Sacral Osteomyelitis after Single-shot Epidural Anesthesia via  the Caudal Approach in a Child

A 6-yr-old, 130-cm, 20-kg healthy boy (American Society of Anesthesiologists classification I), without preexisting systemic or regional infection, had general anesthesia and a caudal block for circumcision on an outpatient basis. After application of the standard monitoring (pulse oximetry, 5-lead ECG, and noninvasive blood pressure measurement), general anesthesia was induced by inhalational technique (sevoflurane 3 minimum alveolar concentration in 40% N²O with oxygen) and an intravenous access was established. After skin disinfection of the puncture area using a solution of octenidine hydrochloride, the caudal block was performed by single-shot technique with no difficulties. A 22-gauge needle was introduced under aseptic conditions (i.e. , sterile gloves, facemask, and operation cap) at a 45-degree angle without lancing the skin. Bony structures were not grazed. Next, 20 ml of 0.125% Bupivacaine with epinephrine 1:200,000 was injected, providing effective intra- and postoperative analgesia. The anesthetic and surgical procedures were uncomplicated, and the early postoperative period was uneventful.

However, on the third postoperative day, the child presented with immobilizing back pain. He was in good clinical condition and was afebrile. A tender spot in the area of the caudal puncture without signs of local inflammation was detected. Sensibility and motor function of the legs were preserved. Leukocyte count and C-reactive protein level were normal. Blood cultures or local tissue-samples were not taken. Magnetic resonance imaging (MRI) revealed a phlegmonous, subcutaneous inflammation of the sacral region. There was no abscess, but edema at the anteroposterior periosteum of the lower sacrum indicated a beginning osteomyelitis (fig. 1).

Amoxicillin-clavulanic acid 120 mg·kg⁻¹daily initially intravenously for 3 days, then 120 mg·kg⁻¹daily orally for 46 days, was used as empiric therapy based on the clinical estimation of highest probability for Gram-positive organisms by the infectious disease consultant. The MRI was repeated after 17 days of antibiotic therapy, at which time the acute osteomyelitis of the sacrum with signs of a soft-tissue infection was diagnosed (fig. 2). Antibiotic therapy was continued for another 30 days, and the patient showed complete recovery: The back pain disappeared 7 days after the second MRI, and a follow-up MRI showed no signs of infection in the sacral area.

Local inflammation of the skin and the subcutaneous tissue after an epidural puncture or catheter insertion is a frequent event with spontaneous resolution. An incidence of up to 11% has been reported. 4In contrast, serious infections are exceptional. The development of an osteomyelitis, especially after single-shot caudal block, has not been reported previously. Spinal epidural abscess, the most frequent form of serious infection after epidural puncture, has a reported incidence of 0.2 to 1.2 per 10,000 in hospitalized children. 5The occurrence of this complication is equally frequent after short-term catheterization and single epidural injection. 2Moreover, a similar incidence of spontaneous epidural abscess has been reported. 6,7 

Epidermal contamination of the subcutis may be the most probable origin of the serious infection in the current case. However, the hematologic spread of bacteria is the common etiologic factor in a spontaneous epidural infection, 6but the child presented with no history or sign of local or systemic infection. Epidural injection and catheterization inherently carry a risk for bacterial colonization. Recent studies found a 20–35% colonization of all epidural catheters in children by various strains of bacteria after a period of 3 days. Gram-positive colonization was similar for caudal and lumbar catheters, whereas Gram-negative strains were found with a higher incidence on caudal catheters. 4,8Furthermore, the epidural space may be contaminated by skin flora, even after correct disinfection. 9Neither a solution of 0.5% Chlorhexidine nor povidone iodine may be able to completely eradicate bacterial skin flora. Octenidin hydrochloride is registered for skin disinfection but may also lack efficacy in bacterial eradication. In addition, bacteria may accumulate in the hair follicles of the skin 10and then be transported to the deeper tissue by a needle track, which could be prevented by lancing the skin. 11 

Early diagnosis of infections of the sacral region after epidural puncture is difficult. The patient may by asymptomatic or may present with minor symptoms: One case with hip pain as the main symptom was reported. 12In the current case, exclusive back pain led to the diagnosis. Unfortunately, however, the infection is frequently not detected until the development of significant neurologic complications. 5Gadolinium-enhanced MRI is considered to be the definitive standard in the diagnosis of spinal epidural abscess. 12,13However, in our case an early MRI showed relatively minor soft-tissue infection and indirect signs of a beginning osteomyelitis, whereas the diagnosis of osteomyelitis could only be confirmed later.

To avoid subsequent surgical intervention, 12an early empiric antibiotic therapy was begun after clinical indications of an infection. The choice of antibiotics must be directed primarily against Gram-positive organisms, especially Staphylococcus aureus , streptococci, and enterococci, the leading pathogens for tissue infection after epidural puncture. 14Furthermore, antibiotics must provide low toxicity to enable treatment over several weeks, and they should easily penetrate bone tissue. 15Amoxicillin-clavulanic acid met the demands in the current case, particularly because susceptibility of the Gram-positive bacteria strains to this drug is still very high in our hospital. If clinical symptoms fail to resolve with therapy, the presence of methicillin-resistant organisms must be considered 16,17and the responsible bacteria should be identified by culture and Gram strain. In the present case, rapid clinical improvement of the patient's complaints warranted continuation of empiric treatment, hence avoiding further intervention in the sacral region for bacterial analysis with general anesthesia in this child.

In summary, we report a case of sacral osteomyelitis after single-shot epidural anesthesia via  the caudal approach under aseptic conditions in a child. Back pain was the predominant clinical sign of infection, whereas the initial MRI did not show the early osteomyelitis. Empiric antibiotic therapy resulted in complete recovery.