Thoracic and Lumbar Epidural Analgesia via  the Caudal Approach Using Electrical Stimulation Guidance in Pediatric Patients: A Review of 289 Patients

After approval from the institutional ethics board (University of Alberta, Edmonton, Alberta, Canada), we examined data from 289 children who were candidates for caudal placement of a lumbar or thoracic catheter between 1999 and 2002 using data collected prospectively in our pediatric pain service database.

All caudal route epidural catheters were inserted using a commercial product (Epidural Positioning System using the Tsui test; Arrow International Inc., Reading, PA), which used the Tsui test to confirm placement of the epidural catheter (fig. 1) as previously described. 9After induction of anesthesia and tracheal intubation without muscle relaxant, the patients were turned to the lateral position. After sterile preparation, an 18-gauge intravenous catheter was inserted into the caudal space, through which a 20-gauge epidural catheter was threaded. The length of the epidural catheter required was predetermined by measuring the distance on the skin from the caudal space to the desired level for the catheter tip. After priming the catheter and adaptor with 1–2 ml normal saline, an electrical impulse can be conducted through the fluid to the tip of the catheter. An electrical current (1–10 mA) was applied through the epidural catheter as it was advanced cranially. The level of muscle twitch (unilateral or bilateral) was observed to advance from the lower limb muscles to the upper abdominal muscles and/or intercostal muscles as the catheter was threaded cranially. Minor resistance to the passage of the catheter was overcome by simple flexion or extension of the infant’s vertebral column and by the injection of a small amount of normal saline through the advancing epidural catheter. The catheter was pulled back and reinserted if it did not reach the desired level. When the catheter was optimally positioned, the 18-gauge intravenous catheter was withdrawn from the caudal space, and the stylet was removed from the epidural catheter. The catheter was affixed immediately cephalad (fig. 2) with several layers of transparent occlusive dressing. The choice of the epidural infusion was left to the anesthesiologist. Patients did not leave the recovery room without evidence of good analgesia as determined by the intraoperative anesthesiologist and/or without radiologic confirmation of the position of the epidural catheter tip. The definition of technical success was the placement of the epidural catheter into a satisfactory location as assessed by the intraoperative attending anesthesiologist. Epidural catheters that did not reach the determined level during the initial insertion were noted.

The pediatric pain service database was initiated to prospectively collect information on all children receiving epidural analgesia, as well as all pertinent information regarding the placement and characteristics of each epidural catheter. This included information such as the method of insertion, the final level of the catheter, and the medication administered. Adverse effects, such as respiratory depression, nausea and vomiting, sedation, urinary retention, pruritus, skin irritation, and excessive neurologic blockade, were also noted. Our pediatric acute pain service team set the criteria used to assess these adverse effects. Any child who needed supplemental oxygen for reduced saturation or had a respiratory rate less than 10 breaths/min was considered to have respiratory depression. Significant respiratory depression necessitating naloxone was also recorded. Sedation was scored on a three-point scale: 0 = none (awake and alert); 1 = mild (occasionally drowsy, easy to rouse); 2 = moderate (frequently drowsy, easy to rouse); and 3 = severe (somnolent, difficult to rouse). Patients with a score of 3 were considered to have sedation. Urinary retention was monitored in those patients who did not have a Foley catheter placed during their surgery. Urinary retention was documented if any patient had difficulty with urine excretion, according to the nursing staff, that may or may not have necessitated a catheter for treatment. Any child treated with an antiemetic was considered to have nausea and vomiting, and if given an antipruritic agent, the child would be considered to have pruritus. Skin irritation was noted if the insertion site of the epidural catheter was red and irritated, and excessive block was recorded if the patient’s extremities were numb or immobilized because of the epidural.

Patients were seen twice daily by the pediatric pain service nurse, who assessed analgesia with specific tools: behavioral for infants and cognitively impaired children, the Faces scale for young children, and the visual analog scale for children older than 7 yr. Residents were not involved with the care of these patients, and analgesic decisions were made by the attending acute pain pediatric anesthesiologist responsible for service that week. The acute pain attending anesthesiologist determined analgesic success. Epidural analgesia was considered to be unsuccessful if adequate analgesia was not possible with this method. In such cases, the epidural analgesic was discontinued, and pain was managed with parenteral opioids. Reasons for abandonment of the epidural technique were recorded. Epidural analgesics supplemented with opioid were also noted.

Statistical analysis was performed with chi-square and analysis of variance for comparing analgesic success rates between patients aged younger than 1 yr and patients aged older than 1 yr. A P  level less than 0.05 was considered statistically significant.

Between January 1999 and December 2002, a total of 289 caudally placed lumbar or thoracic catheters were attempted in our pediatric patient population. In five patients (aged 5 months–1.6 yr), the catheter did not thread to the desired level and was abandoned in the operating room (technical success rate, 98.2%). Of the remaining 284 patients, the median age of the patient group was 0.7 yr (interquartile range, 0.3–1.5 yr), with a median weight of 7.2 kg (interquartile range, 4.7–10.5 kg). Clinical characteristics of the patients are listed in table 1. One incident of intravascular placement was noted after blood was aspirated from the catheter during the initial insertion. Consequently, the epidural catheter was readjusted until blood could no longer be aspirated in the operating room.

Of the remaining 284 patients (table 1), 259 (91.2%) received an initial infusion rate of 0.2- 0.4 ml · kg⁻¹· h⁻¹bupivacaine (0.05–0.125%) and 1–2 μg/ml fentanyl, whereas 16 received bupivacaine alone, and 9 received infusions including epidural morphine. The overall analgesic success rate of all caudal route epidural analgesia procedures was 84.9%. Twenty-nine patients received continuous epidural analgesic supplemented with morphine infusion (28 patients received 5–20 μg · kg⁻¹· h⁻¹and 1 patient received up to 60 μg · kg⁻¹· h⁻¹). There was no significant difference in adequate pain control (analgesic success rate) in infants (aged 1 day–1 yr) versus  older children (aged older than 1 yr) (fig. 3).

Pruritus, as defined by the use of an antipruritic, was the most common adverse effect noted in our patient population (26.1%). Nausea and vomiting was reported in 16.9% of the patient population. Of the patients in our study, 57.7% had urinary catheters in situ ; of those who did not have a catheter placed, 20.8% experienced urinary retention. Sedation and excessive block each occurred in less than 2% of the patients. The incidence of respiratory depression was 4.2%, but the administration of naloxone for severe respiratory depression was never necessary. Technical problems relating to the epidural analgesic administration were divided into three categories: connector problems (6.3%), catheter leak (6.3%), and catheter occlusion (1.4%). The occurrence of each did not always necessitate discontinuation of the epidural analgesic, in fact; only five epidurals were abandoned because of technical problems. Frequencies of the most common adverse effects are listed in table 2.

Major adverse effects were rare. No patients developed epidural abscesses or neurologic deficits related to the epidural analgesic, although one case of a possible seizure in a severely spastic cerebral palsy patient was reported in the postanesthetic recovery room. This epidural analgesic was immediately discontinued and changed to parenteral opioid in the postanesthetic recovery room without further complications. One dural puncture during the epidural catheter insertion in the operating room was also noted. The epidural catheter was reinserted after the dural puncture was detected and used for 3 postoperative days, with no additional complications.

Placing epidural catheters via  the caudal route may be more desirable than direct insertion because of the potential risks of spinal cord damage associated with direct placement. 2–5The success of the Tsui test in placing thoracic and lumbar epidural catheters via  the caudal route has been documented. 9However, the routine use of this technique, the adverse effects, and the analgesic success of the technique have not been reported.

The commercially available Epidural Positioning System using the Tsui test with FlexTip Plus Catheter (fig. 1) has been used routinely in our institution since 1999. The ability to identify the catheter tip position while performing the procedure allows adjustments to be made at the time of insertion. Another important feature of this setup is that it allows the simultaneous use of a stylet and saline injection to permit catheter advancement. The original design of the Arrow Flextip Plus catheter has a flexible and soft tip. This Arrow design uses a soft polyurethane polymer, which has greater elasticity than the nylon used in other catheters. 10Because the stylet ends 10 mm proximal to the tip of the epidural catheter, it is possible for the tip of the catheter to fold back on itself during insertion in a “J” configuration. This feature creates a soft and blunted tip of the catheter, while the stylet wire provides stiffness for ease of advancement from the caudal space. Minor resistance to the passage of the catheter can be overcome by simple flexion or extension of the infant’s vertebral column and, perhaps more importantly, by the injection of a small amount of normal saline through the advancing epidural catheter. By injecting normal saline, the catheter is lubricated and the epidural space is dilated to allow the catheter to advance more easily. Although the time of placing the epidural catheter was not recorded in the database, the average time required to place the epidural was found to be 14 min in a previously published prospective study. 9 

The combination of the injected saline and the soft tip of the catheter may have contributed to the extremely low incidence of intravenous placement. 11Although only one case was documented of positive aspiration of blood while manipulating the catheter, slightly blood-stained saline at the insertion site after manipulating the catheter is frequently observed. This observation may not have always been recorded in the database and represents a possible limitation of this study. Puncture of an epidural vein is not uncommon even with direct epidural placement. 12Fortunately, the bleeding usually stops within a short time and without any consequences except when a local anesthetic is directly injected into an epidural vein or when the vein is punctured in patients with coagulation defects. It is our common practice to avoid placing epidural catheters via  the caudal space in patients with known increased risk factors for prolonged bleeding.

For the five patients in whom we were unable to advance the catheter cranially despite repeated attempts, parenteral opioids were given for postoperative pain. This may have reduced unnecessary exposure of our patients to discomfort and risk of toxicity of local anesthetic with suboptimally positioned catheters. In the study, a high overall technical success of 98.2% (284 of 289) of correctly placed epidural catheters via  the caudal space using the Tsui test was noted. Valairucha et al.  suggested that the development of the lumbar curve during infancy prevents easy threading of the catheter and reduces the success of the epidural catheter placement. 8For this reason, some studies suggest that caudal placement should be used for patients who are younger than 1 yr. Between 1999 and 2002, 192 of the 284 caudal epidural catheters placed at our institution were in patients 1 yr and younger; the remaining 92 were placed in children older than 1 yr (fig. 3). Both groups had similar analgesia success rates of approximately 85%. This challenges the idea that threading the catheter is only feasible in younger patients. Indeed, two adolescent patients, aged 14 and 17 yr, with severe scoliosis had caudal catheters threaded to avoid potential difficulty with direct epidural placement. However, patient selection may have been biased toward smaller patients for the caudal approach, whereas direct thoracic and lumbar approaches were chosen for larger patients. Nevertheless, we suspect that our high technical success rate in larger patients was likely due to the epidural setup used with the stylet catheter, simultaneously allowing injection of saline during advancement, and, more importantly, the Tsui test, monitoring the process of the advancement of the catheter tip.

Despite satisfactory placement of the catheters, 28 of the patients received continuous epidural analgesics with supplementary low-dose morphine infusions (5–20 μg · kg⁻¹· h⁻¹) for sedation. Most of these infusions were planned for patients who were agitated preoperatively. In addition, one patient was irritable even with 60 μg · kg⁻¹· h⁻¹concurrent morphine, and the symptoms were managed only after midazolam and baclofen were added. Because sedation is not a feature of epidural analgesia, children may seem agitated and irritable because of nausea, hunger, the presence of nasogastric tubes and/or intravenous and urinary catheters, or general discomfort of being in a strange environment, rather than because of incisional pain. However, whenever there was doubt as to the analgesic efficacy of the epidural analgesic in small infants or nonverbal patients, we elected to remove the epidural catheters and provide parenteral opioids.

Our overall analgesic success of caudally placed epidural catheters for all age groups was 84.9% (i.e. , 15.1% were abandoned during the postoperative period), despite a high proportion of painful upper abdominal and thoracic surgeries. Our choice of an operational definition of analgesic success is grounded in a realistic approach to these patients and may be applicable to most hospitals with similar acute pain services. The ability to run such a service without physicians in-house at night indicates the effectiveness and low maintenance of this postoperative analgesic regimen.

The most common adverse effects after neuraxial opioid administration include pruritus, urinary retention, and nausea and vomiting. These adverse effects also seem to be dependent on the hydrophilic and lipophilic activity of the drug used. Fentanyl, a lipophilic opioid, was the primary agent used during the study period. Our young patient population (mean age, 1.5 yr) presented some difficulties when assessing and reporting adverse effects because pharmacologic interventions, such as antipruritics and antiemetics, are often used as prophylactic agents in infants and nonverbal children. We elected to err on the side of overreporting adverse effects and recorded all uses of antinausea and antipruritic agents as adverse effect occurrences. Despite this drawback, the incidence of adverse effects in our patients was low and comparable with that of other studies. 13–17 

The two most frequent adverse effects experienced were pruritus (26.1%) and urinary retention (20.8%). Pruritus related to epidural analgesia in pediatric patients has been reported to be as low as 5.3% and as high as 88%. 13,14The wide range of reported pruritus likely reflects the inability of clinicians to accurately distinguish the cause of infant irritability. 18The incidence of urinary retention was 20.8% (25 of 120) in patients who did not have a Foley catheter placed in the operating room, but only 8% (2 of 25) of these patients needed subsequent catheterization. Because approximately 57% of the patients had a Foley catheter placed prophylactically in the operating room, it is difficult to determine the true incidence of urinary retention related to the epidural analgesic. Wood et al.  13reported the incidence of urinary retention as 6.3%, whereas Dalens et al.  14reported it to be 36%, although neither study reported how many of these patients needed subsequent catheterization. Nausea and vomiting occurred in only 16.9% of the patients, which is low compared with other studies in which either epidural morphine (22–40%) 14,15or bupivacaine and fentanyl (23%) 13were used.

One of the most serious adverse effects related to epidural analgesia is respiratory depression. Wilson et al.  16reported the incidence of major respiratory complications in pediatric patients who had undergone fundoplication with epidural analgesia to be 11%. Only 4.2% of our total patient population and 1.8% of patients who had undergone fundoplication needed supplemental oxygen or had a respiratory rate less than 10 breaths/min. None needed rescue naloxone. Excessive blockade due to epidural analgesia can be unpleasant and distressing for small children and could potentially cause major problems for patients with regard to pressure sores. 13Only 1.8% of our patients experienced excessive blockade, but none needed medical treatment for related skin lesions. Oversedation also occurred in only 1.8% of the patients and was treated by a decrease in the epidural infusion rate. We believe the aid of the epidural stimulation test to correctly place the epidural catheter at the level of the incision provided adequate pain relief, with a reduced incidence of adverse effects. The choice of infusion medication could have also attributed to our low adverse effect profile. Epidural morphine use in our patient population was only 1.8%, whereas the use of bupivacaine and fentanyl was 91.2%. This reflects our confidence with nerve stimulation guidance to properly place epidural catheters because good pain relief with fentanyl is general limited to the immediate dermatomes surrounding the catheter tip. 19 

Compared with direct thoracic or lumbar approaches, the potential for contamination with anal bacteria has made the caudal approach a concern for some anesthesiologists. Although studies have not found clinical evidence of higher infection rates with the caudal approach, bacterial colonization is greater with catheters inserted caudally as opposed to the lumbar or the thoracic approach. 20Improving the care of the site of the caudally placed epidural catheter may reduce the risk of contamination. We were particularly vigilant and meticulous in the management of the catheters and of their insertion sites. Using an aseptic technique, we secured the catheter using several layers of a transparent and occlusive dressing, such as Tegaderm (3M Health Care, St. Paul, MN), to cover the insertion site. The catheter was affixed immediately cephalad without making a loop (fig. 3) to maximize the distance of the catheter from the anus and to reduce the risk of contamination by stool and urine. If there was any question of contamination, the catheter was promptly removed. During the study period, 2.5% of the patients had their catheters removed because of suspected contamination. We did not have any cases of epidural abscesses or systemic infection related to caudal epidural catheter placement.

The results of this study suggest that epidural catheter placement via  the caudal approach using the Tsui test is an effective and reasonable alternative to direct lumbar and thoracic epidural analgesia in pediatric patients. This easily performed caudal epidural approach using the Tsui test can optimize positioning of the epidural catheter tip, which in turn leads to effective pain control.

The authors thank Stephan Malherbe, M.B., Ch.B. (Pediatric Anesthesia Fellow, Department of Anesthesiology and Pain Medicine, University of Alberta, Edmonton, Alberta, Canada), and Leeann Philips, B.Sc.N., and Kathy Reid B.Sc.N. (Pediatric Acute Pain Service Nurses, Department of Anesthesiology and Pain Medicine, and Pediatric Anesthesiologists, University of Alberta Hospital), for their contribution.