Tracheobronchial foreign body aspiration in children: how reliable are clinical and radiological signs in the diagnosis?

Abstract

To evaluate the yield of clinical and radiological features in the diagnosis of suspected foreign body aspiration in children and to assess factors associated with delayed diagnosis of foreign body aspiration. Retrospective review of 10 years of experience in tertiary referral centre. Data were extracted from clinical records of children who underwent rigid bronchoscopy for suspected foreign body aspiration at the University of Nigeria Teaching Hospital Enugu from 2000 to 2009. Clinical features and radiological findings were validated against bronchoscopic findings. Data of 103 children, (mean=2.7 years, range=months to 14 years; 64% boys and 36% girls, were analysed. Majority (73%) were under 3 years of age. Foreign body aspiration was proven bronchoscopically in 85 (83%) patients. The most common symptoms were sudden choking crisis (74%) and paroxysms of cough (73%). Independent predictors of proven foreign body aspiration were witnessed aspiration, choking crisis and unilateral decreased breath sounds in univariate (P=0.001, <0.001, and 0.001 respectively) and multivariable analyses (P=0.02, 0.001, and <0.001 respectively). The most sensitive and specific clinical features were choking (86%) and witnessed aspiration episode (89%), respectively. Available chest radiographs revealed radio-opaque objects in 27% of patients. Delayed diagnosis of foreign body aspiration (>72 h) was significantly more in younger children (t=3.29; P=0.001), as well as in children with no history of witnessed aspiration, negative chest examination and radiological signs (P<0.001, P=0.02 and P=0.04 respectively). To prevent the delayed diagnosis, witnessed aspiration, choking crisis, unilateral decreased breath sounds and radiopaque objects should be checked in all suspected cases. When history is doubtful, regardless of radiological findings, bronchoscopy can be strongly recommended in the presence of two factors.