The Impact of Batching Advanced Imaging Tests in Emergency Departments

Using detailed electronic health record data from two major U.S. emergency departments (EDs), we use practice variation across physicians to uncover the operational impact of batch ordering imaging tests. Using causal inference approaches, we find that assigning an arriving patient to an ED physician who is a "batcher" versus a "sequencer" causally increases the patient's length of stay, time to disposition, and the number of imaging tests ordered. We find evidence that the impact of batching on length of stay is heavily mediated by ordering additional imaging tests and the probability of being admitted to the hospital post-ED service. This suggests that this ordering strategy may lead to clinical decision-making that introduces bottlenecks in patient flow. Conversely, sequencing imaging tests by ED physicians poses an "information gain" advantage compared to batching: the information obtained from a prior test allows for eliminating the need for ordering some future tests. Put together, our findings indicate that batch ordering may not be an optimal strategy for managing diagnostic imaging in emergency care and that interventions to reduce physician batch ordering may be warranted.