Usage of prophylactic perinephric drains is variable in renal transplantation, and there is a lack of consensus as to the relative benefit of placing an abdominal drain intraoperatively in this patient cohort [1]. Drainage of post-operative fluid collections and prevention of the development of perinephric collections are the main indications for placing such drains in this cohort of immunosuppressed surgical patients [2]. However, there is debate over the necessity of these drains, and whether they may introduce more risks. For example, placement of a drain can result in several complications, including but not limited to post-operative pain, visceral injury, surgical site infection, bleeding or malposition [3, 4]. Prospective studies in general and colorectal surgery have shown a higher surgical site infection risk when drains are inserted intraoperatively [4]. Furthermore, meta-analysis of randomised trials as well as prospective interventional studies suggest drain insertion results in more pain for patients who received intraoperative drain placement [5, 6].

The pathological basis for the development of collections is multifactorial, however immunosuppression, increasing age, obesity, smoking, difficulty of the operation such as bleeding or damage to surrounding structures such as lymphatic tissue in the recipient’s iliac lymph trunk are all thought to contribute to fluid collections post-operatively [7, 8]. Placing a drain during the index transplantation operation therefore is thought to serve as prophylaxis against these relatively common surgical complications. However, these complications are often sub-clinical, may occur after a surgical drain is removed, and not all post-operative collections require drainage. In addition, intraoperative haemostatic techniques may also be utilised to minimise fluid effusion post renal transplantation [9].

This systematic review aims to investigate the impact of prophylactic perinephric drains placed during renal transplantation surgery on immediate and short-term post-operative surgical complication rates. In addition, the broader impact on graft function will be assessed, as well as relevant important outcomes such as deep wound complications, and surgical site infection.

This study was carried out following the Preferred Reporting Items in Systematic Reviews and Meta-Analysis (PRISMA) [10]. The protocol was prospectively registered on the PROSPERO system from the University of York (CRD42021255795) on 10th May 2023 [11].

Across all four databases, 1,627 papers were identified, of which 87 were identified as duplicates and discarded. Our search therefore revealed 1,540 unique titles and abstracts across all four databases. Of these, four retrospective cohort studies were selected according to the methodology outlined above, and these are presented in Table 1. Figure 1 outlines a PRISMA flow diagram in selecting articles for inclusion. Across the four studies selected, a total of 2,002 patients’ outcomes data were extracted for analysis. 1,046 had an intraoperative drain placed, 956 did not. Drains were removed when the output recorded less than <50 mL/24 h consistently across three of the studies, and was not reported in the remaining study. Only Farag et al. reported the type of drain used (a Jackson-Pratt suction drain). Furthermore, three out of the four studies reported complete data on type of donor (live vs. deceased), Table 1.

This review found no overall benefit when placing perinephric drains prophylactically during renal transplantation, including when assessing need for re-intervention post-operatively. Similarly, this review found no overall benefit of prophylactic drainage on reducing superficial or deep wound complications. Finally, there is not good evidence that perinephric drain placement is associated with improved graft survival outcomes at 12 months post renal transplantation.

Current literature demonstrates the range of complications associated with prophylactic drain insertion. One prospective study suggests that surgical site infection risk is increased when drains are inserted during general surgery procedures (OR 2.41, 95% CI 1.32–4.30, p = 0.004), however less of an effect is seen in vascular and orthopaedic surgery [4]. Furthermore, a systematic review and meta-analysis of twelve randomised controlled trials involving 1763 patients showed patients who underwent drainage had significantly higher pain scores as measured by the visual analogue scale (MD 10.08, 95% CI 5.24 to 14.92; p < 0.00001) [5]. There are limited studies reporting the incidence of bleeding and iatrogenic visceral injury secondary to perinephric drain placement. In one case series of deep pelvic collection drainage, a 2% haemorrhage rate was reported [16]. Fluid collections within the liver parenchyma may be amenable to percutaneous drainage, however this carries a reported 4% risk of major complications such as hepatocolic fistula creation, biliary peritonitis, and arterioportal fistula formation [17–19]. For retroperitoneal perinephric drains, a treatment failure rate exceeding 30% has been reported, often due to drain malposition [17].

Early post-operative collections such as seromas and haematomas occur post-transplant but the majority are discovered incidentally and are usually managed conservatively. The incidence of post operative surgical site haemorrhage detected by imaging and associated with a concurrent serum haemoglobin drop of more than 20 g/L over a 24 h period is relatively low (4.9%), with 90% of cases occurring within 1 day of implantation [20]. Collections more likely to require intervention such as urinomas, abscesses and lymphoceles, typically present later in the post-operative course, and the association with drain insertion is unclear. Lymphoceles in particular are common post renal transplant, with an incidence of 0.6%–51% reported in the literature, and 6.4% according to one recent retrospective study [21]. Urine leak has a reported incidence of 0.6%–6% and generally appears in the early post-transplant period [22, 23].

There have been two similar reviews in this area published previously. In 2019, D’Souza et al. showed that drain placement is associated with a higher incidence of peri-transplant fluid collections (RR 0.62; 95% confidence interval, 0.42–0.90), however no significant difference in the development of wound related complications [24]. A later review by Zawistowski et al. provided an update with the inclusion of a 2021 retrospective single-centre cohort study by Farag et al. The primary end-point in the Zawistowski meta-analysis was also perigraft collections [1]. No significant difference was seen between drain-free and drainage groups (pooled unadjusted OR = 0.77, 95% CI: 0.28–2.17). Similarly, there was no statistically significant difference in the secondary end points of surgical site infection, lymphocele, haematoma, and wound dehiscence between patients who did or did not receive prophylactic drainage. This review provides the most recent and extensive review of the current literature assessing the role of prophylactic perinephric drainage on short and long term clinically significant complications post kidney transplant. While previous reviews focused on the incidence of common post-operative complications, these are not necessarily clinically significant, as not all collections require drainage. By focusing our primary outcome on reintervention rate for post-operative collections, we aimed to better demonstrate the clinical significance of prophylactic drainage on renal transplant patients. More generally, the search criteria were robust and consistent across a range of generic and transplant-specific databases, with no language or time-period restrictions applied during article selection. All studies were rated as “good quality” when rated for quality via the Newcastle-Ottawa Score.

However, this review and analysis has several limitations. Given the retrospective nature of the studies identified in the literature, it is not possible to confidently demonstrate causality between our exposure (drain placement) and outcome (reintervention rate) of interest. The control groups included in the studies (no drain placement) would likely also be affected by selection bias. For instance, the Derweesh et al. study shows significant differences between the groups with respect to patient body mass index (BMI) and immunosuppression use (specifically sirolimus). These are both factors which are known to affect wound healing, surgical site infection, and wound complications specifically in renal transplantation, therefore the results cannot reliably be interpreted due to the selection bias present in the cohorts [25, 26]. Owing to the small number of studies included in this analysis (less than 10), publication bias could not be accurately assessed using Egger’s regression test for funnel plot asymmetry [27]. We found significant heterogeneity in the reporting of outcomes, and so meta-analyses were performed where specific outcomes were published. We also intended to record outcomes such as post-operative pain around the wound or drain site, opiate usage, length of hospital stay, and overall mortality, however these data were not available in the published literature in relation to drain use. Analysis of these outcomes would allow us to more effectively examine the complications associated with drain insertion, however due to the lack of availability we were not able to do so. Patient-reported outcomes and measures following drain insertion in particular would be an important aspect of drain insertion to assess and report upon, and one which we advocate should be investigated in future prospective studies. Regarding Figure 3, we intended to include Cimen et al. results in our meta-analysis, however were unable to contact the authors to obtain raw data to include in the meta-analysis. This represents a drawback to our review because Cimen et al. found lower odds of peri-graft collection, thus favouring drain insertion (p = 0.01). Finally, there was heterogeneity in the definitions of parameters such as “wound complications,” whereby authors divided into either clinically significant vs. not significant, or superficial vs. deep, or specifically looking at individual complications such as surgical site infection, wound dehiscence, or superficial wound collection. We therefore only included data from studies where we were confident that the data reflected the specific outcomes of interest described above.

One of the key rationales for intraoperative drain placement is pre-emptive control of post-operative collections such as lymphocele, seroma, haematoma, urinoma or infected tissue fluid. Ongoing monitoring for bleeding and infective collection around the graft site are the main indication for routine placement of a perinephric drain, however placement of the drain itself is associated with risks. In a meta-analysis of 28 randomised trials involving 3,659 patients, Gurusamy et al. showed that a drain-free approach to open cholecystectomy was associated with significantly lower wound infection rates, and no difference in the incidence of post-operative abdominal collection [28]. Partly as a result of this, drains are now no longer placed for uncomplicated open cholecystectomy operations. Furthermore, a single-centre experience of combined liver-kidney transplants showed no difference in the incidence of superficial/deep wound complications, collection size, intervention rate, graft failure, and overall patient survival between drainage and non-drain patient cohorts [29].

Better access to cross-sectional imaging provides a non-invasive tool for surgeons to utilise in the investigation for post-operative collections. Ultrasound provides accurate assessment of vascular flow to the graft, and can assess the presence of perinephric fluid collection and associated graft parenchymal compression [30]. Imaging is not always performed routinely, however in association with symptoms such as fever or pain, signs of graft failure such as high serum creatinine, ipsilateral leg swelling or hydronephrosis, drainage of these collections is indicated [31].

Current practice also shows a variety approaches to prophylactic drainage. In 2020, a survey of 43 renal transplant surgeons across Australia and New Zealand revealed 61% of surgeons practising routine drain insertion, while 21% rarely inserted drains [32]. A more recent (2023) survey of UK-based transplant surgeon practices suggests over two-thirds of respondents routinely insert one drain, while 8.3% indicated insertion of two or more drains on a routine, prophylactic basis. Only one-fifth of surgeons insert drains selectively as reported in this study [33]. This suggests the need for a paradigm shift in how prophylactic drainage in renal transplantation is viewed, especially in the absence of overwhelming evidence supporting its impact on favourable post-operative outcomes.

Given the lack of clear benefit of placing perinephric drains intraoperatively during renal transplantation, negative impact on patient experience, and the potential risks, we advocate for a an approach whereby drains are only placed for specific indications on a case by case basis. Prospective data is needed to support this position, and trial-level evidence is warranted to support or discourage routine perinephric drainage.