Extending the boundaries of emergency medical simulation into high-risk, low probability events