Plain radiographic and clinical predictors of tracheobronchial foreign body in children who proceed to rigid bronchoscopy: Single institution experience.

Abstract

Purpose: To evaluate the diagnostic performance of clinical and plain radiographic predictors of paediatric tracheobronchial foreign body (TBFB) at rigid bronchoscopy (RB). Methods and Materials: Retrospective, observational study of consecutive paediatric patients (<16 years old) at a single, Australian tertiary-level teaching hospital between November 2010-May 2016 with suspected TBFB who underwent RB. Blinded evaluation of preoperative chest radiograph (CXR) and pre - operative clinical data (desaturation, wheeze, reduced air entry) were used to test diagnostic performance of individual and combined predictors of TBFB. Result(s): 72 cases children had RB for suspected TBFB during the study period, but only 45 (63%) had CXR performed at our institution due to being transferred from other hospitals. 2/45 of these cases were excluded due to inadequate inspiration (non - diagnostic) CXR, leaving 43 subjects for analysis. Of the 43 cases, 27 (63%) were male. Median age was 25 months. 34 (79%) had TBFB on RB. 47% of TBFBs were aspirated organic material (n = 20/43). Site of aspirated FB was bronchial in 60% (n = 26), tracheal in 9% (n = 4), laryngeal in 7% (n = 3) and post cricoid in 2% (n = 1). 35% had an inspiratory and expiratory radiograph. 16% had a lateral as well as an AP view. Radiographic features assessed included: collapse/consolidation (SENS 26%, SPC 44%), hyperinflation/obstructive emphysema (SENS 24%, SPC 100%), and radio-opaque FB (SENS 24%, SPC 89%). 6/6 metallic pins were identified on CXR. 2 non-metallic TBFB seen on CXR. 1 metallic TBFB on CXR not identified on RB. The combination of abnormal CXR and/or wheeze (SENS 82%, SPC 33%), reduced air entry (TPR 79%, SPC 22%), desaturation <95% (TPR 71%, SPC 33%), and/or any clinical abnormality (includes stridor cases, SENS 91%, SPC 22%, accuracy 77%). Conclusion(s): The current study indicates inconsistent radiographic technique and poor diagnostic performance of clinical and plain radiographic features, either individually or in combination, for diagnosis of TBFB, resulting in 1 in 5 children having negative RB. The role of low dose CT in better triaging children for this invasive procedure requires further evaluation.