A. J. Cunningham1, M. Boulos1, K. V. McClellan1, S. Krishnaswami1, N. A. Hamilton1 1Oregon Health And Science University,Division Of Pediatric Surgery, Department Of Surgery,Portland, OR, USA
Introduction:
Technical complications following image-guided central venous line (CVL) placement in children are uncommon, occurring in 1.3-1.6% of cases. Severe complications requiring intervention are even less frequent. Despite this and the accuracy of modern, image-guided line insertion techniques, the practice of routine postoperative chest x-ray to detect occult complications persists. This study aims to investigate the utility of this practice.
Methods:
A retrospective review was conducted of all CVLs placed by 15 pediatric surgeons at 2 institutions from January 1, 2010 through June 30, 2016. Demographic data, vessel accessed, use of intraoperative imaging and technical complications were analyzed. Statistical analysis was performed using analysis of variance as appropriate for demographic data and Fisher’s exact test for the relationship of postoperative imaging to technical complications.
Results:
During the period of analysis, 1102 lines were placed in 937 patients. Demographic data and technical complications are shown in Figure 1. Fluoroscopy was utilized in 1078 lines (97.8%) and 556 were ultrasound-guided, internal jugular placements (50.4%). There were 914 postoperative chest radiographs (82.9%). An abnormality was seen on 44 radiographs (4.8%), with only 28 (3.1%) ultimately identifying any complication. The utilization of a postoperative chest radiograph was independent from complications (p = 0.52). There were 39 (3.5%) postoperative complications occurring in 35 patients, including 12 pneumothoraces (1.1%), 12 hemothoraces (1.1%), 7 malpositioned catheters (0.6%), and 1 mortality (0.09%) in a patient who did not have any findings of a technical complication on autopsy. Twenty-one of these patients (60%) were managed with observation and a repeat chest radiograph alone. Thirteen (37%) patients required 16 interventions: 7 tube thoracostomies, 7 re-operations and 2 thoracotomies. Of all patients, only 1 complication requiring intervention (0.09%) was identified solely by post-operative chest radiograph without either prior intraoperative recognition or postoperative symptomatology.
Conclusion:
To our knowledge, our data represent the largest analysis of technical complications following image-guided CVL placement in the pediatric population. Routine postoperative chest radiographs offer minimal value in identifying clinically significant pathology. Of abnormal postoperative chest radiographs, more than 2/3 are false positives or can be managed by observation, and the majority of the remainder represent previously identified complications. We recommend abandoning routine postoperative chest x-ray following image-guided CVL placement in favor of a clinical, symptom-driven approach to postoperative imaging.
