Continuous Ambulatory Peritoneal Dialysis Catheter Insertion Technique: A Comparative Study of Percutaneous versus Surgical Insertion

Introduction

The chronic kidney disease (CKD) burden is increasing worldwide, and approximately 7% of existing patients are added yearly to renal replacement therapy (RRT).[1] Nearly 85% of dialysis centers are in a private set-up and in urban areas. The demographic data and socioeconomic profile of CKD patients in India are available with Indian CKD registry.[2] Continuous ambulatory peritoneal dialysis (CAPD) is a viable option of RRT for ESRD patients, who do not have access to hemodialysis centers due to long travelling distance. However, the long-term viability of peritoneal dialysis catheter (PDC) is always a concern. The success of PDC depends on catheter insertion technique in addition to other factors. Tenckhoff and Schechter described a very practical percutaneous method of peritoneal catheter placement in 1968, with higher risk of bowel or vessel injury, and malpositioned catheters at that time, resulting in failure rates of up to 65% at 2 years.[3] Walking in line with the higher safety of surgical PDC insertions, safety in percutaneous PDC insertions by nephrologists is now well established.[456] Laparoscopic PDC insertion is another safe procedure;[789] though its expertise is not easily available in rural India. Very few Indian studies are available comparing PDC placement using surgical and percutaneous techniques. We present a single-center retrospective analysis of percutaneous PDC insertions (Group “P”) versus surgical PDC insertion (Group “S”) using mini laparotomy.

Materials and Methods

We analyzed outcomes of PDC insertion techniques by two methods between January 2012 and March 2015 at a tertiary care government hospital in eastern India. Hospital Ethics Committee approved the study. Outcome was analyzed until September 2016. The inclusion criteria were ESRD patients with age >12 years who underwent PDC insertion at this center, patients with regular follow-up, and body mass index <30 kg/m². The exclusion criteria were obese, critically ill patients, prior abdominal surgery and/or simultaneous abdominal herniorrhaphy, recurrent CAPD peritonitis, poor follow-up, and laparoscopy-assisted PDC insertion. There was no restriction of cross over between the two groups in case of failure of one technique.

The primary outcome was primary non-functional PDC. It was defined as catheter malfunction either immediately after its insertion (during on-table flushing) or an inability to perform CAPD exchanges after break-in, thereby prompting either PDC removal or laparoscopic repositioning. The secondary outcomes were catheter survival (death-censored and overall), patient survival, catheter infection rate (exit site infection or tunnel infection, primary peritonitis rate, and secondary peritonitis rate), catheter removal, and mechanical complications (peri-catheter leak, catheter migration, flow problems, and scrotal swelling).

Catheter survival was analyzed at 1, 3, and 6 months, 1 year, and at the end of the study. Catheter survival was estimated, as “overall catheter survival” and death-censored catheter survival. In the event of death with a functioning catheter, the date of death was taken as the last follow-up. The definition of catheter loss was PDC removal due to mechanical complications or peritonitis. Causes of patient death were sudden death and death due to sepsis of any etiology. Primary peritonitis was defined as peritonitis within 1 month of PDC insertion whereas secondary peritonitis defined as peritonitis occurring beyond 1 month.

Data collation

We evaluated CAPD insertions between January 2012 and March 2015 and their follow-up until September 2016 with inclusion and exclusion criteria. The data included the personal details of the patients, date of insertion, type of procedure, type of catheter (2-cuff straight or coiled), operator (nephrologist-1 or 2 or surgeon-1), on-table flushing result, catheter position, outpatient department flushing, size of incision, break-in, mechanical complications, peritonitis episodes and duration, patient status, and catheter status. A catheter reinsertion was taken as a new case. All data were cross-checked for accuracy. The operative differences of two techniques are deliberated in Table 1.

Table 1 Comparison of peritoneal dialysis catheters insertion procedure between percutaneous (Group “P”) and surgical minilaparotomy techniques (Group “S”)

The percutaneous placement of PDC (Group “P”) was done by nephrologist by modified Seldinger technique[3] using Quinton^® PDC insertion kit [Figure 1a] by standard percutaneous technique. The surgical insertion of PDC (Group “S”) was done by either nephrologist or surgeon through minilaparatomy technique [Figure 1b].

Close

Figure 1

Schematic illustration of surgical anatomy of peritoneal dialysis catheters insertion. (a) Percutaneous technique, (b) surgical minilaparotomy technique

Export to PPT

Results

One hundred and fifty-nine PDCs were inserted in the Nephrology Department of our hospital during the study period, and 140 insertions qualified for analysis [Figure 2]. Of the 140 PDC insertions, 47 PDCs were inserted by percutaneous technique (Group P) and 93 insertions were done by surgical minilaparatomy technique (Group S). The baseline characteristics of patients were comparable [Table 2]. We compared and statistically analyzed the primary and secondary outcomes between both groups [Table 3]. The success rate of PDC implantation and the immediate catheter survival was good and comparable in both groups. One PDC in each group was primarily nonfunctional (P = 0.9). One catheter removal occurred in “P” group within 1 month of insertion due to refractory primary peritonitis.

Close

Figure 2

Details of peritoneal dialysis catheters insertions

Export to PPT

Table 2 Baseline characteristics of patients in both groups

Table 3 Comparison of outcomes and other major events in peritoneal dialysis catheters insertion between Group “P” and Group “S”

The catheter survival at 6 months was slightly better in “S” group (87.0% vs. 80.8%, P = 0.05). At the end of our study, both overall catheter survival (50.5% in “S” group and 53.1% in “P” group, P = 0.2) and death censored catheter survival rates were equal in both groups (P = 0.17) [Figure 3]. Subgroup analysis did not reveal any significant relation between catheter survival and prior hemodialysis >1 month duration (RR = 0.86, 95% confidence interval [CI] =0.53–1.39, P = 0.55; OR = 0.80, 95% CI = 0.39–1.65, P = 0.55) or with prior catheter-related bloodstream infection (CRBSI) (RR = 1.13, 95% CI = 0.69–1.85, P = 0.6; OR = 1.20, 95% CI = 0.56–2.58, P = 0.62).

Close

Figure 3

Kaplan–Meier curves for catheter survival in percutaneous (Group “P”) and surgical minilaparotomy technique (Group “S”)

Export to PPT

Patients belonging to the different etiology subsets depicted a comparable catheter survival (P = 0.37) [Figure 4]. Patient survival was equal in both techniques of insertion till 24 months, but thereafter, it worsened in “S” group probably because of longer duration of patient follow-up [Figure 5]. It was comparable among the different etiological groups (P = 0.62), though the diabetic patients' survival worsened beyond 26 months [Figure 6]. There was no relation between patient deaths to prior hemodialysis (RR = 0.76, 95% CI = 0.39–1.48, P = 0.42; OR = 0.71, 95% CI = 0.30–1.65, P = 0.4) or to those with prior CRBSI (RR = 0.58, 95% CI = 0.23–1.45, P = 0.24; OR = 0.52, 95% CI = 0.18–1.51, P = 0.23).

Close

Figure 4

Kaplan–Meier curves for catheter survival in the various etiologies of end-stage renal disease

Export to PPT

Close

Figure 5

Kaplan–Meier curves for patient survival in percutaneous (Group “P”) and surgical minilaparotomy technique (Group “S”)

Export to PPT

Close

Figure 6

Kaplan–Meier curves for patient survival in the various etiologies of end-stage renal disease

Export to PPT

Mechanical complications were more in percutaneous group (P = 0.049) [Table 4]. Percutaneous group had more catheter removals due to migration and omental wrap (10.6%, vs. 3.2% in “S” group, P = 0.033), whereas catheter removal due to refractory peritonitis was comparable (23.4%, vs. 22.5%, P = 0.45). The peritonitis rate was more in “P” group and was statistically significant (1 episode per 24.8 catheter-months, “P” group vs. 1 episode per 34.8 catheter-months in”S” group, P = 0.026) [Table 4]. Peritonitis rates were slightly more in rural population (P = 0.2). No relation was found between peritonitis rates and prior hemodialysis >1 month (RR = 1.25, 95% CI = 0.86–1.80, P = 0.23; OR = 1.54, 95% CI = 0.77–3.08, P = 0.22) or with prior CRBSI (RR = 1.18, 95% CI = 0.82–1.68, P = 0.36; OR = 1.38, 95% CI = 0.66–2.88, P = 0.37).

Table 4 Mechanical and infectious complications in Group “P” and Group “S”

Discussion

The long-term survival of PDC has many variables, the major players being operator competence, type of catheter inserted, insertion technique, mechanical, and infective complications including peritonitis episodes. The evolution of success in PDC implantation is a continuous process, and the past five decades have seen many success stories linked to various insertion techniques.[35611121314151617] Initiated as a domain of surgeons doing minilaparatomy PDC insertions, nephrologists gradually started with the introduction of percutaneous Seldinger technique with peel-away sheath,[3456] only to lose the battle again to the surgeons with laparoscopic intervention,[78918192021] which shortened the in-patient time and break-in period, faster recovery, less procedural morbidity, and a swift initiation of CAPD. In this study, PDCs were inserted by two methods. The basis of allocation of cases to either technique was the preference of the nephrologist.

Break-in period in our study was significantly shorter in “P” group (8.04 ± 1.17 days vs. 10.51 ± 1.71 days in “S” group, P = 0.002). This was comparable to Sivaramakrishnan et al. (9.7 ± 0.84 days, “P” group vs. 13.9 ± 3.39 days, “S” group),[4] whereas Sampathkumar et al. comparatively had much-shortened break-in period (4.6 ± 2.44 days, “P” group vs. 6.31 ± 2.68 days, “S” group).[22] Coiled two-cuff catheters had a better success rate than straight two-cuff catheters. The primary catheter nonfunction in our study was minimal (2.1% in “P” group vs. 1.1% in “S” group) as compared to other studies where it was higher.[4518] The results were better as we practised the procedure of filling peritoneal cavity with CAPD fluid in “P” group[3] and an atraumatic peritoneal and omental inspection in “S” group.

Medani et al.[17] reported 3-month catheter survival as 86.6% versus 77.0%, P = 0.037, whereas our data showed 95.7% and 98.9%, P = 0.66, in “P” group and “S” groups, respectively. The 6-month death-censored catheter survival in our study was 80.8% and 87%. The 1-year catheter survival was better in the surgical group in our study (68.0%, “P” group vs. 76.3%, “S” group, P = 0.05). This result was also seen in studies from AIIMS, India (77.9% vs. 78.4%),[4] Park et al. (89.9% vs. 93.3%),[23] and Perakis et al. (89.5% vs. 91.1%)[24] in “P” and “S” groups, respectively. However, 1-year survival was better in percutaneous insertion as reported by Ozener et al. (90% vs. 82%)[5] and by Medani et al. in the two studies,[1725] in “P” group and “S” groups, respectively. Meta-analysis published by Boujelbane et al.[26] and Tullavardhana et al.[27] did not show any difference in the 1-year catheter survival between both techniques. Surgical group fared better against the percutaneous group in our study compared to others wherein percutaneous insertions had better results.

The death-censored catheter survival at the end of our study was comparable in both groups, as compared to the previously reported poorer surgical outcomes.[451726] Considering the period of follow-up, catheter survival balance was variable in both groups. Initial catheter survival (<8 months) was equal in both groups, whereas it was better in surgical group at 8–16 months period and became equal in both groups later on (>16 months). In our study, the overall catheter survival follow-up was longer in group “S” compared to group “P.” This is in contrast to other observational studies where patients with percutaneous insertion had a longer catheter survival.[451724] Catheter survival and patient survival were comparable in a subgroup of patients with diabetes, CGN, and chronic tubulointerstitial disease.

Patient survival was marginally better in “P” group beyond 24 months but statistically not significant. The cause of mortality in our patients was death at home (cause not known) (n = 18) and peritonitis-related severe sepsis (n = 4). The probable reason for death at home was a cardiovascular cause such as arrhythmias or acute coronary syndrome, and the inability to timely shift the patient to a health facility.

Studies have shown a decreasing trend in infectious complications over the years due to better hygiene, CAPD training, and follow-up. Our center had more peritonitis rates in rural population and less literate patients; however, the second peritonitis episodes were less due to retraining and hygiene neoawareness after the first peritonitis episode, thus a better catheter survival later. The catheter removal due to refractory peritonitis was more in the monsoon months (July–August) every year. The peritonitis rates in our study were comparable with others[45172627] though better in “S” group. The overall infectious complications were less with percutaneous insertions as seen in meta-analysis with 13 studies and 2481 patients (P = 0.02)[26] and by Tullavardhana et al.[27] with seven studies and 996 patients (P = 0.003) though these studies showed significant heterogeneity across studies. The authors submit that incidence of peritonitis is entirely a function of the degree of CAPD training, literacy, and personal hygiene; hence, adjudging their relation to insertion techniques is not justifiable.

Mechanical complications were comparably minimal at our center and better than pre-2010 era.[5172228] Meta-analysis[2627] showed a lower incidence of overall mechanical complications in “P” group. Park et al.[23] revealed that percutaneous group had a higher incidence of early mechanical complications (11.2% vs. 0%, P = 0.002) and higher number of catheter removals due to mechanical complications (7.9% vs. 1.3%, P = 0.047), which was consistent with our results where “P” group had more mechanical complications (P = 0.049) and more catheter removal (P = 0.033). The reason for this is probably the direct visualization of falling back of omentum and bowels, and under vision guidance of catheter to its position in the surgical group. We did not have any bowel injury in our patients because of the protocols and the practice of prior filling of the peritoneal cavity in “P” group. In the surgical group, we prevented bowel injury by a diligent peritoneal entry by the prior lifting of posterior rectus sheath and peritoneum by approximately 1.5 cm to drop bowels and omentum down and placing a stay suture.

There was no pericatheter leak in immediate or late postoperative period. Many studies have reported pericatheter leak rates of varying percentage and the rates being comparable between groups. “P” group had nil incidence of early leak despite using midline insertion approach compared with higher incidences at other centers, which used paramedian approach for percutaneous insertion.[451725] The reason for nil incidence of the early leak was the avoidance of undue stretch on linea alba entry by gentle pull-peeling of peel-away sheath approximately 0.5–1.0 cm above the skin along with simultaneous firm inward push of PDC and tight purse string secure on the rectus sheath. However, two patients developed a late pericatheter leak (at 16 and 19 months after insertion), which was managed with purse string with prolene in the first and repair of a concomitant incisional hernia in the other. There was no pericatheter leak in surgical group as the anterior and posterior rectus sheath closure was watertight and adequate time for healing of peritoneum was given (late break-in) compared to other case series.

Overall, both groups had equal catheter technical survival and patient survival. The mechanical complications, peritonitis rates, and catheter removal due to migration were less in surgical group compared to percutaneous group. The limitation of this study is that it is a retrospective analysis and the patient number is less. Our study confirms that both the insertion techniques have their own pros and cons. Advantages of percutaneous insertion as analyzed by our study are shorter break-in period, less morbidity, relatively less tissue trauma, and supposedly less equipment dependence but slightly more mechanical complications and peritonitis rates. The advantages of surgical insertions using minimally invasive techniques with a small incision and better tissue retraction are lesser mechanical complications, surety of not injuring the underlying bowels, and adequate time to peritoneal healing. The success of either insertion technique depends entirely on the expertise, experience, and sensitization of the operator to the impending complications and management thereof.

Conclusions

The catheter and patient survival in the two techniques of PDC insertion are comparable. Both the techniques have their pros and cons. Both techniques are cost effective and performed either at the bedside or in operation theater with the same number of assistants, instruments, the same set of local anesthetic and analgesic agents, and catheters. Nephrology resident should get themselves trained in both the techniques.