- Email: [email protected]
Risk factors for early peritoneal dialysis catheter failure in children
Journal of Pediatric Surgery (2010) 45, 585–589
www.elsevier.com/locate/jpedsurg
Risk factors for early peritoneal dialysis catheter failure in children Randolph K. Cribbs a , Larry A. Greenbaum b , Kurt F. Heiss a,⁎ a
Division of Pediatric Surgery, Department of Surgery, Emory University, Atlanta, GA 30322, USA Division of Pediatric Nephrology, Department of Pediatrics, Emory University and Children's Healthcare of Atlanta, Atlanta, GA 30322, USA b
Received 5 February 2009; revised 11 June 2009; accepted 12 June 2009
Key words: Children; Peritoneal dialysis catheters; Laparoscopy
Abstract Background: There is uncertainty regarding the optimal approach for surgical placement of peritoneal dialysis (PD) catheters in children. Operative technique, catheter selection, and patient variables (eg, age or prior surgical history) may influence catheter lifespan. Methods: A retrospective review of all PD catheters placed at a tertiary children's medical center during a 6-year period was performed. Our primary outcome was catheter function 2 months after placement. Data were analyzed using Student 2-tailed t test or χ2 analysis. Results: There were 121 PD catheters placed in 81 patients. The median primary functional catheter lifetime was 109 days. Primary PD catheter failure (within 2 months) occurred in 36 catheters (30%). Patients with primary catheter failure (8 ± 7 years) were younger than patients with a functioning catheter at 2 months (12 ± 5 years; P = .002). Catheters placed without simultaneous omentectomy were more likely to fail (P = .042). Catheter failure rate was not significantly different based upon operative technique or catheter type. Conclusion: Omentectomy at the time of catheter placement decreased the risk of early catheter failure. In contrast, type of catheter or laparoscopic placement did not influence the likelihood of early catheter failure. © 2010 Elsevier Inc. All rights reserved.
For children with renal failure, peritoneal dialysis (PD) is the preferred method of dialysis [1]. Compared to hemodialysis, which requires the child to be attached to a dialysis machine at a hospital or outpatient clinic for up to 4 hours a day, 3 days every week, PD can be performed at home by caregivers with a minimum of training.
⁎ Corresponding author. Tel.: +1 404 727 3779; fax: +1 404 727 2120. E-mail address: [email protected] (K.F. Heiss). 0022-3468/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2009.06.019
Placing PD catheters in children poses unique challenges reflected in the high complication rates for children, as high as 70% in some series [1]. The abdominal muscle layers are thinner in children compared to adults, making it difficult to affix the catheter in place, prevent movement of the catheter tip inside the abdominal cavity, and prevent liquid leakage from the abdomen to the skin. Major causes of PD failure are catheter-related infection and malfunction. Abdominal hernias, although not a threat to the PD catheter itself, are also common in children after PD catheter placement [2].
586 Laparoscopy was initially used in adults to rescue blocked PD catheters and, later, for initial PD catheter placement [3]. By facilitating a thorough exploration of the abdominal cavity, permitting identification of coincidental intra-abdominal disease and careful placement of the PD catheter tip, laparoscopy has been suggested to promote longer catheter survival than open placement. However, the optimal placement technique remains uncertain. Other factors that may influence PD catheter lifespan include catheter selection (straight vs. curled), resection of the omentum during catheter placement, and patient variables (e.g., age or prior surgical history). At our institution, open PD catheter placement and laparoscopic-assisted PD catheter placement have been used in nearly equal numbers, based upon individual surgeon preference, since 2002. In addition, the two PD catheter types (straight or curled) have also been employed nearly equally during the period of study. This situation has provided us with a unique opportunity to analyze the short-term outcome of a large number of PD catheter placements in an attempt to identify perioperative factors associated with successful catheter placement.
1. Methods We performed a retrospective chart review of all patients who had operative peritoneal dialysis catheter placement at Children's Healthcare of Atlanta at Egleston, a tertiary children's hospital in the Southeastern United States, from January 2002 to December 2007. Institutional review board approval was obtained prior to record retrieval. Operative and nonoperative variables and dates of initiation and cessation of dialysis were recorded. All catheters had a synthetic cuff to secure to the abdominal wall; catheter tips were placed in the pelvis without additional suture to the pelvic wall. All catheters were tunneled above the fascia to the opposite side of the abdominal wall before exiting at the skin site. The skin exit site was placed in the left upper quadrant in 45 cases (37%), right lower quadrant in 30 cases (25%), left lower quadrant in 24 cases (20%), and the right upper quadrant in 22 cases (18%). In all cases, the external portion of each PD catheter was oriented downward toward the feet at the exit site. The PD catheters were placed by 7 different pediatric surgeons and 1 transplant surgeon: 5 predominantly used a laparoscopic approach; the remaining 3 typically used an open approach. Of 7 surgeons, 6 performed simultaneous omentectomy more frequently during PD catheter placement, whereas 1 surgeon tended not to perform a simultaneous omentectomy. Two surgeons used curled PD catheters exclusively, whereas the remaining surgeons tended to prefer the straight catheters (irrespective of patient age). There was also no consensus regarding exit site placement: 4 surgeons had a predominant site preference, whereas the remaining
R.K. Cribbs et al. surgeons used several different exit sites. Individual surgeons performed a total of 37, 26, 22, 14, 12, 7, 2 and 1 case, respectively. All participating surgeons had prior experience in performing PD catheter placement before the study. Our primary outcome was catheter function 2 months after placement. We excluded PD catheters placed for acute renal failure, those not used for dialysis within 3 weeks of placement, and those without postoperative follow-up. We did not consider a catheter as having failed if it was functioning when removed following successful renal transplantation. Data analysis of successful vs. failed groups was performed using Student 2-tailed t test and χ2 analysis, as appropriate. P b .05 was considered significant. Statistical software was SPSS 15.0 for Windows (SPSS, Chicago, Ill).
2. Results There were 121 PD catheters placed in 81 patients. There was a wide range of patient ages and weights in our cohort (Figs. 1 and 2). Demographic information and surgical characteristics are displayed in Table 1. One laparoscopic placement was converted to an open procedure after a sustained rise in blood pressure was noted with peritoneal insufflation; this case was grouped with open procedures for statistical calculations. Of 121 catheters, 24 were placed in children weighing less than 10 kg (20%); 15 of these catheters were placed in children under the age of 1 year (12%). The mean time from catheter placement to initiation of PD was 4 ± 5 days (median, 2 days). Of PD catheters placed in children younger than 1 year, 9 were straight catheters (60%), and 6 were curled. Simultaneous omentectomies were performed in 8 of these 15 catheter placements (53%). Laparoscopic technique was used to place 7 of the 15 catheters in this group of infants (47%). During the study period, 55 children (68%) had a single PD catheter placement, 14 children (17%) had 2 PD catheter placements, 11 children (14%) had 3 catheter placements, and 1 child (1%) had 5 catheters placed. A total of 15 catheter revisions were performed during the study period; 7 (47%) were successful in achieving a functional PD catheter at the 2-month postoperative end point. There were a total of 36 catheter failures (30%) within 2 months post placement (Table 2). Two children died during the 2-month postoperative period. One child was a 1-monthold male with bilateral renal agenesis who died with a rectal wall perforation discovered 10 days after PD catheter replacement (for leakage around the catheter exit site). This infant presented with dialysate leaking from his rectum, and the PD catheter tip was seen shortly thereafter extending into the infant's rectal vault. The dialysis catheter had been functioning normally from the time of replacement until the day the catheter tip was discovered in the rectum. His parents chose to decelerate care at that point, and the infant died 10
Risk factors for early peritoneal dialysis catheter failure in children
Fig. 1
The ages of the patients at the time of catheter insertion.
days after cessation of dialysis. The other child was a 2month-old male with a congenital diaphragmatic hernia repair who died of a pneumothorax 30 days after placement of a PD catheter. His catheter was functioning at the time of death; he was not considered a catheter failure. The principal cause of failure was catheter occlusion (75% of cases), either from abutment against an external object (eg, omentum, epiploic coli, or fallopian tube) or from formation of a fibrin plug or clot within the catheter. Failure owing to infection of the peritoneal cavity or the subcutaneous tunnel occurred in 4 cases (11%). Dialysate leakage from the skin site led to catheter failure in 2 cases (6%). The remaining causes of catheter failure included one
Fig. 2
587
case each of severe postoperative hemorrhage, inadvertent catheter removal by the patient, and perforation of the rectum (as noted above). Of the 36 catheter failures, 24 were removed without revision (67%), 10 were revised once (28%), 1 catheter was revised twice (3%), and 1 catheter had three revisions (3%) within the 2-month postoperative period. The failed PD catheters were replaced in 21 cases (58%); 7 children were converted to hemodialysis following removal of a failed PD catheter (19%). Children weighing less than 10 kg had a failure rate of 62.5% compared to a failure rate of 21.6% in children weighing at least 10 kg (P b .001). Omentectomy at the time
The weights of the patients at the time of catheter insertion.
588
R.K. Cribbs et al.
Table 1 Patient demographics and surgical characteristics (n = 121) Male sex Median age (range) Median weight (range) Previous abdominal surgery Previous PD catheter Previous omentectomy Straight tipped catheter Laparoscopic placement Omentectomy at time of surgery Mean catheter lifespan (median) Mean revised catheter lifespan (median)
68 (56%) 12 y (7 d-19 y) 38 Kg (2.3-88.2) 67 (55%) 53 (44%) 33 (27%) 62 (52%) 69 (57%) 47 (39%) 177 ± 204 d (109 d) 207 ± 212 d (145 d)
of catheter placement was associated with a lower catheter failure rate (19%) than occurred when no omentectomy was performed (36%; P = .04). However, the failure rate was 27.5% among children with a previous omentectomy and 34% in children without an omentectomy (nonsignificant). The postoperative infection rate of PD catheters placed via laparoscopic technique (9%) did not differ significantly from the infection rate following open PD catheter placement (6%). There was no significant difference in the mean number of days prior to using a catheter when we compared laparoscopic placement (5 days) to open placement (4 days).
3. Discussion In 2004, Washburn et al [1] published the results of a survey of members of the American Pediatric Surgical Association and the Canadian Association of Pediatric Surgeons regarding surgical technique for PD catheter placement. Their results demonstrated significant variability in all aspects of surgical technique, with the most common practice being the placement of a single-cuffed, curled, non– Swan-necked catheter using open technique with a superiorpointing exit site. Routine omentectomy was reported by 59% of respondents. The largest recent series of PD catheter outcomes was a multi-center study from Italy with 503 catheters placed over a 15-year time period [4]. While failure rates improved over the study period, the predominant cause of catheter failure in this series was exit-site and skin tunnel infection, with obstruction in only 5% of catheters. All of these catheters were placed using open technique. Another recent study evaluated the placement of PD catheters in infants and children weighing less than 10 kg [5]. This study showed a higher incidence of catheter failure and other catheter-related complications with smaller-weight children. A study of percutaneously-placed PD catheters, similar in size to our study [6], evaluated 108 PD catheters placed percutaneously in 93 children with an average age of 8 ± 4 years. The chief cause of catheter failure in this series was obstruction, followed by dislocation and infection.
Most of the recent studies of PD catheter placement in children focus on the feasibility of the laparoscopic technique [2,3,6-9], with the largest such study comparing 25 laparoscopic PD catheter placements to 23 open catheter placements [2]. That study found no difference in catheter complication or failure rates between laparoscopic and open placement. An advantage of the laparoscopic technique mentioned by all of the above studies is the opportunity to perform a complete omentectomy and to evaluate for the presence of occult inguinal herniation at the time of catheter placement. Unlike other studies evaluating PD catheter placement in children, our investigation (the largest of its kind published to date) had as its principal objective the identification of perioperative factors leading to catheter failure within 2 months after placement. This period was determined to be the ideal end point for assessment of all factors having to do with catheter placement, while minimizing the effect of more chronic factors, such as intraabdominal adhesion formation or infection introduced separately from surgical placement. Admittedly, our study suffers from the limitations inherent in retrospective studies; specifically, the choice of surgical technique, type of catheter placed, decision to perform omentectomy, and placement of catheter exit site were entirely dependent upon the operative surgeon. Several surgeons were consistent in their choices, regardless of the size or age of the patient; others were more selective. No attempt is made in this study to determine the rationale for those choices at the time of surgery. Our study supports the results of Daschner et al [2], showing no advantage of the laparoscopic technique as it pertains to successful catheter function in the early postoperative period. In addition, other factors that may appear important, including previous PD catheter placement, previous abdominal surgery, type of catheter, or location of catheter exit site, did not determine the successful outcome of PD catheter placement. Table 2
Catheter failure within two months of placement
Median age ⁎ Laparoscopic catheters Open catheters Curled catheters Straight catheters Upper abdominal exit site Lower abdominal exit site Previous abdominal surgery Simultaneous omentectomy ⁎⁎ Weight b10 kg ⁎⁎⁎ Weight N10 kg Postoperative infection
Functioning catheter
Catheter failure (n = 36)
13 y 57.6% 42.4% 49.4% 50.6% 51.8% 48.2% 51.8% 72% 10.6% 89.4% 5.9%
9y 52.8% 47.2% 44.4% 55.6% 63.9% 36.1% 63.9% 28% 41.7% 58.3% 8.3%
⁎ P = .002 for age. ⁎⁎ P = .042 vs. omentectomy. ⁎⁎⁎ P b .001 for weight below 10 kg.
Risk factors for early peritoneal dialysis catheter failure in children The most common cause of catheter failure in this study was catheter malfunction owing to occlusion: either from an internal fibrin plug, or from abutment against an external surface, such as omentum, bowel, or other structure. Interestingly, despite the claims of laparoscopic proponents, laparoscopic placement appeared to offer no advantage in preventing this complication in this series. Similarly, curled catheters offered no advantage against occlusion compared to straight catheters. The single controllable factor that was associated with a decreased risk of occlusion was the performance of simultaneous omentectomy; even a history of previous omentectomy appeared to offer no advantage. Our postoperative infection rates were lower than those reported in other large series: it may be that our standard use of perioperative antibiotics may decrease our skin exit site and subcutaneous tunnel infection rates. In addition, parents of children with new PD catheters at our institution receive extensive education regarding the importance of sterile technique during catheter handling to minimize contamination of the catheter. We did find that children weighing less than 10 kg have a higher risk of PD catheter failure, supporting the findings of Beanes et al. [5]. In addition, we found that older children are more likely to have a successfully functioning PD catheter. The most important finding of our study is that the performance of an omentectomy at the time of PD catheter
589
placement was associated with a decreased risk of primary catheter failure.
References [1] Washburn KK, Currier H, Salter KJ, et al. Surgical technique for peritoneal dialysis catheter placement in the pediatric patient: a North American survey. Adv Perit Dial 2004;20:218-21. [2] Daschner M, Gfrorer S, Zachariou Z, et al. Laparoscopic Tenckhoff catheter implantation in children. Perit Dial Int 2002;22:22-6. [3] Mattioli G, Castagnetti M, Verrina E, et al. Laparoscopic-assisted peritoneal dialysis catheter implantation in pediatric patients. Urology 2007;69:1185-9. [4] Rinaldi S, Sera F, Verrina E, et al. Chronic peritoneal dialysis catheters in children: a fifteen-year experience of the Italian Registry of Pediatric Chronic Peritoneal Dialysis. Perit Dial Int 2004;24:481-6. [5] Beanes SR, Kling KM, Fonkalsrud EW, et al. Surgical aspects of dialysis in newborns and infants weighing less than ten kilograms. J Pediatr Surg 2000;35:1543-8. [6] Lessin MS, Luks FI, Brem AS, et al. Primary laparoscopic placement of peritoneal dialysis catheters in children and young adults. Surg Endosc 1999;13:1165-7. [7] Numanoglu A, McCulloch ML, Van Der Pool A, et al. Laparoscopic salvage of malfunctioning Tenckhoff catheters. J Laparoendosc Adv Surg Tech 2007;17:128-30. [8] Bensard DD, Partrick DA, Ford D, et al. Efficacy of laparoscopic peritoneal dialysis catheter placement in children. Pediatr Endosurg Innov Tech 2001;5:241-6. [9] Stringel G, McBride W, Weiss R. Laparoscopic placement of peritoneal dialysis catheters in children. J Pediatr Surg 2008;43:857-60.
www.elsevier.com/locate/jpedsurg
Risk factors for early peritoneal dialysis catheter failure in children Randolph K. Cribbs a , Larry A. Greenbaum b , Kurt F. Heiss a,⁎ a
Division of Pediatric Surgery, Department of Surgery, Emory University, Atlanta, GA 30322, USA Division of Pediatric Nephrology, Department of Pediatrics, Emory University and Children's Healthcare of Atlanta, Atlanta, GA 30322, USA b
Received 5 February 2009; revised 11 June 2009; accepted 12 June 2009
Key words: Children; Peritoneal dialysis catheters; Laparoscopy
Abstract Background: There is uncertainty regarding the optimal approach for surgical placement of peritoneal dialysis (PD) catheters in children. Operative technique, catheter selection, and patient variables (eg, age or prior surgical history) may influence catheter lifespan. Methods: A retrospective review of all PD catheters placed at a tertiary children's medical center during a 6-year period was performed. Our primary outcome was catheter function 2 months after placement. Data were analyzed using Student 2-tailed t test or χ2 analysis. Results: There were 121 PD catheters placed in 81 patients. The median primary functional catheter lifetime was 109 days. Primary PD catheter failure (within 2 months) occurred in 36 catheters (30%). Patients with primary catheter failure (8 ± 7 years) were younger than patients with a functioning catheter at 2 months (12 ± 5 years; P = .002). Catheters placed without simultaneous omentectomy were more likely to fail (P = .042). Catheter failure rate was not significantly different based upon operative technique or catheter type. Conclusion: Omentectomy at the time of catheter placement decreased the risk of early catheter failure. In contrast, type of catheter or laparoscopic placement did not influence the likelihood of early catheter failure. © 2010 Elsevier Inc. All rights reserved.
For children with renal failure, peritoneal dialysis (PD) is the preferred method of dialysis [1]. Compared to hemodialysis, which requires the child to be attached to a dialysis machine at a hospital or outpatient clinic for up to 4 hours a day, 3 days every week, PD can be performed at home by caregivers with a minimum of training.
⁎ Corresponding author. Tel.: +1 404 727 3779; fax: +1 404 727 2120. E-mail address: [email protected] (K.F. Heiss). 0022-3468/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2009.06.019
Placing PD catheters in children poses unique challenges reflected in the high complication rates for children, as high as 70% in some series [1]. The abdominal muscle layers are thinner in children compared to adults, making it difficult to affix the catheter in place, prevent movement of the catheter tip inside the abdominal cavity, and prevent liquid leakage from the abdomen to the skin. Major causes of PD failure are catheter-related infection and malfunction. Abdominal hernias, although not a threat to the PD catheter itself, are also common in children after PD catheter placement [2].
586 Laparoscopy was initially used in adults to rescue blocked PD catheters and, later, for initial PD catheter placement [3]. By facilitating a thorough exploration of the abdominal cavity, permitting identification of coincidental intra-abdominal disease and careful placement of the PD catheter tip, laparoscopy has been suggested to promote longer catheter survival than open placement. However, the optimal placement technique remains uncertain. Other factors that may influence PD catheter lifespan include catheter selection (straight vs. curled), resection of the omentum during catheter placement, and patient variables (e.g., age or prior surgical history). At our institution, open PD catheter placement and laparoscopic-assisted PD catheter placement have been used in nearly equal numbers, based upon individual surgeon preference, since 2002. In addition, the two PD catheter types (straight or curled) have also been employed nearly equally during the period of study. This situation has provided us with a unique opportunity to analyze the short-term outcome of a large number of PD catheter placements in an attempt to identify perioperative factors associated with successful catheter placement.
1. Methods We performed a retrospective chart review of all patients who had operative peritoneal dialysis catheter placement at Children's Healthcare of Atlanta at Egleston, a tertiary children's hospital in the Southeastern United States, from January 2002 to December 2007. Institutional review board approval was obtained prior to record retrieval. Operative and nonoperative variables and dates of initiation and cessation of dialysis were recorded. All catheters had a synthetic cuff to secure to the abdominal wall; catheter tips were placed in the pelvis without additional suture to the pelvic wall. All catheters were tunneled above the fascia to the opposite side of the abdominal wall before exiting at the skin site. The skin exit site was placed in the left upper quadrant in 45 cases (37%), right lower quadrant in 30 cases (25%), left lower quadrant in 24 cases (20%), and the right upper quadrant in 22 cases (18%). In all cases, the external portion of each PD catheter was oriented downward toward the feet at the exit site. The PD catheters were placed by 7 different pediatric surgeons and 1 transplant surgeon: 5 predominantly used a laparoscopic approach; the remaining 3 typically used an open approach. Of 7 surgeons, 6 performed simultaneous omentectomy more frequently during PD catheter placement, whereas 1 surgeon tended not to perform a simultaneous omentectomy. Two surgeons used curled PD catheters exclusively, whereas the remaining surgeons tended to prefer the straight catheters (irrespective of patient age). There was also no consensus regarding exit site placement: 4 surgeons had a predominant site preference, whereas the remaining
R.K. Cribbs et al. surgeons used several different exit sites. Individual surgeons performed a total of 37, 26, 22, 14, 12, 7, 2 and 1 case, respectively. All participating surgeons had prior experience in performing PD catheter placement before the study. Our primary outcome was catheter function 2 months after placement. We excluded PD catheters placed for acute renal failure, those not used for dialysis within 3 weeks of placement, and those without postoperative follow-up. We did not consider a catheter as having failed if it was functioning when removed following successful renal transplantation. Data analysis of successful vs. failed groups was performed using Student 2-tailed t test and χ2 analysis, as appropriate. P b .05 was considered significant. Statistical software was SPSS 15.0 for Windows (SPSS, Chicago, Ill).
2. Results There were 121 PD catheters placed in 81 patients. There was a wide range of patient ages and weights in our cohort (Figs. 1 and 2). Demographic information and surgical characteristics are displayed in Table 1. One laparoscopic placement was converted to an open procedure after a sustained rise in blood pressure was noted with peritoneal insufflation; this case was grouped with open procedures for statistical calculations. Of 121 catheters, 24 were placed in children weighing less than 10 kg (20%); 15 of these catheters were placed in children under the age of 1 year (12%). The mean time from catheter placement to initiation of PD was 4 ± 5 days (median, 2 days). Of PD catheters placed in children younger than 1 year, 9 were straight catheters (60%), and 6 were curled. Simultaneous omentectomies were performed in 8 of these 15 catheter placements (53%). Laparoscopic technique was used to place 7 of the 15 catheters in this group of infants (47%). During the study period, 55 children (68%) had a single PD catheter placement, 14 children (17%) had 2 PD catheter placements, 11 children (14%) had 3 catheter placements, and 1 child (1%) had 5 catheters placed. A total of 15 catheter revisions were performed during the study period; 7 (47%) were successful in achieving a functional PD catheter at the 2-month postoperative end point. There were a total of 36 catheter failures (30%) within 2 months post placement (Table 2). Two children died during the 2-month postoperative period. One child was a 1-monthold male with bilateral renal agenesis who died with a rectal wall perforation discovered 10 days after PD catheter replacement (for leakage around the catheter exit site). This infant presented with dialysate leaking from his rectum, and the PD catheter tip was seen shortly thereafter extending into the infant's rectal vault. The dialysis catheter had been functioning normally from the time of replacement until the day the catheter tip was discovered in the rectum. His parents chose to decelerate care at that point, and the infant died 10
Risk factors for early peritoneal dialysis catheter failure in children
Fig. 1
The ages of the patients at the time of catheter insertion.
days after cessation of dialysis. The other child was a 2month-old male with a congenital diaphragmatic hernia repair who died of a pneumothorax 30 days after placement of a PD catheter. His catheter was functioning at the time of death; he was not considered a catheter failure. The principal cause of failure was catheter occlusion (75% of cases), either from abutment against an external object (eg, omentum, epiploic coli, or fallopian tube) or from formation of a fibrin plug or clot within the catheter. Failure owing to infection of the peritoneal cavity or the subcutaneous tunnel occurred in 4 cases (11%). Dialysate leakage from the skin site led to catheter failure in 2 cases (6%). The remaining causes of catheter failure included one
Fig. 2
587
case each of severe postoperative hemorrhage, inadvertent catheter removal by the patient, and perforation of the rectum (as noted above). Of the 36 catheter failures, 24 were removed without revision (67%), 10 were revised once (28%), 1 catheter was revised twice (3%), and 1 catheter had three revisions (3%) within the 2-month postoperative period. The failed PD catheters were replaced in 21 cases (58%); 7 children were converted to hemodialysis following removal of a failed PD catheter (19%). Children weighing less than 10 kg had a failure rate of 62.5% compared to a failure rate of 21.6% in children weighing at least 10 kg (P b .001). Omentectomy at the time
The weights of the patients at the time of catheter insertion.
588
R.K. Cribbs et al.
Table 1 Patient demographics and surgical characteristics (n = 121) Male sex Median age (range) Median weight (range) Previous abdominal surgery Previous PD catheter Previous omentectomy Straight tipped catheter Laparoscopic placement Omentectomy at time of surgery Mean catheter lifespan (median) Mean revised catheter lifespan (median)
68 (56%) 12 y (7 d-19 y) 38 Kg (2.3-88.2) 67 (55%) 53 (44%) 33 (27%) 62 (52%) 69 (57%) 47 (39%) 177 ± 204 d (109 d) 207 ± 212 d (145 d)
of catheter placement was associated with a lower catheter failure rate (19%) than occurred when no omentectomy was performed (36%; P = .04). However, the failure rate was 27.5% among children with a previous omentectomy and 34% in children without an omentectomy (nonsignificant). The postoperative infection rate of PD catheters placed via laparoscopic technique (9%) did not differ significantly from the infection rate following open PD catheter placement (6%). There was no significant difference in the mean number of days prior to using a catheter when we compared laparoscopic placement (5 days) to open placement (4 days).
3. Discussion In 2004, Washburn et al [1] published the results of a survey of members of the American Pediatric Surgical Association and the Canadian Association of Pediatric Surgeons regarding surgical technique for PD catheter placement. Their results demonstrated significant variability in all aspects of surgical technique, with the most common practice being the placement of a single-cuffed, curled, non– Swan-necked catheter using open technique with a superiorpointing exit site. Routine omentectomy was reported by 59% of respondents. The largest recent series of PD catheter outcomes was a multi-center study from Italy with 503 catheters placed over a 15-year time period [4]. While failure rates improved over the study period, the predominant cause of catheter failure in this series was exit-site and skin tunnel infection, with obstruction in only 5% of catheters. All of these catheters were placed using open technique. Another recent study evaluated the placement of PD catheters in infants and children weighing less than 10 kg [5]. This study showed a higher incidence of catheter failure and other catheter-related complications with smaller-weight children. A study of percutaneously-placed PD catheters, similar in size to our study [6], evaluated 108 PD catheters placed percutaneously in 93 children with an average age of 8 ± 4 years. The chief cause of catheter failure in this series was obstruction, followed by dislocation and infection.
Most of the recent studies of PD catheter placement in children focus on the feasibility of the laparoscopic technique [2,3,6-9], with the largest such study comparing 25 laparoscopic PD catheter placements to 23 open catheter placements [2]. That study found no difference in catheter complication or failure rates between laparoscopic and open placement. An advantage of the laparoscopic technique mentioned by all of the above studies is the opportunity to perform a complete omentectomy and to evaluate for the presence of occult inguinal herniation at the time of catheter placement. Unlike other studies evaluating PD catheter placement in children, our investigation (the largest of its kind published to date) had as its principal objective the identification of perioperative factors leading to catheter failure within 2 months after placement. This period was determined to be the ideal end point for assessment of all factors having to do with catheter placement, while minimizing the effect of more chronic factors, such as intraabdominal adhesion formation or infection introduced separately from surgical placement. Admittedly, our study suffers from the limitations inherent in retrospective studies; specifically, the choice of surgical technique, type of catheter placed, decision to perform omentectomy, and placement of catheter exit site were entirely dependent upon the operative surgeon. Several surgeons were consistent in their choices, regardless of the size or age of the patient; others were more selective. No attempt is made in this study to determine the rationale for those choices at the time of surgery. Our study supports the results of Daschner et al [2], showing no advantage of the laparoscopic technique as it pertains to successful catheter function in the early postoperative period. In addition, other factors that may appear important, including previous PD catheter placement, previous abdominal surgery, type of catheter, or location of catheter exit site, did not determine the successful outcome of PD catheter placement. Table 2
Catheter failure within two months of placement
Median age ⁎ Laparoscopic catheters Open catheters Curled catheters Straight catheters Upper abdominal exit site Lower abdominal exit site Previous abdominal surgery Simultaneous omentectomy ⁎⁎ Weight b10 kg ⁎⁎⁎ Weight N10 kg Postoperative infection
Functioning catheter
Catheter failure (n = 36)
13 y 57.6% 42.4% 49.4% 50.6% 51.8% 48.2% 51.8% 72% 10.6% 89.4% 5.9%
9y 52.8% 47.2% 44.4% 55.6% 63.9% 36.1% 63.9% 28% 41.7% 58.3% 8.3%
⁎ P = .002 for age. ⁎⁎ P = .042 vs. omentectomy. ⁎⁎⁎ P b .001 for weight below 10 kg.
Risk factors for early peritoneal dialysis catheter failure in children The most common cause of catheter failure in this study was catheter malfunction owing to occlusion: either from an internal fibrin plug, or from abutment against an external surface, such as omentum, bowel, or other structure. Interestingly, despite the claims of laparoscopic proponents, laparoscopic placement appeared to offer no advantage in preventing this complication in this series. Similarly, curled catheters offered no advantage against occlusion compared to straight catheters. The single controllable factor that was associated with a decreased risk of occlusion was the performance of simultaneous omentectomy; even a history of previous omentectomy appeared to offer no advantage. Our postoperative infection rates were lower than those reported in other large series: it may be that our standard use of perioperative antibiotics may decrease our skin exit site and subcutaneous tunnel infection rates. In addition, parents of children with new PD catheters at our institution receive extensive education regarding the importance of sterile technique during catheter handling to minimize contamination of the catheter. We did find that children weighing less than 10 kg have a higher risk of PD catheter failure, supporting the findings of Beanes et al. [5]. In addition, we found that older children are more likely to have a successfully functioning PD catheter. The most important finding of our study is that the performance of an omentectomy at the time of PD catheter
589
placement was associated with a decreased risk of primary catheter failure.
References [1] Washburn KK, Currier H, Salter KJ, et al. Surgical technique for peritoneal dialysis catheter placement in the pediatric patient: a North American survey. Adv Perit Dial 2004;20:218-21. [2] Daschner M, Gfrorer S, Zachariou Z, et al. Laparoscopic Tenckhoff catheter implantation in children. Perit Dial Int 2002;22:22-6. [3] Mattioli G, Castagnetti M, Verrina E, et al. Laparoscopic-assisted peritoneal dialysis catheter implantation in pediatric patients. Urology 2007;69:1185-9. [4] Rinaldi S, Sera F, Verrina E, et al. Chronic peritoneal dialysis catheters in children: a fifteen-year experience of the Italian Registry of Pediatric Chronic Peritoneal Dialysis. Perit Dial Int 2004;24:481-6. [5] Beanes SR, Kling KM, Fonkalsrud EW, et al. Surgical aspects of dialysis in newborns and infants weighing less than ten kilograms. J Pediatr Surg 2000;35:1543-8. [6] Lessin MS, Luks FI, Brem AS, et al. Primary laparoscopic placement of peritoneal dialysis catheters in children and young adults. Surg Endosc 1999;13:1165-7. [7] Numanoglu A, McCulloch ML, Van Der Pool A, et al. Laparoscopic salvage of malfunctioning Tenckhoff catheters. J Laparoendosc Adv Surg Tech 2007;17:128-30. [8] Bensard DD, Partrick DA, Ford D, et al. Efficacy of laparoscopic peritoneal dialysis catheter placement in children. Pediatr Endosurg Innov Tech 2001;5:241-6. [9] Stringel G, McBride W, Weiss R. Laparoscopic placement of peritoneal dialysis catheters in children. J Pediatr Surg 2008;43:857-60.