After institutional review board approval was obtained, a retrospective chart review was conducted of all general surgical inpatients with a chest tube in place after a thoracic procedure using International Classification of Diseases, 9th Edition (Centers for Disease Control and Prevention, National Center for Health Statistics, 1979) codes between January 1, 2010 and December 31, 2012 (Box).
Patients who had cardiac surgery were excluded. Cardiac surgical patients were excluded because their chest tubes are most often mediastinal and it is not possible to tell from the medical record if the mediastinum communicated with the pleural spaces or not. Cardiac surgical patients were also excluded because they differ significantly from general surgical patients in postoperative complications. The electronic medical record was reviewed to acquire demographic data; location, size, date, and time of chest tube insertion and removal; vital signs including respiratory rate and pulse oximetry before chest tube removal and at 4 hours and 12 hours after removal; and chest radiographs before and after chest tube removal. We used the official radiology report in lieu of reviewing films.
The most recent chest film done before chest tube removal was compared with the postremoval film to identify any changes. In the circumstance that a patient had more than one thoracic procedure during this timeframe, we used data from the first chest tube placement. For patients with more than one chest tube in place simultaneously, the last chest tube removal was used for data analysis. Our standard was a chest radiograph immediately after chest tube removal, and most patients in this cohort had their film taken within 2 hours after chest tube removal. A clinically significant change in physiology or film findings was considered to be one that was either documented or led to a change in management. A clinically significant change was a change in vital signs or oxygen saturation that was recognized and noted by a clinician. In addition, new findings noted on the radiology report were also considered clinically significant if they would typically be associated with a change in clinical status. The data were then entered into a Red Cap database for analysis (Vanderbilt University, Nashville, TN).
J Pediatr Health Care. 2017;31(5):588-593. © 2017 Mosby, Inc.