Full Length ArticleRisk factors and clinical features of acute pulmonary embolism in children from the community

- Childhood pulmonary embolism typically affects female adolescents in the community.
- Oral contraceptive pill use is the most significant risk factor for acute pulmonary embolism in community-based children.
- Chest pain, dyspnoea, tachycardia and multiple thromboembolic risk factors are common but non-specific clinical features.

Background and aimsCurrent clinical decision rules for pulmonary embolism are based on adult populations and have not been validated in children. The objective was to identify and evaluate clinical features for a first lifetime episode of pulmonary embolism in children presenting to the emergency department.Materials and methodsWe present a case-control study of children (≤ 18 years) presenting to the emergency department of the Royal Children's Hospital, Melbourne between November 2007 and February 2015. Children with radiologically proven pulmonary embolism formed the case group, whilst children in whom there was a clinical suspicion of pulmonary embolism but negative diagnostic imaging formed the control group. Charts, electronic medical and imaging records of both cases and controls were reviewed and analysed.ResultsThere were a total of 50 patients in this study (11 cases and 39 controls). Current or recent (within three months) use of the oral contraceptive pill was the most significant risk factor in our study (odds ratio 14.667, 95% confidence interval 3.001-71.678, P < 0.001). Most other features failed to discriminate between cases and controls, although there was a trend towards increased heart rate in cases.ConclusionsPulmonary embolism is perhaps the most common presenting spontaneous venous thromboembolism in the community and teenage girls on the oral contraceptive pill are most at-risk amongst children. The clinical signs and symptoms are often non-specific. Additional larger studies are required to determine the significance and magnitude of potential clinical predictors identified in this study. This may lead to derivation of a paediatric-specific pre-test probability tool.