Statistical test results by subgroups.

Drug shortages are prominent, persistent operational challenges that hospital pharmacies have been facing for years. During a drug shortage, hospital pharmacists must solve the problem of how best to invest resources to mitigate the effect of the drug shortage on patient health care. One piece of data they use to inform their decision-making is the estimated release date (ERD) of a drug, a lead time estimate given from the pharmaceutical manufacturer specifying when the next release of a drug (that is on shortage) will occur. Working with a hospital collaborator, we collected a novel set of ERD and shipment data to analyze the accuracy of this information and the impact on decision-making at hospitals. We show via statistical analysis that ERD information tends to be an inaccurate indicator of when the hospital should expect to receive more product and is subject to change unpredictably, adding additional complexity to managing drug shortages. We discuss managerial insights that stem from this analysis and lay a foundation for future research studying decision-making with unreliable lead time information.