The straightforward simplicity of interventions, including a paper checklist and changes to order sets, makes them reproducible in a variety of other hospital settings. Another key facilitator was the involvement of the nursing staff. Educating nurses on local guidelines for intermittent oximetry use empowered them to proactively transition patients who met the goals and presented no clinical concerns, which removed the step of calling the resident to change the order.3
Stanford University Hospital
Dr. Shieh
Lisa Shieh, MD, PhD, reports that Stanford began by creating a steering committee of representatives from all hospital departments, which resulted in the development of evidence-based guidelines for necessary transfusions. The hospital’s EHR was programmed to fire a best practice alert (BPA) should a physician order a transfusion in a hemodynamically stable (hemoglobin reading 7 or above) patient; however, the smart BPA did not fire for patients diagnosed with bleeding disorders, hematology and oncology patients, and other special populations. Should a physician choose to transfuse despite the BPA, a reason had to be entered into the EHR.
Prior to the intervention, 50% of transfusions at Stanford were given to patients with a hemoglobin reading of 8 or above. After the intervention, only 30% fell into that category.
“Giving physicians information via the BPA in the moment they are treating the patient really made the difference,” Dr. Shieh says. “The fact that it allows them to explain why they are choosing to transfuse a particular patient allows us to understand appropriate blood usage better.”
“There are patients who really do need blood but exist outside the guidelines.”
“Adding the educational component really helped make the change in how physicians practice. When they understand why they are being asked to do things differently, they are more likely to change their behavior.” —Manya Gupta, MD
University of California, San Diego
Dr. Seymann
Remus Popa, MD, and Gregory Seymann, MD, SFHM, instituted a program at the University of California at San Diego hospital to reduce the number of patients in a telemetry bed who receive no benefit from cardiac monitoring. When physicians entered telemetry orders into a patient’s EHR, they were prompted to choose a diagnosis from the list of accepted indications programmed into the system. They were also able to choose a diagnosis of “Other” for patients who existed outside the guidelines and then explain their reasoning for ordering cardiac monitoring.
“This system is not attempting to limit the physician’s autonomy,” Dr. Popa explains. “Choosing ‘Other’ allows the clinician to order telemetry for patients with special situations, maybe not entirely addressed by the guidelines, and entering a reason can help tweak the telemetry order set going forward.”
The intervention successfully reduced the use of telemetry to a post-intervention rate of 20% from a baseline of 44%. An additional benefit was the 1.2-hour reduction in time elapsed from presentation at the ED to being placed in the telemetry bed. Because fewer of these beds were in use at any time, patients who could benefit from cardiac monitoring were in place faster.
Rush University Hospital, Chicago
Hospitalist Manya Gupta, MD, an assistant professor in the department of internal medicine, and her team at Rush embarked on a program to decrease the number of blood transfusions performed per year to be more in line with new data demonstrating that more restrictive transfusions result in better patient outcomes. The hospital had originally tried to reduce the number of transfusions by amending the blood transfusion order set in the EHR system, requiring physicians to select the indication for the transfusion from a checklist. If the desired indication was not found on the checklist, they could check “Other” and proceed with the transfusion, even if the patient did not require a transfusion. This system did not result in a meaningful decrease in transfusion.