Chest Infection, Pulmonary Edema - Acute
Conditions
Keywords
AI, Emergency Department, Radiology, Imaging
Brief summary
Does a radiology clinical dashboard increase radiologist use of specific reporting language for chest X-rays (CXRs) and thereby decrease rates of unnecessary Emergency Department (ED) prescriptions and follow-up imaging?
Detailed description
Chest X-rays (CXR) are the most common radiologic exam, comprising nearly one third of total KPNC imaging volume. Despite high CXR utilization, diagnostic yield is frequently low, partly due to the paucity of clinical information provided to radiologists at the time of CXR interpretation, resulting in frequent nonspecific CXR assessments which include terms like opacities instead of diagnostic terms such as pulmonary edema or pneumonia. This may contribute to over-treatment and unnecessary follow-up CT imaging. This study proposes the production deployment, active promotion, and evaluation of an innovative clinical dashboard called Grasshopper for radiologists to provide pertinent clinical information at the time of CXR interpretation. The dashboard leverages the electronic medical record to programmatically retrieve, and display select clinical notes, patient lab values, and vital signs. The Grasshopper study aims to prospectively evaluate the impact of Grasshopper on CXR report specificity in ED settings across 15 KPNC service areas, and on downstream care process measures such as prescription and imaging utilization.
Interventions
OTHERGrasshopper dashboard
Computer dashboard application for radiologists that displays relevant clinical information directly from the medical record
Sponsors
Kaiser Permanente
Study design
Allocation
NA
Intervention model
SEQUENTIAL
Primary purpose
OTHER
Masking
NONE
Intervention model description
Stepped wedge cluster randomized trial
Eligibility
Sex/Gender
ALL
Healthy volunteers
Yes
Inclusion criteria
Kaiser Permanente Northern California CXR is ordered during emergency room visit or inpatient hospitalization
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Number of Diagnostic CXR Reports | Through study completion, an average of 18 months. | Number of CXR reports that employ diagnostic terms in the report impression (e.g. negative, pneumonia, pulmonary edema), compared with baseline standard of care report impressions. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Number of Prescriptions after Stat CXR | Through study completion, at an average of 18 months. | Number of antibiotic and diuretic prescriptions the patient receives after having undergone a stat CXR, compared with baseline standard of care. |
| Number of CT Imaging exams after Stat CXR | Through study completion, at an average of 18 months. | Number of follow-up CT imaging exams that a patient receives after having undergone a stat CXR, compared with baseline standard of care. |
Countries
United States
Outcome results
None listed