealed some evidence of this in the form of a 'missing' population of small negative studies (Figure 3). Although there is a clustering of trials around the null (RR = 1), there were very few trials that showed an overall negative result for cell salvage for allogeneic red cell transfusion exposure.

Funnel plot of comparison: 1 Cell Salvage ‐ Blood Transfused (All Studies), outcome: 1.1 No. exposed to allogeneic blood (All Studies).

Although this review did not exclude studies on the basis of language, only six non‐English studies fulfilled the inclusion criteria (Fragnito 1995; Lorentz 1991; Menges 1992; Naumenko 2003; Rosencher 1994; Zhang 2008). For those studies published in English language, the relative risk of exposure to allogeneic red cell transfusion was 0.61 (95% CI 0.54 to 0.70) compared to 0.63 (95% CI 0.50 to 0.79) for those studies published in languages other than English. Due to the lack of non‐English language studies it is difficult to interpret these results with any degree of confidence. However, it is of interest to note, that although the heterogeneity in treatment effect for the English language studies was statistically significant (P < 0.00001, I² = 83%) this was not the case with those studies published in non‐English languages (P = 0.89, I² = 0%).

The main study outcome used in these trials was a practice variable (the decision to transfuse a patient with allogeneic red cells) and, as such, may have been a major source of bias. The decision to transfuse a patient requires a degree of subjectivity on the part of the clinicians, and as all the trials reviewed here were unblinded and lacked adequate concealment of treatment allocation. This is a particularly important source of bias that may have potentially influenced the results in favour of cell salvage.

Adverse events and other outcomes

Mortality, re‐operation for bleeding, infection, wound complication, non‐fatal myocardial infarction, thrombosis, stroke, and hospital length of stay did not appear to be adversely affected by the use of cell salvage. Even though one of the known risks associated with cell salvage is infection (due to contamination of the autologous product during the salvaging and reinfusion process), fewer cell salvage patients experienced infection (RR 0.68; 95% CI 0.46 to 0.99). However, cell salvage did not appear to be associated with any significant reductions in wound complications (RR 0.94; 95% CI 0.57 to 1.55).

It should be noted that the event rates were small ranging for 1.2% in the case of stroke to 5.1% in the case of any infection. Therefore it is difficult to draw firm conclusions regarding the impact of cell salvage on important clinical outcomes. There were only two outcomes (any infection and non‐fatal myocardial infarction) where the event rates in the control groups were greater than 3.5% (5.1% and 4.8%, respectively). From the very limited data, it appears that the potential benefit of cell salvage in reducing exposure to allogeneic blood transfusion, is not offset by serious adverse effects.

Clinical significance of the results

In an attempt to avoid allogeneic blood transfusion during the perioperative period, technologies such as cell salvage have been introduced without firm evidence to support their use. Although cell salvage provides peace of mind, knowing a patient's own blood will be transfused should it be needed, cell salvage is not without its own risks and costs (Forgie 1998). The risks associated with cell salvage are well documented, and include, bacterial contamination of the salvaged blood, air embolism, nephrotoxicity, and coagulation disorders (Faught 1998; Huet 1999; Semmens 2000; Spahn 2000). In its most simplistic form, unwashed cell salvage merely represents a very laborious means of obtaining an autologous volume expander, which is not necessarily advocated due to the potential serious side effects (Huet 1999). Although washed cell salvage provides a better quality blood product, the overall cost of this technology is rather substantial. However, a recent cost‐effectiveness analysis indicated that cell salvage was cost‐effective compared with all other transfusion strategies except ANH (Davies 2006). This study indicated that the net benefit of cell salvage was between £112 and £359 per person, compared with the allogeneic blood transfusion strategy, PAD, PAD plus erythropoietin (EPO), fibrin sealants, antifibrinolytics, and EPO. This study claimed that the use of cell salvage could result in net reductions in the volume of allogeneic blood transfused of between 6500 and 320,000 units per year, translating into annual savings to the National Health Service (NHS) of £0.73 million to £36 million (Davies 2006).

Any intervention that forms part of a blood conservation strategy needs to be critically examined in respect to its cost‐effectiveness. On the basis of cost alone, cell salvage may appear to be an attractive alternative to the other currently available technologies, in particular aprotinin. However, as highlighted by Fergusson 1999a, in many instances neither the cost of the technologists needed to operate the device nor the cost of the cell saver device itself are considered in the overall calculations of cost. Further to this, Fergusson 1999a propose that the conclusions of studies that suggest cell salvage is cost saving should be interpreted with caution.

The true value of avoiding allogeneic blood transfusion remains debatable. Those concerned with the risks of transfusion transmitted disease (TTD) will be more interested in avoiding blood transfusion completely, rather than reducing the volume of blood transfused. However, the importance of avoiding transfusion depends on the probability of avoiding disease transmission, or other adverse effects that have been attributed to blood transfusion, such as alloimmunisation or febrile non‐haemolytic transfusion reaction (FNHTR). The rate of HIV or viral hepatitis transmission in most developed countries is very low, due to the presence of quality blood screening programmes (Coyle 1999; Whyte 1997). However, this assumption does not equally apply to developing countries where allogeneic blood is frequently administered without adequate screening in an environment where there is a high prevalence of viral pathogens amongst donors (Kimball 1995; McFarland 1997). In these settings there may be much greater clinical value in a range of interventions that diminish or avoid the need for allogeneic blood. However, the costs associated with such interventions may be prohibitive in developing countries, a situation that may well apply to cell salvage.

Most of the data have been collected in the context of major cardiac and orthopaedic surgery, where blood loss is often substantial. Consequently the applicability of the results to clinical settings where blood loss is minor is questionable. This review has highlighted the fact that cell salvage is frequently used alongside other interventions designed to minimise the need for allogeneic blood transfusion. This is particularly evident in the area of cardiac surgery, where over the last 10‐20 years there has been a steady increase in the use of anti‐fibrinolytic drugs (for example aprotinin, tranexamic acid, epsilon aminocaproic acid), acute normovolaemic haemodilution (ANH), and pre‐operative autologous donation (PAD). A meta‐analysis (Henry 2007) of the aforementioned anti‐fibrinolytic drugs showed that both aprotinin and tranexamic acid were highly efficacious in reducing surgical blood loss and allogeneic blood transfusion in cardiac surgery. However, the findings of a Canadian study (Fergusson 2008), which reported an increased risk of death in cardiac surgery patients treated with aprotinin compared with the lysine analogues (tranexamic acid and epsilon aminocaproic acid), led to the market suspension of aprotinin on November 5, 2007. The loss of aprotinin from the armamentarium of the cardiac surgeon has lead to a re‐exploration of alternative approaches to haemostasis management (Baker 2009).

The evidence on the efficacy and safety of ANH and PAD has been reviewed by the International Study of perioperative Transfusion (ISPOT) group (Bryson 1998; Forgie 1998). The literature regarding these interventions is generally viewed as being of indifferent quality because of inadequate randomisation and lack of blinding of outcomes assessment. However, these techniques have been shown to have modest blood sparing effects. This and the growing evidence on the efficacy of transfusion triggers indicates that a more conservative approach to blood transfusion is generally desirable in patients without cardiovascular risk factors, such as ischaemic heart disease or cerebral vascular disease (Carson 1998; Hebert 1999). This conservative approach, combined with the use of anti‐fibrinolytic drugs may well offer the best approach for managing the transfusion requirements of patients in high‐risk settings such as cardiac surgery. However, in settings other than cardiac, such as vascular and orthopaedic surgery, the choice of intervention that best minimises patient exposure to allogeneic blood transfusion is not that clear cut, although a more conservative approach to transfusion practice has been shown to be efficacious across a range of clinical domains (Carson 2002; Hill 2003). The decision to use any of the current available technologies as an alternative to allogeneic blood transfusion, including cell salvage, will be primarily based on availability, cost, and surgeon preference. To delineate the efficacy of the various technologies used in non‐cardiac settings is rather difficult as the current available evidence is equivocal.

Authors' conclusions

Implications for practice.

The use of washed cell salvage appears justified in orthopaedic surgery. However, in situations where there are concerns about the safety of the blood supply, the use of unwashed, filtered cell salvage may well be justified. Although the use of washed or unwashed cell salvage in cardiac surgery may well be justified on the basis of the evidence reviewed here, due consideration needs to be given to those technologies (that is anti‐fibrinolytic agents) that, unlike cell salvage, have been shown to significantly reduce peri‐operative blood loss and re‐operation due to bleeding in the context of cardiac surgery (Henry 2007). Re‐operation alone is associated with substantial additional costs due to additional surgery and costs associated with increased length hospital stay (Ray 1999).

Implications for research.

There is no need for further small randomised controlled trials of cell salvage in orthopaedic and cardiac surgery. The principal need is for large, methodologically rigorous, comparative trials to assess the relative efficacy, safety, and cost‐effectiveness of cell salvage in different surgical procedures.

What's new

Date	Event	Description
19 December 2011	Amended	The Plain Language Summary Title has been shortened to comply with new guidelines.

Date	Event	Description
10 February 2010	New citation required and conclusions have changed	The review has been updated with the results of 24 additional trials.
1 June 2006	New search has been performed	May 2006   The searches were updated in January 2004 as part of a Health Technology Assessment (HTA) project. Two new studies have been included (Naumenko 2003; Zhao 2003), with the results of the review amended accordingly.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 No. exposed to allogeneic blood (All Studies)	67	6025	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.55, 0.70]
2 No. exposed to allogeneic blood (Transfusion Protocol)	67		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Transfusion Protocol	52	4755	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.54, 0.71]
2.2 No Transfusion Protocol	15	1270	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.41, 0.82]
3 No. exposed to allogeneic blood (Type of Surgery)	67		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Cardiac	31	2518	Risk Ratio (M‐H, Random, 95% CI)	0.77 [0.69, 0.86]
3.2 Orthopaedic	32	3240	Risk Ratio (M‐H, Random, 95% CI)	0.46 [0.37, 0.57]
3.3 Vascular	4	267	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.34, 1.15]
4 No. exposed to allogeneic blood ‐ (Washed vs Unwashed)	67		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1 Washed	27	2225	Risk Ratio (M‐H, Random, 95% CI)	0.60 [0.51, 0.70]
4.2 Unwashed	40	3717	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.57, 0.77]
5 No. exposed to allogeneic blood (Timing)	66		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Intra‐operative cell salvage	11	805	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.46, 0.76]
5.2 Post‐operative cell salvage	46	4361	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.54, 0.74]
5.3 Intra & post‐operative cell salvage	9	737	Risk Ratio (M‐H, Random, 95% CI)	0.70 [0.54, 0.92]
6 Units of allogeneic blood transfused (All Studies)	32	2321	Mean Difference (IV, Random, 95% CI)	‐0.68 [‐0.88, ‐0.49]
7 Units of allogeneic blood transfused (Transfusion Protocol)	32		Mean Difference (IV, Random, 95% CI)	Subtotals only
7.1 Transfusion Protocol	27	1950	Mean Difference (IV, Random, 95% CI)	‐0.69 [‐0.90, ‐0.49]
7.2 No Transfusion Protocol	5	371	Mean Difference (IV, Random, 95% CI)	‐0.64 [‐1.30, 0.01]
8 Units of allogeneic blood transfused (Type of Surgery)	32		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.1 Cardiac	19	1497	Mean Difference (IV, Random, 95% CI)	‐0.67 [‐0.89, ‐0.44]
8.2 Orthopaedic	10	638	Mean Difference (IV, Random, 95% CI)	‐0.81 [‐1.22, ‐0.39]
8.3 Vascular	3	186	Mean Difference (IV, Random, 95% CI)	0.02 [‐0.34, 0.38]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 No. exposed to allogeneic blood (Cardiac)	31		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Washed	13	1158	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.55, 0.80]
1.2 Unwashed	18	1360	Risk Ratio (M‐H, Random, 95% CI)	0.85 [0.76, 0.95]
2 No. exposed to allogeneic blood (Orthopaedic)	32		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Washed	10	800	Risk Ratio (M‐H, Random, 95% CI)	0.48 [0.36, 0.64]
2.2 Unwashed	22	2357	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.36, 0.63]
3 No. exposed to allogeneic blood (Vascular)	4		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Washed	4	267	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.34, 1.15]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 No. exposed to allogeneic blood (Active vs Control)	43		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 Cell Salvage vs Control	43	3666	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.52, 0.71]
2 No. exposed to allogeneic blood (Transfusion Protocol)	43		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Transfusion Protocol	34	3030	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.51, 0.73]
2.2 No Transfusion Protocol	9	636	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.35, 0.89]
3 No. exposed to allogeneic blood (Type of Surgery)	43		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
3.1 Cardiac	18	1257	Risk Ratio (M‐H, Random, 95% CI)	0.78 [0.68, 0.91]
3.2 Orthopaedic	21	2142	Risk Ratio (M‐H, Random, 95% CI)	0.45 [0.34, 0.60]
3.3 Vascular	4	267	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.34, 1.15]
4 No. exposed to allogeneic blood (Washed vs Unwashed)	42		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
4.1 Washed	18	1322	Risk Ratio (M‐H, Random, 95% CI)	0.57 [0.45, 0.72]
4.2 Unwashed	24	2224	Risk Ratio (M‐H, Random, 95% CI)	0.67 [0.56, 0.81]
5 No. exposed to allogeneic blood (Timing)	43		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Intra‐operative cell salvage	8	564	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.50, 0.83]
5.2 Post‐operative cell salvage	29	2852	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.48, 0.73]
5.3 Intra & post‐operative cell salvage	6	250	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.35, 0.95]
6 Units of allogeneic blood transfused (Active vs Control)	23		Mean Difference (IV, Random, 95% CI)	Subtotals only
6.1 Cell Salvage vs Control	23	1608	Mean Difference (IV, Random, 95% CI)	‐0.81 [‐1.08, ‐0.54]
7 Units of allogeneic blood transfused (Transfusion Protocol)	23		Mean Difference (IV, Random, 95% CI)	Subtotals only
7.1 Transfusion Protocol	19	1339	Mean Difference (IV, Random, 95% CI)	‐0.86 [‐1.17, ‐0.55]
7.2 No Transfusion Protocol	4	269	Mean Difference (IV, Random, 95% CI)	‐0.68 [‐1.63, 0.27]
8 Units of allogeneic blood transfused (Type of Surgery)	23		Mean Difference (IV, Random, 95% CI)	Subtotals only
8.1 Cardiac	13	995	Mean Difference (IV, Random, 95% CI)	‐0.93 [‐1.27, ‐0.59]
8.2 Orthopaedic	7	427	Mean Difference (IV, Random, 95% CI)	‐0.82 [‐1.36, ‐0.27]
8.3 Vascular	3	186	Mean Difference (IV, Random, 95% CI)	0.02 [‐0.34, 0.38]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Total blood loss (All Studies)	32	2311	Mean Difference (IV, Random, 95% CI)	‐39.02 [‐85.10, 7.05]
2 Total blood loss (Active vs Control)	24	1570	Mean Difference (IV, Random, 95% CI)	‐38.98 [‐99.91, 21.95]
3 Total blood loss (Active vs Active)	8	741	Mean Difference (IV, Random, 95% CI)	‐48.32 [‐116.38, 19.74]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Language of Publication (All Studies)	66		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
1.1 English	60	5711	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.54, 0.70]
1.2 Non‐English	6	274	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.50, 0.79]
2 Methodological Quality ‐ Allocation concealment (All Studies)	67		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Allocation concealment ‐ Yes	1	47	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.04, 3.11]
2.2 Allocation concealment ‐ Unclear	42	3812	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.53, 0.72]
2.3 Allocation concealment ‐ No	24	2166	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.51, 0.75]

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Mortality ‐ All Studies	22	1788	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.49, 1.88]
2 Mortality ‐ Active vs Control	16		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
2.1 Active vs Control	16	1132	Risk Ratio (M‐H, Random, 95% CI)	1.41 [0.66, 3.05]
3 Mortality ‐ Active vs Active	6	656	Risk Ratio (M‐H, Random, 95% CI)	0.27 [0.07, 1.08]
4 Re‐operation for bleeding ‐ All Studies	19	1683	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.53, 1.53]
5 Re‐operation for bleeding ‐ Active vs Control	10		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
5.1 Active vs Control	10	688	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.45, 2.24]
6 Re‐operation for bleeding ‐ Active vs Active	9	995	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.41, 1.68]
7 Any infection ‐ All Studies	23	2892	Risk Ratio (M‐H, Random, 95% CI)	0.68 [0.46, 0.99]
8 Any infection ‐ Active vs Control	16		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
8.1 Active vs Control	16	1860	Risk Ratio (M‐H, Random, 95% CI)	0.63 [0.41, 0.96]
9 Any Infection ‐ Active vs Active	7	1032	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.37, 2.07]
10 Wound complication ‐ All Studies	16	1962	Risk Ratio (M‐H, Random, 95% CI)	0.94 [0.57, 1.55]
11 Wound complication ‐ Active vs Control	12		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
11.1 Active vs Control	12	1464	Risk Ratio (M‐H, Random, 95% CI)	0.88 [0.49, 1.57]
12 Wound complication ‐ Active vs Active	4	498	Risk Ratio (M‐H, Random, 95% CI)	1.13 [0.43, 2.99]
13 Any thrombosis ‐ All Studies	11	925	Risk Ratio (M‐H, Random, 95% CI)	1.13 [0.48, 2.66]
14 Any thrombosis ‐ Active vs Control	10		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
14.1 Active vs Control	10	807	Risk Ratio (M‐H, Random, 95% CI)	1.13 [0.48, 2.66]
15 Stroke ‐ All Studies	7	696	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.21, 1.98]
16 Stroke ‐ Active vs Control	5		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
16.1 Active vs Control	5	439	Risk Ratio (M‐H, Random, 95% CI)	0.79 [0.20, 3.21]
17 Stroke ‐ Active vs Active	2	257	Risk Ratio (M‐H, Random, 95% CI)	0.43 [0.06, 2.90]
18 Non‐fatal myocardial infarction ‐ All Studies	11	951	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.43, 1.46]
19 Non‐fatal myocardial infarction ‐ Active vs Control	6		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
19.1 Active vs Control	6	509	Risk Ratio (M‐H, Random, 95% CI)	0.65 [0.32, 1.31]
20 Non‐fatal myocardial infarction ‐ Active vs Active	5	442	Risk Ratio (M‐H, Random, 95% CI)	1.46 [0.43, 4.88]
21 Deep vein thrombosis (DVT)	7		Risk Ratio (M‐H, Random, 95% CI)	Subtotals only
21.1 Active vs Control	7	645	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.31, 2.23]
22 Hospital length of stay (LOS) ‐ Active vs Control	10		Mean Difference (IV, Random, 95% CI)	Subtotals only
22.1 Active vs Control	10	772	Mean Difference (IV, Random, 95% CI)	‐1.38 [‐2.79, 0.03]
23 Hospital length of stay (LOS) ‐ Active vs Active	1	196	Mean Difference (IV, Random, 95% CI)	2.8 [‐2.11, 7.71]
23.1 Active vs Active	1	196	Mean Difference (IV, Random, 95% CI)	2.8 [‐2.11, 7.71]

Methods	Patients were randomised to receive an autologous reinfusion drain or a standard suction drain using the computer program MINIM. The method used to conceal treatment allocation was not described.
Participants	104 consecutive patients undergoing primary cemented total knee arthroplasty were randomised to one of two groups: Group 1 (Autotransfusion group): n=52 Group 2 (Control group): n=52 NB: Of the 104 randomised patients 43 were male and 61 were female. The mean age of randomised subjects was 68.5 years.
Interventions	Group 1: Autotransfusion group (Bellovac autotransfusion system) had one deep drain inserted at the end of the operation. The drain was opened in the recovery room 20 minutes after the tourniquet was released. If blood collected in the reinfusion drain was more than 150mls it was transfused back into the patient unwashed and a new bag was then attached to the drain. The process was repeated if the amount of blood collected again exceeded 150mls. Group 2: Control group (Redivac standard suction drain) had their collected blood discarded.
Outcomes	Outcomes reported: number of patients transfused allogeneic blood, amount of allogeneic blood transfused, blood loss, hospital length of stay, Hb & Hct levels, wound problems, knee range of motion.
Notes	Transfusion threshold: allogeneic blood transfusion was given if the haemoglobin level was less than 9.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	The software program MINIM was used to randomise patients to intervention or control.
Allocation concealment (selection bias)	Unclear risk	The method used to conceal treatment allocation was not described.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Concealment of treatment allocation was by use of sealed envelopes. Method of generating allocation sequences was not described.
Participants	90 patients undergoing primary total knee arthroplasty were randomised to one of three groups: Group 1 (No drain group): n=30, M/F=11/13, mean age (95% CI) = 70 (67‐74) years Group 2 (Autotransfusion group): n=30, M/F=4/20, mean age (95% CI) = 71 (69‐74) years Group 3 (Control group): n=30, M/F=9/16, mean age (95% CI) = 72 (69‐75) years
Interventions	Group 1: No drain was used. Group 2: Solcotrans autotransfusion system collected blood for 6 hours or until the unit was full. Acid citrate dextrose‐anticoagulant (ACD‐A) was not added to the collection unit. Continuous suction was applied at 20cm H2O. Drains were maintained for 24 hours post‐operatively. Group 3: A standard disposable closed suction drainage system (Redon) was used with two standard drains maintained for 24 hours post‐operatively.
Outcomes	*Outcomes reported:* number of patients transfused allogeneic blood, blood loss, hospital length of stay, Hb and Hct levels.
Notes	Transfusion threshold: allogeneic blood transfusion was given if the haemoglobin level was less than 9.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Randomisation was carried out with sealed envelopes, opened just before closure of the wound. Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Sealed envelopes were used to conceal treatment allocation.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Patients undergoing bi‐ or tri‐compartmental total knee arthroplasties with a diagnosis of primary osteoarthritis were included in the study.
Participants	32 patients undergoing total knee arthroplasty were randomised to one of two groups: Group 1 (Autotranfusion group): n=16, M/F=0/16, mean (sd) age = 66.9 (9.1) years Group 2 (Control group): n=2/14, M/F=2/14, mean (sd) age = 66.2 (7.1) years
Interventions	Group 1: Autotransfusion group (ConstaVac CBCII autotransfusion system) had wound drainged connected at the end of the operation. The drain fluid was collected during the first 6 hours. Collected blood was transfused at the end of the 6th hour. Reinfusion was performed using a standard 40um blood filter between the collection bag and the intravenous site. After the 6 hours any blood collected from the reinfusion drain was discarded. Group 2: Control group received standard care without autotransfusion.
Outcomes	Outcomes reported: number of patients transfused allogeneic blood, blood loss, hospital length of stay, adverse events.
Notes	Transfusion threshold: allogeneic blood transfusion was given if the haemoglobin level was less than 9.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was not described.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Between May 2005 and December 2005, 178 patients were entered into the study. All patients over 55 years with osteoarthritis and/or inflammatory arthritis of the knee, and awaiting total knee replacement (TKR), were considered for the study. In a pre‐assessment clinic patients were randomly assigned into two groups by sealed envelopes.
Participants	178 patients undergoing total knee replacement were randomised to one of two groups: Group 1 (Autologous retransfusion group): n=92, M/F=43/49, mean (range) age = 70.3 (55.2‐88.5) years Group 2 (Control group): n=86, M/F=39/47, mean (range) age = 70.4 (57.9‐87.1) years
Interventions	Group 1: Autologous retransfusion group (Bellovac ABT autotransfusion sytem) had the blood collection suction bellows connected to an autologous transfusion bag with a 200mm filter and a one‐way valve. The transfusion bag was connected to a transfusion set with a 40um filter. The drain was opened 20 minutes after tourniquet release. The shed blood was returned to the patient after collecting up to 500mls and no later than 6 hours after surgery. A maximum of 1200mls was re‐transfused. Group 2: Control group (standard vacuum drain) had blood collected in the vacuum drains discarded.
Outcomes	Outcomes reported: number of patients transfused allogeneic blood, hospital length of stay, adverse events.
Notes	Transfusion threshold: allogeneic blood transfusion was given if the haemoglobin level was less than 8.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Sealed envelopes were used to conceal treatment allocation.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Between June 1988 and August 1989, 103 patients who gave informed consent to participate in the study underwent cardiopulmonary bypass. Of the initial 103 patients, 71 were excluded from the study. Method of randomisation and allocation concealment was not described.
Participants	32 patients undergoing cardiac surgery requiring cardiopulmonary bypass were randomised to one of two groups: Group 1 (Autologous retransfusion group): n=16; mean (sd) age = 60 (8.0) years Group 2 (Control group): n=16; mean (sd) age = 61 (8.0) years
Interventions	Group 1: Autologous retransfusion group (Pleur‐evac Autotransfusion System ‐ A‐5005‐ATS) had their mediastinal shed blood collected in a polyvinyl chloride blood bag containing an inline 200um nylon mesh filter by means of a closed system with ‐20cmH2O suction applied. This collection system contained no anticoagulant and none was added. Mediastinal shed blood was transfused without washing by detaching the autotransfusion replacement bag and reinfusing the blood through a standard 40um screen blood filter (Pall SQ40S) via a peripheral intravenous line. Group 2: Control group received either autologous packed cells if available or allogeneic packed red blood cells (standard citrate‐phosphate‐dextrose ADSOL‐preserved cross‐matched packed RBCs units stored at 4 degrees celsius for up to 42 days).
Outcomes	*Outcomes reported:* amount of allogeneic blood transfused, amount of autologous blood transfused, number of patients transfused autologous and/or allogeneic blood, complications, bleeding times, plasma protein variables, post transfusion febrile reactions.
Notes	*Transfusion threshold:* the decision to transfuse a patient post‐operatively was made by the clinician who was responsible for the patient's post‐operative care, and who was not involved in the study. The clinical criteria used to determine the need for transfusion consisted of the following: systolic BP less than 80mmHg; mean arterial pressure less than 50mmHg; central venous pressure (CVP) less than 5mmHg; pulmonary capillary wedge pressure (PCWP) less than 5mmHg; cardiac index (CI) less than 2.0L/min/m2; evidence of inadequate end‐organ perfusion (ie: urine output less than 20ml/hr), or anaemia (Hct less than 25%). Any patient who bled more than 400ml in the first 4 hours post‐operatively and who met any of these criteria underwent transfusion.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to concealment treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	The study was conducted between October 15, 1991 through to January 1, 1993. The patients included 125 women and 107 men who were 20‐89 years of age (mean age = 72 years). All patients were advised to donate blood pre‐operatively. The 156 patients (67%) who were scheduled to have a primary procedure were advised to donate 2 units of autologous blood, and the 76 patients (33%) who were scheduled to have a revision procedure were a advised to donate 4 units of autologous blood.
Participants	232 patients undergoing total hip arthroplasties were randomly assigned to one of two groups: Group 1 (Autotransfusion group): n=103 Group 2 (Control group): n=129 NB: Demographic data were not reported.
Interventions	Group 1: Autotransfusion group (Autovac Post‐operative Orthopaedic Autotransfusion Canister) had blood loss collected for 4 hours post‐operatively. The autotransfusion canister was injected with 40mls of acid‐citrate‐dextrose anticoagulant (ACD‐A) before activation. The autotransfusion canister was connected to wall suction with use of an Autovac Autotranfusion Regulator that limited maximum collection pressure to 100mmHg. If at least 300mls of blood was collected within 4 hours, the unwashed blood was reinfused through a microaggregate filter; if less than 300mls of blood was collected, the blood was discarded. Any blood that had not been reinfused within 6 hours after the beginning of collection was discarded. Group 2: Control group had a closed suction drainage system used (Hemovac system).
Outcomes	Outcomes reported: number of patients transfused allogeneic and./or autologous blood, blood loss, Hb levels.
Notes	Transfusion threshold: Transfusion protocol not reported.   All revision patients were exposed to cell salvage intra‐operatively.   85% of Group 1 patients pre‐deposited blood pre‐operatively (PAD).   77% of Group 2 patients pre‐deposited blood pre‐operatively (PAD).
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Patients were randomly assigned on the basis of their hospital record number.
Allocation concealment (selection bias)	High risk	Inadequate allocation concealment.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Study was conducted between January 1993 and May 1993. Consecutive patients underwent elective or urgent coronary artery bypass surgery. All procedures were performed by the same cardiac surgeon. Method of randomisation and allocation concealment was not described.
Participants	75 consecutive patients undergoing coronary artery bypass graft surgery were randomised into one of two groups: Group 1 (Autotransfusion group): n=42; mean (sd) age = 60 (7.0) years Group 2 (Control group): n=33; mean (sd) age = 59 (8.0) years
Interventions	Group 1: Autotransfusion group received autotransfusion of shed mediastinal blood using the cardiotomy reservoir, after the completion of the coronary artery bypass grafting (CABG). As soon as the chest was closed, the mediastinal tubes were attached to the inlet port of the cardiotomy reservoir, which allows the chest tube drainage to pass through a 20 micron filter. The filtered blood was collected in the bottom of the cardiotomy reservoir, ready for reinfusion. The vacuum port was attached to wall suction apparatus and negative pressure was instituted at 20cm H2O. The chest drains were milked every 30 minutes. The collected blood was reinfused using a standard infusion pump. The hourly volume of mediastinal drainage was measured and the infusion pump adjusted to deliver this amount of blood over the next hour. Reinfusion was continued until the drainage was less than or equal to 50ml per hour for two consecutive hours. Group 2: Control group received standard chest drainage.
Outcomes	Outcomes reported: amount of blood collected by the cell saver, amount of blood re‐transfused from the cell saver, amount of allogeneic blood transfused, number of patients transfused allogeneic blood, complications, wound infection, re‐operation for bleeding, hospital length of stay, fever, mortality.
Notes	Transfusion threshold: allogeneic packed cells were transfused intra‐operatively or post‐operatively when the haematocrit fell below 30%.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Between June 2002 and May 2004, 60 patients undergoing unilateral total knee arthroplasty (TKA) were enrolled in this prospective randomised trial. Randomisation was by sealed opaque envelopes which were mixed by independent personnel and consecutively assigned a case number from 1 to 60. All surgeries were performed by specialists of the joint and reconstruction team using an identical surgical approach and technique. Near the end of the operation the corresponding envelope was opened and the surgeon was informed at the time of drain insertion to achieve a single‐blind effect.
Participants	60 patients undergoing unilateral total knee arthroplasty (TKA) were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=26; M/F=6/20; mean (range) age = 72 (57‐84) years Group 2 (Control group): n=34; M/F=12/22; mean (range) age = 69.4 (55‐78) years
Interventions	Group 1: Autotransfusion group (DONOR system) had their blood reinfused from drains using a 40um blood filter between the collection bag and the intravenous site within 6 hours of surgery. All patients had their drains removed on post‐operative day 2 or 3. The DONOR system is an integrated, closed system designed for the collection and reinfusion of drained wound blood. It consists of an 800ml chlorine‐free, pre‐evacuated collection vessel, a vacuum regulator, and a 40um integrated filter for salvaged blood. Group 2: Control group received no post‐operative autotransfusion.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, number of patients transfused allogeneic blood, febrile complications, adverse events, blood loss.
Notes	Transfusion threshold: allogeneic blood transfusion was given if the haemoglobin level was less than 9.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Method used to generate allocation sequences was adequate.
Allocation concealment (selection bias)	High risk	Sealed opaque envelopes were used to conceal treatment allocation.
Blinding (performance bias and detection bias)   All outcomes	High risk	Single‐blind design.

Methods	Patients undergoing elective abdominal aortic aneurysm (AAA) repair or aortofemoral bypass (AFB) for occlusive disease were eligible for entrance into the study. Randomisation was carried out in blocks of 10 stratified for AAA repair or AFB. Patients were randomised by means of drawing sealed envelopes that contained prescriptions for either intra‐operative autotransfusion (IAT) or control therapy. The study was unblinded.
Participants	100 patients undergoing aortic surgery were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=50; M/F=41/9; mean (sd) age = 63 (11.0) years Group 2 (Control group): n=50; M/F=43/7; mean (sd) age = 65 (9.0) years
Interventions	Group1: Intra‐operative autotransfusion group had their blood processed by either a Cell Saver 3 Plus or Cell Saver 5 device. Both systems consist of polyvinyl aspiration tubing with a separate channel for introducing small amounts of heparised saline solution to anticoagulate aspirated blood, a plastic cardiotomy reservoir with microaggregate filter, a continuous flow, disposable washing bowl driven by a centrifuge, and a transfusion setup that consists of a plastic transfer pack passed to the anaesthesiologist for administration. The maximum allowable amount of IAT‐PRBCs administered to a single patient was 1500mls. Group 2: Control group did not receive autotransfusion.
Outcomes	Outcomes reported: number of patients transfused allogeneic blood, amount of allogeneic blood transfused, blood loss, hospital length of stay, ICU length of stay, adverse events.
Notes	Transfusion threshold: patients were transfused allogeneic RBCs intra‐operatively if the haemoglobin level was less than 10.0g/dL. Post‐operatively patients were transfused allogeneic RBCs if the haemoglobin level was less than 8.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Treatment allocation was inadequately concealed.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Patients undergoing either coronary artery bypass grafting, valve replacement/repair operations or a combination of the two were randomised pre‐operatively into two groups using a binary random number table. Method used to conceal treatment allocation was not described.
Participants	112 patients undergoing cardiac surgery were randomised to one of two groups: Group 1 (Autotransfusion group): n=56; M/F=36/20; mean (sd) age = 67.4 (9.0) years Group 2 (Control group): n=56; M/F=41/15; mean (sd) age = 65.3 (10.5) years
Interventions	Group 1: Autotransfusion group were transfused with washed post‐operative drained blood processed by a Fresenius Continuous Autotransfusion System (C.A.T.S). Group 2: Control group received usual care management without autotransfusion.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, number of patients transfused allogeneic blood, mortality, re‐operation for bleeding, blood loss, coagulopathy, Hb levels.
Notes	Transfusion threshold: patients were transfused allogeneic RBCs intra‐operatively if the haemoglobin level was less than 7.0g/dL. Post‐operatively patients were transfused allogeneic RBCs if the haemoglobin level was less than 10.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was not described.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	The study was conducted between September 2003 to October 2004. Patients admitted for elective or sub‐acute coronary bypass surgery without the use of the cardiopulmonary bypass (CPB) machine were included. If the CPB machine became necessary during the operation the patient was excluded. Patients were randomised to intervention or control by means of sealed opaque envelopes numbered in sequence.
Participants	60 patients undergoing 'off‐pump' coronary artery bypass surgery were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=30; M/F=11/19; mean (IQR) age = 77 (74‐79) years Group 2 (Control group): n=30; M/F=14/16; mean (IQR) age = 76 (70‐79) years
Interventions	Group 1: Autortransfusion group (Medtronic Autolog system) received intra‐operative autotransfusion. Immediately after surgery the suctioned blood was processed by the cell saver device and autotransfused before the patient was transferred to the intensive care unit (ICU). Group 2: Control group had their intra‐operative suctioned blood discarded. NB: The cell saver reservoir with a 40um filter was used in the ICU for mediastinal drained blood collection and for post‐operative autotransfusion in both groups. A maximum of 12 hours of post‐operative unwashed autotransfusion from the drains was routine practice.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, number of patients transfused allogeneic blood, blood loss, Hb levels, adverse events, costs.
Notes	Transfusion threshold: patients were transfused allogeneic RBCs if the haematocrit level was less than 30%.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Method used to conceal treatment allocation was inadequate.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Fifty patients having aortic surgery for either abdominal aortic aneurysm or aorto‐iliac occlusive disease were selected for study. Method of randomisation and allocation concealment was not described.
Participants	50 patients undergoing aortic surgery were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=25; M/F=21/4; mean (sd) age = 68 (8.0) years Group 2 (No intraoperative salvage group): n=25; M/F=22/3; mean (sd) age = 70 (8.0) years
Interventions	Group 1: Autotransfusion group (Sorenson autotransfusion system) had their blood loss from the surgical site suctioned into the Sorenson receptacle device and then retransfused at the time of surgery. Additional blood loss which was not able to be collected was replaced according to haematocrit levels, 3.5% polygeline being given if the haematocrit was above 30% and allogeneic blood if the haematocrit was below 30%. The collected blood was anticoagulated with an acid citrate dextrose solution and administered via a burette at a rate of 70ml for every 430ml of autologous blood collected. The scavenged blood was collected in a 1900ml sterile disposable Sorenson receptal ATS trauma liner contained within the rigid reusable receptal canister. When approximately 1 litre of autologous blood had been scavenged the liner was removed and this blood then administered to the patients after being filtered through a Pall 40um filter. Group 2: intraoperative blood loss was replaced with either 3.5% polygeline or allogeneic blood according to the measured Hct. If the Hct was above 30% then polygeline was used; if the Hct was below 30% then allogeneic blood was administered.
Outcomes	Outcomes reported: amount of blood re‐transfused from the cell saver, amount of allogeneic blood transfused, number of patients transfused allogeneic blood, mortality, re‐operation for bleeding, haemodialysis, blood loss, coagulopathy, Hb levels, organisms cultured from autologous vs allogeneic blood.
Notes	Transfusion threshold: patients received allogeneic RBC transfusion if the haematocrit level fell below 30%.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	The efficacy of four different blood conservation techniques in decreasing allogeneic blood transfusion in different cardiac operations were studied in 100 patients undergoing myocardial revascularisation. Method of randomisation and allocation concealment was not described.
Participants	100 patients undergoing myocardial revascularisation were randomly assigned to one of four groups: Group 1 (Control group): n=25; mean (sd) age = 56.0 (6.6) years Group 2 (Intra‐operative autotransfusion group): n=25; mean (sd) age = 54.1 (6.8) years Group 3 (Acute normovolaemic haemodilution + intra‐operative autotransfusion group): n=25; mean (sd) age = 55.0 (9.4) years Group 4 (Acute normovolaemic haemodilution + intra‐ and post‐operative autotransfusion group): n=25; mean (sd) age = 55.7 (6.3) years
Interventions	Group 1: patients received unprocessed oxygenator blood after the termination of extracorporeal circulation (ECC). Group 2: the blood remaining in the oxygenator after ECC was processed to packed cells with a cell separator (Haemonetics Cell Saver) and re‐transfused until the end of the operation. Group 3: after the induction of anaesthesia and before the start of the operation, isovolumetric hemodilution (harvesting of 10ml/kg autologous blood) was performed under electrocardiographic and haemodynamic control. The blood loss was replaced with hydroxyethyl starch. After termination of ECC, the blood remaining in the oxygenator was processed by a cell separator. The preoperatively drawn blood and the packed cells were retransfused before the end of the operation. Group 4: patients in Group 4 were managed as in Group 3. In addition, the shed mediastinal blood was retransfused in the Intensive Care Unit (ICU). The cardiotomy reservoir of the heart lung machine was used to collect this blood. The drained blood was retransfused intermittently according to the circulatory state of the patient and when at least 250ml of blood had been collected in the reservoir. The last retransfusion was performed 6 hours post‐operatively.
Outcomes	Outcomes reported: amount of blood re‐transfused from the cell saver, amount of allogeneic blood transfused, number of patients transfused allogeneic blood, complications, mortality, ICU length of stay, blood loss, re‐exploration for bleeding, operation time, haematological variables, Hct levels.
Notes	Transfusion threshold: in all patients signs of hypovolaemia and haematocrit values below 30% were indications for allogeneic blood transfusion.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

PERMALINK

Cell salvage for minimising perioperative allogeneic blood transfusion

Paul A Carless

David A Henry

Annette J Moxey

Dianne O'Connell

Tamara Brown

Dean A Fergusson

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

Plain language summary

Background

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

Data extraction and management

Assessment of risk of bias in included studies

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Results

Description of studies

Methods of cell savage

Types of cell salvage devices

Transfusion 'triggers' or thresholds

Risk of bias in included studies

1.

2.

Adequate sequence generation

Allocation concealment

Blinding

Effects of interventions

Aggregated analysis

Transfusion protocol

Type of surgery

Type of cell salvage ‐ washed versus unwashed

Timing of cell salvage

Volume of blood transfused

Type of cell salvage ‐ washed versus unwashed ‐ type of surgery

Blood loss

Active versus control

Transfusion protocol

Type of surgery

Type of cell salvage ‐ washed versus unwashed

Timing of cell salvage

Volume of blood transfused

Volume of blood transfused ‐ transfusion protocol

Volume of blood transfused ‐ type of surgery

Blood loss

Active versus active

Transfusion protocol

Type of surgery

Type of cell salvage ‐ washed versus unwashed

Timing of cell salvage

Volume of blood transfused

Blood loss

Adverse events and other outcomes

Mortality

Re‐operation for bleeding

Any infection

Wound complications

Any thrombosis

Stroke

Methods	Patients undergoing primary unilateral total knee arthroplasty over a consecutive 30‐day period were studied to assess the efficacy and financial cost of post‐operative reperfusion of drained blood.
Participants	49 patients undergoing primary unilateral total knee arthroplasty were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=32; M/F=11/21; mean (range) age = 69 (49‐83) years Group 2 (Control group): n=17; M/F=4/13; mean (range) age = 72 (62‐91) years
Interventions	Group 1: Autotranfusion group (CellTrans system) had their drained blood filtered through a 40um filter before being reinfused. Before closure of the wound two drainage tubes were inserted. The tubes were connected through a Y‐connector to the CellTrans assembly which contains two transfusion bags. The clamps remained closed for 20 minutes after the wound had been closed off. The drainage was started in the recovery room and collected for 6 hours or until 600mls of blood had accumulated at which point reinfusion took place. Collection up to a maximum of 12 hours ‐ thereafter the blood collected in the drains was discarded. Group 2: Control group received a standard vacuum drain (Redivac high vacuum drainage system). Drains were removed routinely at 48 hours. Contents were discarded.
Outcomes	Outcomes reported: number of patients transfused allogeneic blood, amount of allogeneic blood transfused, Hb levels, cost.
Notes	Transfusion threshold: the trigger for transfusing allogeneic blood was a post‐operative haemoglobin level of less than 9.0g/dL or clinical symptoms of anaemia.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Patients with severe hip arthrosis undergoing total hip arthroplasty were studied to evaluate the efficacy of different peri‐operative blood saving techniques to reduce allogeneic blood transfusion. Method of randomisation and allocation concealment was not described.
Participants	45 patients undergoing total hip arthroplasty were randomly allocated to one of three groups: Group 1 (Control group): n=15 Group 2 (Autotransfusion group): n=15 Group 3 (Autologous predonation + autotransfusion group): n=15 NB: Demographic data were not reported.
Interventions	Group 1 (Control group): blood loss was replaced with heterologous erythrocyte concentrate (SAGM‐ERC) and 3% dextran 60 in a ratio of 1:1. If necessary, additional SAGM‐ERC was transfused to correct erythrocyte volume fraction (EVF)>27%. Group 2 (Autotransfusion group): blood loss was replaced with 3% dextran and by autotransfusion of washed and haemconcentrated blood salvaged by intraoperative suction and from wound drains up to 4 hours postoperatively. As in Group 1, additional SAGM‐ERC was transfused to correct erythrocyte volume fraction (EVF)>27%. Group 3 (Autologous predonation + Cell Saver group): blood loss was replaced with 3% dextran and by autotransfusion of washed and haemconcentrated blood salvaged by intraoperative suction and from wound drains up to 4 hours postoperatively. Predonated autologous SAGM‐ERC was used instead of heterologous blood to maintain erythrocyte volume fraction (EVF)>27%. In 2‐3 sessions within 6 weeks prior to the operation, 2 to 3 units of SAGM‐ERC had been withdrawn. If necessary, heterologous SAGM‐ERC was used if transfusion of all predonated autologous blood failed to maintain EVF>27%. Autotransfusion technique: Haemonetic Cell Saver 4, Althin model AT 1000, or Shiley/Dideco STAT were used. Blood was retrieved from the operation site by suction through a double lumen catheter and was then anticoagulated with heparin (30,000 IU heparin in 1000ml of physiological saline). The blood was collected into a reservoir where a macrofilter removed debris. Thereafter, the blood was pumped into a spinning centrifuge bowl (125ml of blood) and washed with 1500ml of physiological saline. The erythrocytes were concentrated to an EVF of about 50‐60% and pumped into an infusion bag. The effluent containing platelets, free haemoglobin and anticoagulants was disposed.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, amount of autologous blood transfused, number of patients transfused allogeneic blood, complications, adverse events.
Notes	Transfusion threshold: patients were transfused allogeneic blood to maintain the erythrocyte volume fraction (EVF) >27%.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Randomised trial to study the quality and effect of blood produced by the cell saver compared with allogeneic blood in primary total hip arthroplasty.
Participants	40 patients undergoing primary total hip arthroplasty were randomly allocated to one of two groups: Group 1 (Autotransfusion group): n=20; M/F=9/11; mean (range) age = 68 (59‐89) years Group 2 (Control group): n=20; M/F=8/12; mean (range) age = 74 (48‐89) years
Interventions	Group 1: Autotransfusion group received autologous blood processed intra‐operatively by a cell saver device (Electromedic Autotrans AT1000 autotransfusion system). Blood was retrieved from the operative field with a double lumen suction catheter. The blood was immediately anticoagulated with sodium citrate. Larger debris was removed by a 240um filter in the cardiotomy reservoir. The filtered blood was pumped into a bowl centrifuge and washed with 1500mls of saline. The supernatant was discarded. The erythrocyte concentrate was pumped into a reinfusion bag and then reinfused into the patient. Group 2: Control group received allogeneic blood and no autotransfusion. NB: Thromboprophylaxis was given to all patients using dextran 70, 6% in saline (Macrodex) 500mls during the operation, another 500mls during the remainder of the operation day, and 500mls on post‐operative days 1,3 and 5.
Outcomes	Outcomes reported: amount of allogeneic units transfused, number of patients receiving allogeneic blood, complications, blood loss, haematological variables.
Notes	Transfusion threshold: a transfusion of allogeneic blood was given if the haemoglobin was less than 8.5g/dL or if there were symptoms of anaemia.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Method used to conceal treatment allocation was inadequate.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Prospective randomised trial to investigate the safety and efficacy of post‐operative autologous blood transfusion carried out in two matched groups of twenty patients undergoing elective coronary artery bypass surgery. Method of randomisation and allocation concealment were not described.
Participants	40 patients (33 males and 7 females) undergoing elective coronary artery bypass surgery were randomised to one of two groups: Group 1 (Autologous blood transfusion group): n=20 Group 2 (Control group): n=20 Mean (range) age for both groups = 55.75 (33‐69) years.
Interventions	Group 1: received postoperative autologous blood transfusion (AT) using the Shiley hardshell venous reservoir. At the end of the operation in theatre, the chest drains were connected to the Shiley hardshell venous reservoir using the Shiley drainage set. After the system was primed and specimens obtained for haematological, biochemical, and bacteriological analyses, transfusion of the shed blood was commenced, the rate depending on the amount of drainage, reinfusing the previous hours blood loss over the subsequent hour. At the end of 6 hours the AT was discontinued and further specimens were obtained. Group 2: patients were managed in the same manner without the use of autologous blood transfusion.
Outcomes	Outcomes reported: amount of blood re‐transfused from the cell saver, amount of allogeneic blood transfused, number of patients transfused allogeneic blood, hospital length of stay, mortality, blood loss, haematological variables, adverse events.
Notes	Transfusion threshold: allogeneic blood transfusion was used only when the haematocrit fell below 25%, haemoglobin below 9.0g/dL or the blood loss exceeded 500mls in the first 4 hours.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	To determine if autotransfusion of unwashed shed mediastinal blood led to a reduction in post‐operative banked blood requirements a prospective randomised study of 82 patients undergoing myocardial revascularisation was conducted in 1994 at the Cardiovascular Center of Parma. Method of randomisation and allocation concealment were not described [Italian article].
Participants	82 patients undergoing myocardial revascularisation were randomised to one of two groups: Group 1 (Autotransfusion group): n=41; M/F=37/4; mean (sd) age = 60.2 (9.3) years Group 2 (No Autotransfusion group): n=41; M/F=33/8; mean (sd) age = 62.7 (8.9) years
Interventions	Group 1: Autotransfusion group (Atrium 2550 autotransfusion system) had their drained blood processed using the autotransfusion system. Group 2: Control group did not receive autotransfusion.
Outcomes	Outcomes reported: amount of blood collected by the cell saver, number of patients transfused allogeneic blood, amount of allogeneic blood transfused, blood loss, mortality.
Notes	Transfusion threshold: allogeneic blood transfusion was given during surgery if the haemoglobin level fell below 7.5g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was unclear.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Consecutive patients undergoing total hip or total knee replacement procedures between January 1989 and April 1989 were included in this study. A computer‐generated random number list was used to pre‐operatively assign patients to intervention or control groups. Method used to conceal treatment allocation was not described.
Participants	239 consecutive patients undergoing total knee replacement procedures were randomly assigned to one of two groups: Group 1 (Autotransfusion group): n=124; M/F=59/65; mean age = 65 years Group 2 (Control group): n=115; M/F=46/69; mean age = 69 years
Interventions	Group 1: Autotranfusion group (Solcotrans autotransfusion system) had their wounds drained into post‐operative blood salvage canisters. There was a 6 hour total time limit for collection and reinfusion of blood. Because 40ml of citrate ACD‐A was entered in each Solcotrans canister prior to use, a minimum of 320mls of blood and citrate volume was necessary before reinfusion to prevent citrate toxicity. If wound drainage was slow and an adequate volume had not been collected before the 6‐hour time limit, the canister and blood were discarded, and a standard collection canister was attached to the drainage tube for the duration. If wound drainage was rapid, the canister was allowed to fill completely (500mls volume). The blood was then infused at an appropriate rate as long as the 6‐hour pre‐canister limit was not exceeded. Another Solcotrans canister could then be attached, beginning a new 6‐hour time interval. Intra‐operative blood salvage was not used. Group 2: Control group had their wounds drained into standard 400ml suction canisters. Autotransfusion was not performed.
Outcomes	Outcomes reported: amount of blood re‐transfused from the cell saver, number of patients transfused allogeneic blood, adverse events.
Notes	Transfusion threshold: All patients whose post‐operative haemoglobin value was less than 9.0g/dL were transfused allogeneic blood. The decision to transfuse patients with haemoglobin values greater than 9.0g/dL was made by the internist on the basis of each patient's medical condition.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer‐generated random number list was used to pre‐operatively assign patients to either intervention or control.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Between March 2004 and June 2004, all patients admitted for elective or urgent first‐time coronary artery bypass grafting (CABG) were enrolled in the study.
Participants	50 patients undergoing 'off‐pump' first‐time coronary artery bypass grafting (CABG) were randomised to one of two groups: Group 1 (Autotransfusion group): n=24; M/F=21/3; mean (sd) age = 58.2 (8.7) years Group 2 (Control group): n=25; M/F=21/4; mean (sd) age = 61.9 (10.0) years NB: One patient in the autotransfusion group (intervention group) was excluded from the final analysis due to conversion to cardiopulmonary bypass ('on'pump').
Interventions	Group 1: Autotransfusion group (Dideco autotransfusion system) had all intra‐operative shed blood from the time of incision till skin closure collected by means of a single lumen high‐pressure suction cannula flushed with heparinised saline and was collected in the reservoir of the cell saver device. The collected blood was then subjected to washing and centrifugation. The processed red blood cells were collected in sterile blood bags and were made available to the anaesthetic staff for autotransfusion. Group 2: Control group had their intra‐operative shed blood discarded.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, volume of blood re‐transfused from the cell saver, blood loss, adverse events.
Notes	Transfusion threshold: the indication for allogeneic blood transfusion in the intra‐operative period was a haemoglobin level less than 9.0g/dL or a haematocrit level less than 27%. In the autotransfusion group, all the processed red blood cells collected during surgery were re‐transfused as required. Banked allogeneic blood was used only if the haemoglobin level remained less than 9.0g/dL despite autotransfusion.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Sealed envelopes were used to conceal treatment allocation.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	The efficacy of autologous shed blood in reducing allogeneic blood transfusion was evaluated at four medical centres in a prospective randomised study. Method of randomisation and allocation concealment was not described.
Participants	128 patients undergoing total hip arthroplasty, total knee arthroplasty, or spine fusion were randomly allocated to one of three groups: Group 1 (Autotransfusion ‐ Orth‐Evac group): n=44; M/F=18/26; mean (range) age = 67.9 (41‐82) years Group 2 (Autotransfusion ‐ Solcotrans group): n=40; M/F=20/20; mean (range) age = 66.3 (54‐82) years Group 3 (Control group): n=44; M/F=23/21; mean age = 62.5 years.
Interventions	Group 1: Autotransfusion group received autologous shed blood reinfusion collected from wound drainage by an Orth‐evac device. Group 2: Autotransfusion group received autologous shed blood reinfusion collected from wound drainage by a Solcotrans device. Group 3: Control group received either autologous predonated blood or allogeneic banked blood. In control patients a standard wound drainage system (Hemovac) was used, and these patients received liquid‐preserved autologous predonated blood or allogeneic blood filtered with a standard 170 micron screen filter. NB: Patients randomised to the autologous shed blood groups (Group 1 and Group 2) were randomly assigned to one of two infusion filters (Pall 40 micron screen filter or Pall RC100 polyester filter) for the transfusion phase of the study. With the Solcotrans drainage system, 40ml acid citrate detrose (ACD) was used. No anticoagulant was added with the Ortho‐evac drainage system.
Outcomes	Outcomes reported: amount of blood collected by the cell saver, amount of blood re‐transfused from the cell saver, number of patients transfused allogeneic blood, amount of allogeneic blood transfused, adverse events.
Notes	Transfusion threshold: transfusion protocol was not reported.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	Consecutive patients undergoing elective knee arthroplasty at two institutions were enrolled in the study. Method of randomisation and allocation concealment was not described.
Participants	81 patients undergoing elective knee arthroplasty were randomly assigned to one of two groups: Group 1 (Autotransfusion group): n=39; M/F=14/25; mean (sd) age = 69.3 (6.9) years Group 2 (Control group): n=40; M/F=14/26; mean (sd) age = 71.0 (9.0) years
Interventions	Group 1: Autotransfusion group underwent drainage and autotransfusion transfusion using a Solcotrans system. The autologous blood collected into the drainage and transfusion device was transfused if specific transfusion guidelines were met. Patients were transfused the initial unit of Solcotrans blood if 350ml or more had been collected within 3 hours of the patients entry to the recovery room. The 3‐hour collection time provided for collection and transfusion of the blood within the maximum interval of 6 hours. After successful collection and transfusion of the first autologous blood unit, a second autologous blood collection device was attached. For this and subsequent collections, autologous blood was transfused if 150ml or more was collected within 3 hours. When the rate of drainage was less than 250ml of blood within a 3 hour period, a subsequent drainage and transfusion device was not attached. The first Solcotrans device attached to the drain contained 40ml of ACD‐A. Group 2: Control group had their drained blood collected by a Davol suction unit and discarded. The Davol unit was the current standard practice in the two study centres. Patients assigned to the Davol suction group received 1 unit of allogeneic red cells if more than 500ml of blood drained from the surgical site within a 2 hour period. Subsequently, whenever drainage exceeded 500ml within a 2 hour period, 1 unit of allogeneic blood was transfused.
Outcomes	Outcomes reported: amount of allogeneic blood transfused, number of patients transfused allogeneic blood, adverse events, blood loss, coagulation variables, venogram tests.
Notes	Transfusion thresholds: on postoperative Day 2 through to Day 5, the criteria for allogeneic red cell transfusions were identical for both groups. Patients were given one unit of red cell concentrate if their haemoglobin was within the range of 8.0 to 8.9g/dL, two units when the value was from 7.0 to 7.9g/dL, three units when the value was from 6.0 to 6.9g/dL, and four units if the value was from 5.0 to 5.9g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	Unclear risk	Method used to conceal treatment allocation was unclear.
Blinding (performance bias and detection bias)   All outcomes	High risk

Methods	This study was prospectively performed on the cases of a single vascular surgeon operating at two institutions from January 1989 to January 1990. Patients undergoing elective infrarenal abdominal aortic bypass for either occlusive or aneurysmal disease were included in the study sample. Patients were randomised on an alternating basis to either intervention or control. Method of randomisation was not described.
Participants	36 patients undergoing aortobifemoral or aortobi‐iliac bypass for occlusive disease were randomised to one of two groups: Group 1 (Autotransfusion group): n=18 Group 2 (Control group): n=18 NB: Demographic data were not reported.
Interventions	Group 1: Autotransfusion group (Haemonetics Cell Saver) was monitored and operated by a technician‐member of the perfusion team. The Haemonetics Cell Saver delivers washed red blood cells at an average haematocrit level of 55% to 60%. Group 2: Control group did not receive autotransfusion.
Outcomes	Outcomes reported: amount of blood re‐transfused from the cell saver, amount of allogeneic blood transfused, adverse events, hospital length of stay, blood loss, haemoglobin levels.
Notes	Transfusion threshold: after the operation allogeneic red cell transfusions were not given to patients who were haemodynamically stable, and had haemoglobin values greater than 8.0g/dL.
*Risk of bias*
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Method used to generate allocation sequences was not described.
Allocation concealment (selection bias)	High risk	Method used to conceal treatment allocation was inadequate.
Blinding (performance bias and detection bias)   All outcomes	High risk