This study is to determine the effectiveness of a computerized clinical decision support tool (Best Practice Alert - BPA) in reducing unnecessary platelet transfusions based on guidelines published by national transfusion societies such as the AABB (formerly American Association of Blood Banks).

A computerized alert will be deployed through the electronic health record. The alert will trigger when a provider orders platelet transfusion above a threshold supported by current guidelines. The alert will provide information on current evidence and give the provider the option to cancel the order or proceed. The alert will not be triggered in procedural areas such as the operating room or as part of any emergency orders to prevent any delays to urgent patient care.

The proposed project will evaluate this alert by randomizing its implementation by patient chart. The investigators will observe the prevalence of this alert and its impact on ordering practices for the next 3 months. For patient charts randomized not to receive the intervention, the providers will not receive any alerts, but the analytics tool will record if the alert would have been triggered and the pertinent clinical information. After a period of three months, the investigators will review patient and provider information through analytics tools, assessing differences in platelet usage between the patients whose charts showed the alerts vs. those (control) patient charts not showing alerts.

Inclusion Criteria:

• Recent platelet count >50k/uL
• Platelet transfusion ordered

Exclusion Criteria:

• Neurosurgery
• Procedure area (OR, catheterization lab, interventional radiology suite)
• Documented history of antiplatelet agent