Impact of surgeon experience on complication rates and functional outcomes of 484 deceased donor renal transplants: a single-centre retrospective study

Authors


Florian Fuller, Department of Urology, Charité University, Hospital Berlin, Charitéplatz 1, 10117 Berlin, Germany. e-mail: florian.fuller@charite.de

Abstract

Study Type – Therapy (outcomes)

Level of Evidence 2c

What's known on the subject? and What does the study add?

Although renal transplantation represents a well established surgical procedure, the learning curve for this procedure has not been studied so far. The published data on surgical complications do not discriminate between surgeons and their stage of expertise.

The present study highlights the importance of a structurized programme in urological sub-speciality training. Renal transplanation represents a standardized procedure with a low learning-curve for uncompliciated cases. When comparing experienced urological transplant surgeons with inexperienced surgeons, the postoperative complications and functional outcome were similar. One exception is seen in ureteroneocystostomy, which should be considered in clinical practice.

OBJECTIVE

  • • To determine how postoperative and functional outcomes after deceased donor renal transplantation (DDRT) are related to surgeon experience.

PATIENTS AND METHODS

  • • The outcomes of 484 adult DDRT performed by 13 urological surgeons were retrospectively reviewed.
  • • After completion of a staged renal transplant training programme under supervision of an attending urological transplant surgeon, the 13 surgeons were either assigned to the inexperienced group (n= 8) or the experienced group (n= 5).
  • • Surgeons in the experienced group had performed more than 30 unsupervised DDRT in a standard fashion with routine ureteric stenting.
  • • Between 1988 and 2005, inexperienced surgeons performed 152 DDRT, whereas experienced surgeons performed 332 DDRT.

RESULTS

  • • Patient and graft survival at 2 hyears were 98% and 94.7%, respectively.
  • • Early graft loss in five recipients was unrelated to surgeon experience.
  • • Delayed graft function occurred in 29% of cases and median 1-year serum-creatinine was 1.48 mg/dL, with no difference between surgeon groups.
  • • Postoperative bleeding and lymphocele formation were the most frequent surgical complications, with an equal distribution between groups.
  • • Ureteric complications had a significantly higher incidence among inexperienced surgeons (6.6% versus 2.7%; P= 0.04).

CONCLUSION

  • • We conclude that DDRT as performed by inexperienced urological renal transplant surgeons has both acceptable short- and long-term outcomes.
Abbreviations
DDRT

deceased donor renal transplantation

DGF

delayed graft function.

INTRODUCTION

Renal transplantation is the treatment of choice for patients with end-stage renal disease [1]. Compared to haemodialysis, deceased donor renal transplantation (DDRT) offers a superior quality of life and a prolonged overall survival [2]. The best results in patient and graft outcomes are achieved by living donor renal transplantation [3].

As with most operative procedures, centre volume is considered to have a major influence on complication rates and graft function after DDRT [4]. Surgical complications still represent a major source of postoperative morbidity in recipients of deceased donor renal transplants. Surgeon-associated risk factors have a major impact on surgical complications [5,6]. However, several studies have shown that, with adequate supervision, surgeon expertise is not a determinant of postoperative outcomes [7,8].

In Germany, ≈30% of renal transplants are performed by urologists. In most German urological transplant centres, renal transplantation is exclusively carried out by experienced surgeons. As a high-volume renal transplant centre with currently more than 100 renal transplants per year, we aimed to expand the pool of urological transplant surgeons by implementing a staged training programme for board-certified urological specialists with no experience in renal transplantation.

In the present single-centre retrospective study, we compared the outcomes of 484 adult DDRT performed by either experienced or inexperienced transplant urologists. All 13 urological surgeons who participated in the present study were assigned to either category based on the number of unsupervised DDRT (less than 30 versus more than 30) performed after finishing our staged renal transplant training programme. The present study aimed to evaluate the impact of surgeon experience on graft outcomes and complication rates after DDRT.

PATIENTS AND METHODS

Between January 1988 and December 2005, 691 adult recipients underwent DDRT at Charité University Hospital, Campus Mitte. Screening of our electronic database and of the clinical charts yielded complete data sets on 484 DDRT recipients. Patient inclusion criteria were a follow-up of at least 1 year and recipient age >18 years. DDRT were performed by 13 urological transplant surgeons, all of whom had participated in a staged renal transplant training programme. Regarding the structure of the training programme, during the first 10 renal transplants, the trainee performed parts of the vascular and vesico-ureteric anastomosis under the guidance of an attending urological transplant surgeon. The next 10 transplants were performed entirely by the trainee himself under close supervision of an attending urological transplant surgeon. After completing the training programme, all 13 surgeons were allowed to perform unsupervised DDRT. To stratify the 13 surgeons into two expertise categories (inexperienced versus experienced), a threshold of 30 unsupervised DDRT was chosen. According to this criterion, eight surgeons who performed less than 30 unsupervised DDRT were retrospectively classified as ‘inexperienced’, whereas five were classified as ‘experienced’. Although the inexperienced group consisted mostly of board certified urological specialists at the beginning of their 2-year sub-speciality training, the experienced group consisted of attending surgeons and of specialists in their final stage of urological sub-specialization. The distribution of renal transplantation per annum is shown in Fig. 1A. The main surgical volume of the inexperienced transplant surgeons was seen in the years 2000–2005, parallel to an increase of centre volume (Fig. 1B). All of the surgeons included in the present study participated in the Eurotransplant organ retrieval programme.

Figure 1.

(A) Number of deceased donor renal transplantations (DDRT) (tx) performed per annum during the study period. An increased DDRT volume is seen after the year 2000; a trend which remains ongoing today (data from 2005 to 2010 not shown). (B) Number of DDRTs per surgeon group per annum. After 2000, a gradual increase of surgeon volume is seen in the inexperienced group.

All donor kidneys were flushed with histidine–tryptophan–ketoglutarate solution. Renal transplantation was performed via a Gibson incision using the standard extraperitoneal technique with placement of the graft in the iliac fossa and end-to-side anastomosis between the renal vessels and the external iliac vessels of the recipient. Vesico-ureteric anastomosis was accomplished using the Politano–Leadbetter technique with routine ureteric stenting.

IMMUNOSUPPRESSION

Triple immunosuppression was started on the day of transplantation and consisted of prednisone, mycophenolatemofetil and a calcineurin inhibitor. Oral prednisone was tapered to a dose of 10 mg per patient during the first 6 months. Target trough serum levels for cyclosporine and tacrolimus were between 150–250 ng/mL and 10–12 ng/mL, respectively.

Patients with a high immunological risk (pre-transplanted patients, panel-reactive antibodies positive or living unrelated donor) received induction therapy with anti-interleukin-2 receptor antibodies (20 mg of basiliximab pre-transplantation and at day 4 post-transplantation).

STATISTICAL ANALYSIS

Statistical analysis was performed using PASW, version 18 for Windows (SPSS Inc., Chicago, IL, USA). Values are reported as the median (sem), unless indicated otherwise. For evaluation of differences between medians, the Mann–Whitney U-test was used. Group means were compared using anova. Survival analysis was calculated by the Kaplan–Meier method using the log-rank test to compare groups. P < 0.05 was considered statistically significant.

RESULTS

DONOR AND RECIPIENT DEMOGRAPHICS

The group of experienced surgeons (n= 5) carried out 332 renal transplants, whereas the group of inexperienced surgeons (n= 8) carried out 152 renal transplants. Demographic data did not differ significantly between the two groups of surgeons (Table 1).

Table 1.  Donor and recipient demographics in 484 deceased donor renal transplantations
VariableExperienced surgeons (n= 5)Inexperienced surgeons (n= 8)
Renal transplants, n 332152
Donor age (years), mean (sem)49 (17)48 (18)
Recipient age (years), mean (sem)48 (15)47 (14)
Male/female recipients (%)60.5/39.562.5/37.5
Recipients with second transplant (%)20.815.8
Recipient mean body mass index (kg/m2)2727
Renal artery multiplicity (%)2627
Cold ischaemia time (h)14.6 (6.3)14 (6.5)
Recipient disease (%)  
 Glomerulonephritis34.340.8
 Pyelonephritis17.813.8
 Polycystic kidney disease9.310.5
 Analgesic nephropathy5.73.9
 Hereditary nephropathy5.13.9
 Hypertensive nephropathy3.37.2
 Unknown/other2.42.6

PERI-OPERATIVE DATA AND POSTOPERATIVE OUTCOME

Mean cold and warm ischaemia times were 14 h and 48 min, respectively. Overall 2-year patient and graft survival were 98% and 94.7%, respectively. Sub-analysis of the surgeon groups showed a 2-year patient survival of 99.7% in the inexperienced group and 97.6% in the experienced group (P= 0.43; Fig. 2A). The 2-year graft survival was 96.7% for the inexperienced surgeons and 94.3% for the experienced surgeons (P= 0.26; Fig. 2B). Median 1-year serum creatinine was 1.48 (1.05) mg/dL. The overall incidence of delayed graft function (DGF), defined as the need for at least one haemodialysis within the first week, was 29%. The 1-year rejection rate was 35%. The differential distribution of postoperative parameters among the two groups of surgeons is outlined in Table 2. No significant differences were seen between the two groups.

Figure 2.

(A) Kaplan–Meier curve for 2-year patient survival for inexpierenced (<30 deceased donor renal transplantations [DDRT]) and experienced (>30 DDRT) surgeons. (B) Kaplan–Meier curve for 2-year graft survival for inexpierenced (<30 DDRT) and experienced (>30 DDRT) surgeons. The number at risk at 2 years was 143 in the inexperienced group (n = 152) and 307 in the experienced group (n = 332).

Table 2.  Postoperative outcome
VariableExperienced surgeons (n= 5)Inexperienced surgeons (n= 8)
Renal transplants, n332152
Follow-up (months), mean (sem)96 (54)72 (70)
Warm ischaemia time (min), mean (sem)48 (15.2)48 (14.4)
1-year serum-creatinine (mg/dL), mean (sem)1.52 (1.07)1.43 (1.02)
Delayed graft function (%)2830
1-year rejection rate (%)3535
Postoperative complications (%)17.718.6

SURGEON EXPERTISE AND POSTOPERATIVE COMPLICATIONS

Differential analysis of surgical complications is shown in Table 3. The incidence of postoperative bleeding and vascular complications, as well as the frequency of lymphocele drainage, were equally distributed between groups. Early graft loss as a result of vascular thrombosis occurred in five recipients. In the experienced group, three grafts were lost as a result of arterial or venous thrombosis within the first few days after transplantation. In the inexperienced group, two grafts were lost within the first week as a result of venous thrombosis. No technical difficulties were encountered during engraftment. Revision surgery was performed by an experienced surgeon and involved in situ flushing of the renal vein with heparinized saline. Eventually, both renal grafts had to be removed as a result of irreversible ischaemic damage.

Table 3.  Surgical complications
VariableExperienced surgeons (n= 5)Inexperienced surgeons (n= 8) P
Renal transplants, n332152 
Postoperative bleeding with intervention, n (%)14 (4.2)6 (3.9)0.63
Lymphocele drainage, n (%)25 (7.5)9 (5.9)0.79
Ureteric complications, n (%)9 (2.7)10 (6.6)0.04
Ureteric stenosis, n (%)5 (1.5)6 (3.9)0.09
Vascular thrombosis, n (%)11 (3.3)4 (2.6)0.74
Graft loss as a result of vascular thrombosis, n (%)3 (0.9)2 (1.3)0.5

The incidence of ureteric complications was significantly higher in the inexperienced group (6.6% versus 2.7%; P= 0.04). Ureteric stenosis represented the most frequent ureteric complication in both goups of surgeons. Although the rate of ureteric stenosis was higher in the inexperienced group, this was not statistically significant (P= 0.09). In six out of 11 patients with a ureteric stenosis, open reanastomosis of the transplant ureter was necessary. The remaining five patients were treated with ureteric stenting for at least 3 months. Ureteric leakage resolved in three out of eight DDRT recipients after conservative treatment. Surgical treatment for ureteric leakage was required in five recipients.

SURGEON-RELATED COMPLICATIONS

The complication rate of each individual surgeon is outlined in Table 4. The number of transplantations performed by an experienced surgeon was in the range 30–101. The complication rate in this group was 17.7%. The number of transplantations carried out by an inexperienced surgeon was in the range 6–27. The complication rate in this group was 18.6%. There was no inverse correlation between surgeon expertise and complication frequency.

Table 4.  Surgeon-related complications
VariableNumber of transplantations, nComplications, n (%)
Experienced surgeons  
 Surgeon 1304 (13.3)
 Surgeon 2429 (21.4)
 Surgeon 37216 (22.2)
 Surgeon 48718 (20.7)
 Surgeon 510111 (10.9)
 Mean rate (%) 17.7
Inexperienced surgeons  
 Surgeon 161 (16.7)
 Surgeon 2113 (27.3)
 Surgeon 3142 (14.3)
 Surgeon 4224 (18.2)
 Surgeon 5233 (13)
 Surgeon 6236 (26)
 Surgeon 7262 (7.6)
 Surgeon 8277 (25.9)
 Mean rate (%) 18.6

DISCUSSION

To our knowledge, the present study is the first reported investigation of the impact of surgeon experience on postoperative complications and early functional outcomes after DDRT. In the present retrospective study, we found that urological surgeons who were inexperienced in renal transplantation, and who completed a staged training programme under the guidance of an attending renal transplant urologist, can safely perform this procedure with outcomes comparable to those of experienced urological renal transplant surgeons.

Donor and recipient demographics, as well as peri-operative parameters, including cold and warm ischaemia times, did not differ between the two patient cohorts assigned to either the experienced or the inexperienced group of transplant surgeons. A 1-year serum creatinine level <1.5 mg/dL has been shown to predict excellent long-term graft survival [8]. In the present study, an overall 1-year serum creatinine level of 1.48 mg/dL translated into a favourable 2-year graft survival rate of 94.7%. Surgeon experience had no significant impact on 1-year serum creatinine levels, with the inexperienced group showing a slightly lower value (Table 2). Delayed graft function is a surrogate marker of donor organ quality and non-immunological injury to the graft resulting from cold ischaemia and surgical trauma. Donor age is the dominant factor determining DGF and graft survival in deceased donor renal transplants [9]. A median donor age of 48 years in the present study cohort resulted in a DGF rate of 29% with no difference between groups. In their recent retrospective study on 113 adult renal transplant recipients, Hokema et al. [10] reported similar values: donor age of 51 years and a DGF rate of 31%. In a previous study from our institution, the incidence of DGF among deceased donor renal transplant recipients reached 40% with a median donor age of 58.5 years [11].

To assess the quality of the renal transplant procedure itself, we compared intra-operative and postoperative complications between experienced and inexperienced urological transplant surgeons. Relevant surgical complications included haemorrhage, vascular thrombosis, urinary leakage, ureteric stricture and lymphocele formation requiring surgery. Postoperative haemorrhage and lymphocele formation requiring surgery occurred in 4% and 6.7% of cases, respectively. In the literature, the incidence of lymphoceles after renal transplantation varies between 0.6% and 26% [12–16].

The incidence of surgical complications does not appear to decrease with growing surgical expertise, suggesting that factors such as donor and recipient morbidity may influence the type and frequency of post-transplant surgical complications (Table 4). However, in-depth analysis of the source of post-transplant surgical complications is beyond the scope of the present study given the relatively small patient numbers.

We found a significantly higher incidence of ureteric complications in the inexperienced group compared to the experienced group. For ureteroneocystostomy, all surgeons used the antirefluxive Politano–Leadbetter technique. With this technique, the transplant ureter is anchored within the bladder using a submucosal tunnel. Previous studies have shown that this intravesical implantation approach is more susceptible to complications compared to other techniques [17,18]. Although a higher stenosis rate was found in the inexperienced group of surgeons in the present study, the overall ureteric complication rate is comparable to that reported in the literature [17,19]. The learning curve associated with the technically demanding Politano–Leadbetter technique could have accounted for the higher rate of ureteric complications among inexperienced surgeons compared to the experienced group. We found that two surgeons in the inexperienced group were responsible for six out of 10 (60%) ureteric complications. In the present study, ureteric complications did not affect graft outcomes, which is in agreement with previously reported results [20,21]. In the literature, the extravesical Lich-Gregoir technique is favoured among transplant surgeons [22]. The overall ureteric complication rate is consistent with previously published data, even with those studies using the Lich-Gregoir technique [16,23,24]. As a consequence of the findings of the present study, we recently switched from the intravesical ureteric implantation technique to the extravesical Lich-Gregoir technique.

In times of limited human resources as a result of economic constraints, high-volume surgical or urological centres with a busy renal transplant programme are forced to steadily increase the number of surgeons performing DDRT. Working time regulations may limit the availability of attending surgeons for scheduled daytime procedures if they have performed renal transplantation the night before [25]. The question arises as to whether or not there is a minimal volume of renal transplants that a board certified urological specialist has to perform under the guidance of an attending surgeon before embarking on unsupervised ‘out-of-hours’ renal transplantations. Our staged renal transplant training programme, including 20 guided procedures, proved to be appropriate to enable urological surgeons with no previous experience in renal transplantation to safely perform unsupervised DDRT. Certification by the American Society of Transplant Surgeons requires a transplant fellow to perform 30 renal transplants over the duration of the fellowship (http://www.asts.org/fellowshiptraining/accreditation.aspx). Choosing a threahold surgeon volume of 30 unsupervised renal transplants, with the aim of discriminating between experienced and inexperienced urological transplant surgeons, we found no significant difference between the two groups in terms of postoperative complications and early graft outcomes. It may be argued that the threshold volume for an experienced transplant surgeon could have been lowered to 10 or 15 procedures, given the apparent success of our training programme. Nevertheless, in the present study, experienced surgeons had a significantly lower ureteric complication rate compared to inexperienced ones. Whether or not the extravesical Lich-Gregoir ureteroneocystostomy yields better results among inexperienced transplant surgeons than the more complex intravesical Politano–Leadbetter technique is a matter to be resolved in future investigations. The limitations of our present study include its retrospective character and the relatively large number of surgeons involved.

We conclude that DDRT can be a technically challenging surgical procedure and, given the extreme shortage of organs from deceased donors, renal transplant surgeons have a high responsibility for the future well-being of their patients. However, because of economic constraints with respect to limited human resources among healthcare providers, increasing the pool of renal transplant surgeons is indispensable, especially in high-volume transplant centres. We have shown that urological surgeons who are inexperienced in renal transplantation are capable of safely performing DDRT after participating in a staged training programme under the supervision of an experienced renal transplant urologist. The higher incidence of ureteric complications among inexperienced surgeons using the intravesical Politano–Leadbetter ureteroneocystostomy technique had no impact on long-term graft survival and may decrease with the use of the more robust Lich-Gregoir technique.

CONFLICT OF INTEREST

The authors declare that there are no conflicts of interest.

Ancillary