Comparison between autologous blood transfusion drainage and closed-suction drainage/no drainage in total knee arthroplasty: a meta-analysis

Background Autologous blood transfusion (ABT) drainage system is a new unwashed salvaged blood retransfusion system for total knee replacement (TKA). However, whether to use ABT drainage, closed-suction (CS) drainage or no drainage in TKA surgery remains controversial. This is the first meta-analysis to assess the clinical efficiency, safety and potential advantages regarding the use of ABT drains compared with closed-suction/no drainage. Methods PubMed, Embase, and the Cochrane Library were comprehensively searched in March 2015. Fifteen randomized controlled trials (RCTs) were identified and pooled for statistical analysis. The primary outcome evaluated was homologous blood transfusion rate. The secondary outcomes were post-operative haemoglobin on days 3–5, length of hospital stay and wound infections after TKA surgery. Results The pooled data included 1,721 patients and showed that patients in the ABT drainage group might benefit from lower blood transfusion rates (16.59 % and 37.47 %, OR: 0.28 [0.14, 0.55]; 13.05 % and 16.91 %, OR: 0.73 [0.47,1.13], respectively). Autologous blood transfusion drainage and closed-suction drainage/no drainage have similar clinical efficacy and safety with regard to post-operative haemoglobin on days 3–5, length of hospital stay and wound infections. Conclusions Autologous blood transfusion drainage offers a safe and efficient alternative to CS/no drainage with a lower blood transfusion rate. Future large-volume high-quality RCTs with extensive follow-up will affirm and update this system review. Electronic supplementary material The online version of this article (doi:10.1186/s12891-016-0993-z) contains supplementary material, which is available to authorized users.


Background
Total knee arthroplasty (TKA) is a highly successful standard procedure for patients who suffer serious knee arthralgia, instability and deformity. It is used after nonsurgical treatments are exhausted, especially in advanced knee osteoarthritis [1,2]. However, TKA can result in significant blood loss, reduction in haemoglobin (Hb) and other clinical risks [3,4]. Reports of blood transfusion rates of 39 %-50 % have been published [5][6][7].
Autologous blood transfusion (ABT) drainage system is a new unwashed salvaged blood retransfusion system for primary TKA. However, whether to use ABT drainage, closed-suction (CS) drainage or no drainage in TKA surgery is still controversial. Some studies have found that ABT significantly reduced the need for homologous blood [8,9], but other research has questioned the benefits of this method [10,11] or demonstrated that post-TKA ABT had a limited effect on blood conservation [12,13]. While gaining worldwide acceptance [14] for effectively decreasing hematoma formation [15,16], conventional suction drains have been theoretically thought to decrease postoperative pain, swelling and incidence of infection [17]. However, a closed suction drainage system inevitably increases bleeding because the tamponade effect of a closed undrained wound is eliminated [14].
Until now, no systematic reviews incorporating metaanalyses (SRMA) have found sufficient evidence to recommend ABT drainage or no drainage in primary TKA. This is the first SRMA to systematically compare the clinical results of ABT drainage with closed-suction (CS)/no drainage in patients undergoing TKA. Previous SRMAs comparing ABT drainage versus CS drainage and CS drainage versus no drainage were published as the standard in evidence-based medicine with conflicting results [6,18,19]. Quinn et al. [19] showed that ABT drainage was superior to CS drainage for reducing blood transfusion rate (OR: 0.25 [0.13, 0.47]; P < 0.0001), and length of hospital stay (WMD: −0.25 [−0.48, −0.01]; P = 0.04). However, data extraction errors occurred in two included studies [20,21] when extracting the number of patients requiring homologous blood transfusion for the metaanalysis. Another flaw is that in meta-analysis extracted data without intention-to-treat (ITT) analysis, treatment effectiveness may be exaggerated. The previous meta-analysis also did not evaluate other outcome measures like wound complication and post-operative haemoglobin on days 3-5. The aim of this SRMA was to pool extracted data from available published RCTs to provide a directly substantiated judgment regarding the use of ABT drainage following TKA surgery.

Methods
In accordance with Preferred Reporting Items for Systematic Reviews and Meta-analysis (Additional file 1) [22], we made a prospective protocol of objectives, literature-search strategies, inclusion and exclusion criteria, outcome measurements and methods of statistical analysis before the research began.

Data sources and search strategies
The following databases were searched in March 2015 without restriction to regions and publication types: Pubmed (1950-March 2015), Embase (1974( -March 2015 and Cochrane Library (March 2015 Issue 3) (Additional file 2). The MeSH terms and their combinations searched in [Title/Abstract] was as follows: "total knee replacement" OR "total knee arthroplasty" OR "total knee prosthesis" OR "unicompartmental" OR "unicondylar" OR "arthroplasty, replacement, knee" [MeSH term] AND ("autologous blood transfusion" OR "autotransfusion" OR "blood transfusion, autologous" [MeSH Terms] OR "intraoperative blood salvage" OR "intraoperative blood" OR "postoperative blood salvage" OR "intraoperative blood cell salvage" OR "operative blood salvage" [MeSH Terms]). The reference lists of related reviews and original articles identified for any relevant studies, including randomized controlled trials (RCTs) involving adult humans were reviewed. The search also included the Controlled Trials Register (http://www.controlled-trials.com). Only articles originally written in English or translated into English were considered. When multiple reports describing the same situation were published, the most recent or complete report was used.

Inclusion and exclusion criteria
Two independent researchers (Pan and Yang) identified studies that met the defined inclusion criteria, with disagreements resolved by consensus (Hong and Liu). Inclusion criteria were: (1) the comparison was between ABT drainage and CS/no drainage post TKA; (2) at least one of the quantitative outcomes we determined to evaluate was reported; (3) study design was a RCT; and (4) full text was published in English. Non-original research (e.g. review article, editorials, letter to the editor), case reports, animal experimental studies and duplicated publications were excluded.

Data extraction and analysis
The data from eligible studies were extracted by two researchers (Hong and Pan) independently to minimize errors and reduce potential biases. In cases of disagreement, a consensus was reached by the adjudicating senior authors (Yang and Liu). The extracted data was input into a computerized spreadsheet, including sample size, study design, patient age, gender, preoperative/postoperative Hb levels, number of patients transfused with homologous blood, length of hospital stay and wound infection. The primary outcome was homologous blood transfusion rate. The secondary outcomes were postoperative haemoglobin on days 3-5, length of hospital stay and wound infection.

Quality assessment and data synthesis
The RCTs were graded according to criteria of the Centre for Evidence-Based Medicine in Oxford, UK [23]. The quality of the RCTs; methodology was evaluated by the Cochrane risk of bias tool [24].
The statistical analysis was conducted with Cochrane Collaboration Review Manager 5.3.5 (Cochrane Collaboration, Oxford, UK). Our analyses were based on ITT or modified ITT data. Odds risk (OR) with 95 % confidence intervals (CIs) was calculated for dichotomous data and weighted mean differences (WMD) with 95 % CIs for continuous data. Statistical heterogeneity was assessed by using the chi-square test and I2 statistic. A random-effects model was used when significant heterogeneity was detected between studies without clinical diversity (P < 0.10; I2 > 50 %). Otherwise, a fixed-effect model was performed [24]. In cases with I 2 values greater than 50 % for outcome measures, sensitivity analyses were conducted for  Fig. 1 Flow diagram of studies identified, included and excluded heterogeneity. When overall results and conclusions are not affected by the different decisions that could be made during the review process, the results of the review can be regarded with a higher degree of certainty [24]. Funnel plots were used to identify potential publication bias.

Characteristics of included studies
The demographic characteristics of the 15 studies are presented in Table 1.

Wound infection
Four studies [11,29,31,35 ] (Fig. 7). Figure 8 shows a funnel plot of the included studies that reported homologous blood transfusion rates. All studies lie inside the 95 % CIs except two studies, with an asymmetric distribution around the vertical indicating presence of obvious publication bias. This obvious publication bias is for the beneficial effect of lowering blood transfusion rate.

Discussion
This SRMA of 15 studies including 1,721 patients comparing the clinical efficacy and safety of ABT drainage and closed-suction/no drainage showed significant statistical  With recent techniques, ABT drainage post TKA manifests the attractive concept of retransfusing collected drainage blood and continues to be a controversial issue in TKA surgery. Some studies have published considerable doubt with respect to its advantages [16,36]. Despite the advantageous results, including reduced homologous blood transfusion rates shown in some studies [29,37,38], some authors have suggested insufficient efficiency for ABT [10,39]. In spite of the paucity of consistent evidence, for many years the majority of orthopaedic procedures were followed by the use of ABT drainage post TKA to reduce the blood transfusion rate. However, the present systematic review and metaanalysis demonstrate a significant beneficial effect of ABT drainage in reducing the blood transfusion rate. The result of this meta-analysis showed no significant difference in post-operative haemoglobin on days 3-5. As those patients who received allogenic blood were not excluded from this analysis and there was a higher rate of allogenic blood transfusion in the closed suction drainage group compared with the  ABT drainage group, it cannot be ascertained whether this is owing to a failure in ABT drainage to produce a beneficial effect on post-operative haemoglobin or to the positive nature of allogenic blood on haemoglobin levels. With the application of any new medical device, the safety of the patients is always of paramount importance. Acting as a channel for the introduction of infection, drainage may increase infection risk by impairing host resistance and allowing pathogens access to a sterile field [16,17,40]. The demands on nursing care and physiotherapy are increased to accommodate the presence of drainage. In orthopaedic surgery, wound infection is a devastating complication. However, the pooled data of postoperative outcomes indicated that the ABT drainage equipment was safe and effective for TKA. There was no significant difference in wound complication and length of hospital stay. This finding indicates that ABTD is as safe and efficient as CS/no drainage.
Some possible limitations of this meta-analysis and future research directions should be noted. The primary limitation is that the selected RCTs in this meta-analysis were moderate-quality studies with small sample sizes. With fewer included studies in the outcome analysis, the statistical heterogeneity assessments, including I 2 text, were able to make false negative errors. Future systematic reviews should evaluate the indications from literature from sufficient, larger multi-centre clinical studies. In addition, this meta-analysis limited the included articles to those published in English. There might be selection bias in language. Finally, no long-term outcome measures were assessed, which is most pertinent to patients [41]. Therefore, other outcomes like range of movement, deep joint infection and component loosening, which are manifested after many years, should be considered.

Conclusions
To our knowledge, this is the first SRMA to systematically compare the results of ABT drainage with closed-suction drainage/no drainage in patients undergoing TKA. The pooled results demonstrated that ABT drainage was more efficacious than CS drainage in clinically reducing blood transfusion rate. This meta-analysis also indicated that ABT drainage and closed-suction drainage/no drainage had similar clinical efficacy and safety with regard to post-operative haemoglobin on days 3-5, length of hospital stay and wound infection. Nevertheless, in spite of our rigorous methodology, the inherent limitations of eligible studies prevented us from reaching definitive conclusions. Based on the above clinical equipoise and potential benefit, future large-volume high-quality RCTs with long-term measures are awaited to affirm and update this system review.