Complications of Long-Term Jejunostomy in Children By David Smith and Pierre Saucy Ottawa, Ontario 0 The authors reviewed 64 jejunostomies performed in 57 patients. Data were collected regarding complications and performance of the catheters. Patient diagnoses were grouped as follows: cystic fibrosis (25), neurological impairment (14). and miscellaneous other (25). Indications were malnutrition (43), inability to feed (17), and gastroesophageal reflux (4). Complications were compared between these groups. The age range was 7 days to 23 years. There were 251 tube changes over 142 years of cumulative site patency, for an average of 1.8 tube changes per year and an average life of 2.2 f 2.4 years per site. The longest duration was 11.7 years. Four tube changes resulted in intraperitoneal insertion (6.2% of changes). The overall complication rate was 37.5%. The major and minor complication rates were 21.9% each. Some patients had more than one complication. Stratification of complications by diagnosis showed that the highest inci- dence was among the neurologically impaired children (64%), followed by those with cystic fibrosis (32%) and then others (28%). Sixty-four percent of major and 54% of minor compli- cations occurred within the first 6 months. The mortality rate was 4.7%. Infections requiring intravenous antibiotics oc- curred in 9.4% of the sites, at an average site age of 8.7 I- 7.7 months. Tube dislodgment requiring surgical replacement occurred in 9.4% of the patients. Our mortality and complica- tion rates compare favorably to those of previously reported series. Surgical jejunostomy is a reliable long-term solution to feeding but is associated with a significant risk of compli- cations, especially in neurologically impaired children. The risk is greatest in the first 6 months after insertion, then decreases as the site “matures.” Copyright o 7996 by W. 6. Saunders Company INDEX WORDS: Jejunostomy, complications. C HILDREN who have severe neurological impair- ment, cystic fibrosis, multiple trauma, and com- plex esophageal malformations often require adjunc- tive enteral feeding. Surgical jejunostomy is one option to provide enteral access. Methods for perform- ing surgical jejunostomy include percutaneous endo- scopic jejunostomy (PEJ), transgastric jejunostomy through a preexisting gastrostomy, Roux-en-Y jejunos- tomy, needle jejunostomy, catheter jejunostomy, and laparoscopic-assisted jejunostomy. Each of these pro- cedures has inherent complications. Transgastric tubes are associated with a high rate of dislodgment, obstruction, and tube leakage. Kaplan et al,’ in 1989, reported that the failure rate of PEJ tubes was 84% at a mean of 39 days. They noted a 22% incidence of aspiration pneumonia and a 30% incidence of upper gastrointestinal bleeding. There were no catheter-related deaths. Wolfsen et a1,2 in 1990, reported that 41% of PEJ tubes required replacement within 2 months of placement. Their incidence of aspiration pneumonia was 17% in the Journa/ofPed/atr/c Surgery, Vol31, No 6 (June), 1996: pp 787-790 first 30 days after placement, and there were no catheter-related mortalities. Roux-en-Y jejunostomies prevent mechanical tube problems by creating a catheterizable stoma. In 1952, Brintnall et al3 reported prolapse of the stoma in 15% of patients and leakage of bile and pancreatic juice in 6%. In 1993 DeCou et al4 reported no complications among six patients with Roux-en-Y limbs closed beneath the fascia over a replaceable catheter. One 30-day death occurred in DeCou’s series, and none occurred in Brintnall’s series. Catheter jejunostomies have the best-character- ized complications of all surgical jejunostomies. A review of several series5-l1 showed a catheter-related mortality rate of lo%, 30-day overall mortality rates of 20% to 80%, overall morbidity rates of 40% to 50%, wound infection rates of 0% to 8%, and catheter dislodgment rates of 0% to 7%. All these series have relatively short follow-up periods and include only adult patients. The majority were placed during major laparotomy for other indications. We perform a permanent catheter jejunostomy with a replaceable catheter at the Children’s Hospital of Eastern Ontario (CHEO). Herein we results of jejunostomies placed during period. MATERIALS AND METHODS Patient Selection report the a 1Zyear Patients who underwent catheter jejunostomy placement be- tween January 1, 1982 and January 1, 1994 were enrolled in the study. All records, including those of inpatient admissions and outpatient visits. were reviewed. Phone interviews were performed for patients who had been discharged from the Children’s Hospital or whose care was transferred elsewhere. The surgical technique for catheter placement was similar among all the surgeons. A loop of jqunum, approximately 20 cm from the ligament of Treitz, was used. A Witzel tunnel of 2 to 4 cm in length was fashioned, and the loop of jejunum was fixed in a longitudinal orientation by three or four sutures. The replaceable Entriflex catheter (Sherwood Medical, St Louis, MO) was secured From the Children S Hospital of Eastern Ontario, Ottawa, Ontario. Presented at the 27th Annual Meeting of the Canadian Association of Paediatk Surgeons, Montreal, Quebec, September 2-4, 1995. Address reprint requests to Pierre Saucy, MD, FRCSC, Department of Surgery, Children’s Hospital of Eastern Ontario, 401 Smyth Rd, Ottawa, Ontario, Canada KIH 8Ll. Copyright o 1996 by WB. Saunders Company 0022-3468/96/3106-0014$03.OOlO 787
Transcript
Complications of Long-Term Jejunostomy in Children By David Smith and Pierre Saucy
Ottawa, Ontario
0 The authors reviewed 64 jejunostomies performed in 57 patients. Data were collected regarding complications and performance of the catheters. Patient diagnoses were grouped as follows: cystic fibrosis (25), neurological impairment (14). and miscellaneous other (25). Indications were malnutrition (43), inability to feed (17), and gastroesophageal reflux (4). Complications were compared between these groups. The age range was 7 days to 23 years. There were 251 tube changes over 142 years of cumulative site patency, for an average of 1.8 tube changes per year and an average life of 2.2 f 2.4 years per site. The longest duration was 11.7 years. Four tube changes resulted in intraperitoneal insertion (6.2% of changes). The overall complication rate was 37.5%. The major and minor complication rates were 21.9% each. Some patients had more than one complication. Stratification of complications by diagnosis showed that the highest inci- dence was among the neurologically impaired children (64%), followed by those with cystic fibrosis (32%) and then others (28%). Sixty-four percent of major and 54% of minor compli- cations occurred within the first 6 months. The mortality rate was 4.7%. Infections requiring intravenous antibiotics oc- curred in 9.4% of the sites, at an average site age of 8.7 I- 7.7 months. Tube dislodgment requiring surgical replacement occurred in 9.4% of the patients. Our mortality and complica- tion rates compare favorably to those of previously reported series. Surgical jejunostomy is a reliable long-term solution to feeding but is associated with a significant risk of compli- cations, especially in neurologically impaired children. The risk is greatest in the first 6 months after insertion, then decreases as the site “matures.” Copyright o 7996 by W. 6. Saunders Company
INDEX WORDS: Jejunostomy, complications.
C HILDREN who have severe neurological impair- ment, cystic fibrosis, multiple trauma, and com-
plex esophageal malformations often require adjunc- tive enteral feeding. Surgical jejunostomy is one option to provide enteral access. Methods for perform- ing surgical jejunostomy include percutaneous endo- scopic jejunostomy (PEJ), transgastric jejunostomy through a preexisting gastrostomy, Roux-en-Y jejunos- tomy, needle jejunostomy, catheter jejunostomy, and laparoscopic-assisted jejunostomy. Each of these pro- cedures has inherent complications.
Transgastric tubes are associated with a high rate of dislodgment, obstruction, and tube leakage. Kaplan et al,’ in 1989, reported that the failure rate of PEJ tubes was 84% at a mean of 39 days. They noted a 22% incidence of aspiration pneumonia and a 30% incidence of upper gastrointestinal bleeding. There were no catheter-related deaths. Wolfsen et a1,2 in 1990, reported that 41% of PEJ tubes required replacement within 2 months of placement. Their incidence of aspiration pneumonia was 17% in the
Journa/ofPed/atr/c Surgery, Vol31, No 6 (June), 1996: pp 787-790
first 30 days after placement, and there were no catheter-related mortalities.
Roux-en-Y jejunostomies prevent mechanical tube problems by creating a catheterizable stoma. In 1952, Brintnall et al3 reported prolapse of the stoma in 15% of patients and leakage of bile and pancreatic juice in 6%. In 1993 DeCou et al4 reported no complications among six patients with Roux-en-Y limbs closed beneath the fascia over a replaceable catheter. One 30-day death occurred in DeCou’s series, and none occurred in Brintnall’s series.
Catheter jejunostomies have the best-character- ized complications of all surgical jejunostomies. A review of several series5-l1 showed a catheter-related mortality rate of lo%, 30-day overall mortality rates of 20% to 80%, overall morbidity rates of 40% to 50%, wound infection rates of 0% to 8%, and catheter dislodgment rates of 0% to 7%. All these series have relatively short follow-up periods and include only adult patients. The majority were placed during major laparotomy for other indications.
We perform a permanent catheter jejunostomy with a replaceable catheter at the Children’s Hospital of Eastern Ontario (CHEO). Herein we results of jejunostomies placed during period.
MATERIALS AND METHODS
Patient Selection
report the a 1Zyear
Patients who underwent catheter jejunostomy placement be- tween January 1, 1982 and January 1, 1994 were enrolled in the study. All records, including those of inpatient admissions and outpatient visits. were reviewed. Phone interviews were performed for patients who had been discharged from the Children’s Hospital or whose care was transferred elsewhere.
The surgical technique for catheter placement was similar among all the surgeons. A loop of jqunum, approximately 20 cm from the ligament of Treitz, was used. A Witzel tunnel of 2 to 4 cm in length was fashioned, and the loop of jejunum was fixed in a longitudinal orientation by three or four sutures. The replaceable Entriflex catheter (Sherwood Medical, St Louis, MO) was secured
From the Children S Hospital of Eastern Ontario, Ottawa, Ontario. Presented at the 27th Annual Meeting of the Canadian Association
of Paediatk Surgeons, Montreal, Quebec, September 2-4, 1995. Address reprint requests to Pierre Saucy, MD, FRCSC, Department
of Surgery, Children’s Hospital of Eastern Ontario, 401 Smyth Rd, Ottawa, Ontario, Canada KIH 8Ll.
Copyright o 1996 by WB. Saunders Company 0022-3468/96/3106-0014$03.OOlO
787
to the skin with a nonabsorbable suture and Hypafix dressing (Smith & Nephew, Lachine, Quebec). A detailed review of the preoperative preparation, anesthesia, surgical technique, and post- operative care used at CHEO has been reported previously.r2
Data Collection Data were collected separately for each jejunostomy site placed.
The information recorded included patient demographics, patient diagnosis, indication for insertion of tube, complications (including age of site at time of occurrence, number of tube changes, including any radiographic confirmation of location of tube), and length of time the site was used.
Data Analysis Sitepatency. The average number of catheter changes per year
per jejunostomy site was calculated. The percentage of changes confirmed by contrast radiological studies was calculated, as was the percentage of changes resulting in intraperitoneal insertion.
Complications. Complications were classified as major if they required surgical correction or were deemed life-threatening. All other complications were considered minor. Complications were stratified for comparison and plotted by age of patient at time of occurrence, age of catheter site at time of occurrence, patient diagnosis, and indication for jejunostomy placement.
Mortalities. Cause of death and age of catheter site were analyzed for each occurrence. Both 30-day mortality (deaths within 30 days of insertion) and catheter-related mortality (deaths result- ing from complications caused directly by the jejunostomy) were calculated.
Infections. The average age of the catheter site at time of infection was calculated, and the most common organisms cultured were tabulated.
DisZodgments. The average age of the catheter site at time of dislodgment was calculated. Dislodgments were stratified as early ( < 30 days) or late (> 30 days).
RESULTS
A total of 64 jejunostomies were placed in 57 patients over the study period. The patients’ age range was 7 days to 23 years. Follow-up was achieved for all patients, and the mean time follow-up period was 6.4 years after tube placement.
The most common diagnosis at the time of tube placement was cystic fibrosis (39%) followed by neurological impairment (22%) and a variety of less common diagnoses. These are summarized in Table 1. Diagnoses were consolidated into three groups for comparison: cystic fibrosis (2.5) neurological impair- ment (14), and other (25).
Indications for insertion of jejunostomy were mal- nutrition in orally feeding patients (43), inability to feed (17), and gastroesophageal reflux (4).
The cumulative site patency for all sites was 142 years; the longest duration was 11.7 years, in an older cystic fibrosis patient. The shortest duration was 3 days. The mean duration of jejunostomy use was 1.1 + 2.4 years. This was highest for the cystic fibrosis group (3.8 f 2.8 years). During this beriod, 251 tube changes were documented, an average of 1.8 changes per patient per year. Four changes resulted in intra-
SMITH AND SOUCY
Table 1. Diagnoses
Dlagnosls No of Patients Percentage
Cystic fibrosrs 25
Neurological rmparrment 14
Wade gap esophageal atresia 4
Trauma 4
Complications related to premature birth 4
Cancer 3
Complex cardiac anomaly 3
Immune deficiency 3
Laryngoesophageal cleft 2
Apert’s syndrome 1
Liver cirrhosrs 1
Total 64
39 22
6 6
6 5 5
5 3
2 2
100
peritoneal insertions (1.6%). Contrast radiological studies were performed to confirm tube location in 41 of the tube-change incidents (16%).
Three patients died of catheter-related complica- tions or in the immediate postoperative period, yield- ing a mortality rate of 4.7%. The first of these was a 13-year-old neurologically impaired boy who had adhesions and a small bowel volvulus away from the jejunostomy site, which caused extensive bowel necro- sis. Surgery was recommended early in his hospitaliza- tion, but was refused by his parents. The second death occurred in a 4-year-old neurologically im- paired child and was attributable to small bowel volvulus around the jejunostomy site. Treatment was delayed because of transfer from an outlying hospital, and the entire small bowel was nonviable. The third patient, a 1 month old, had failure to thrive because of Salmonella sepsis and died on the third postopera- tive day.
A total of 32 complications were identified in 24 different jejunostomy sites, for an overall complica- tion rate of 37.5%. Seventeen major complications occurred in 14 different sites, yielding a major compli- cation rate of 21.9%. The major complications are summarized in Table 2. Similarly, 15 minor complica- tions occurred in 14 sites, for a minor complication rate of 21.9%. Both a major and a minor complication occurred at four sites. The minor complications are summarized in Table 3.
Stratification of all complications by patient age, comparing percentages of complications for each age
group, produced a bell-shaped curve with the maxi- mum incidence being in the l- to 5-year age group. Sixty-four percent of patients in this age group suffered a complication. Neonates had the lowest incidence of complications (14%).
Stratification of complications by diagnosis demon- strated that there were complications in 32% of patients with cystic fibrosis, 28% of those with “other” diagnoses, and 64% of those with neurological impair- ment. Table 4 shows the specific complication rates for each of these groups.
Stratification of complications by indication for jejunostomy placement showed that complications occurred in 35% of patients with malnutrition, 41% of those who were unable to feed, and 50% of those with reflux.
Stratification of complications by age of catheter site showed that 56% of complications occurred within the first 6 months of placement. Over time the number of complications decreased until 60 months after placement, following which no complications occurred.
A mild skin rash occurred at all jejunostomy sites, but infections requiring antibiotics occurred in only 9.4% of sites, at an average age of 8.7 + 7.7 months. The most frequently cultured organisms were Pseudo- monas aeruginosa and Staphylococcus aureus.
Tubes became dislodged in 30% of the jejunostomy sites, 9.4% of which required surgical replacement. Some sites had frequent dislodgment (up to eight times). Twenty percent of dislodgments occurred within the first 30 days of placement.
enteral access, up to 11.7 years in the present series. Tubes can be replaced easily and safely if they become blocked or dislodged. Only 1.6% of tube insertions resulted in intraperitoneal insertions. Cath- eter-related mortality occurred infrequently and may have been preventable altogether had the patients been referred earlier and undergone laparotomy in a timely fashion. Neurologically impaired patients had the highest mortality and complication rates-14% and 64% (respectively) in the present series. They also had the highest intraperitoneal insertion rate for catheter changes-21% versus 4% for patients with cystic fibrosis, and 0% for other patients. This high rate probably is related to the tortuosity of the tracts between the skin and bowel seen in neurologically impaired patients. The tortuosity may be a result of abdominal movements in these patients. Patients with cystic fibrosis had the highest infection rate, which explains the frequency with which Pseudomo- nas was cultured. The frequency of complications decreases as the site matures, and reaches a safe plateau at 5 years.
We found no comprehensive review of pediatric surgical jejunostomies in the literature. Our series compares favorably with the collected adult series: catheter-related mortality rate, 4.7% versus 5% to 10%; 30-day mortality rate, 1.6% versus 10% to 60%; morbidity rate, 37.5% versus 40% to 50%; infection rate, 9.4 versus 0% to 8%; and dislodgment requiring surgical replacement, 9.4 versus 0% to 7%.
DISCUSSION
Surgical jejunostomy using a replaceable tube in a permanent site represents a long-term solution for
We recognize the potential for complications asso- ciated with surgical jejunostomy. The benefits of long-term reliable usage outweigh the risks for most patients except perhaps those who are neurologically impaired.
4. DeCou JM, Shorter NA, Karl SR: Feeding Roux-en-Y jejunostomy in the management of severely neurologically im- paired children. J Pediatr Surg 28:1276-1279,1993
790 SMITH AND SOUCY
5. Al-Shehri M, Makarewicz P, Freeman JB: Feeding jejunos- tomy: A safe adjunct to laparotomy. Can J Surg 33:181-184,199O
6. Sarr MG, Mayo S: Needle catheter jejunostomy: An unappre- ciated and misunderstood advance in the care of patients after major abdominal operations. Mayo Clin Proc 63:565-572,198s
7. Smith-Choban P, Max MH: Feeding jejunostomy: A small bowel stress test? Am J Surg 155:112-117, 1988
8. Weltz CR, Morris JB, Mullen JL: Surgical jejunostomy in aspiration risk patients. Ann Surg 215:140-145, 1992
9. Adams MB, Seabrook GR, Quebbeman EA, et al: Jejunos- tomy: A rarely indicated procedure. Arch Surg 121:236-238, 1986
10. Matino JJ: Feeding jejunostomy in patients with neurologic disorders. Arch Surg 116:169-171,1981
11. McGonigal MD, Lucas CE, Ledgerwood AM: Feeding jejunostomy in patients who are critically ill. Surg Gynecol Obstet 168:275-277,1989
12. Boland MP, Patrick J, Stoski DS, et al: Permanent enteral feeding in cystic fibrosis: Advantage of a replaceable jejunostomy tube. J Pediatr Surg 22:843-846,1987
