|
Purpose |
Determining how best to provide report data to patients, and tracking utilization of MyChart and access of reports by patients -- what do patients actually want to see vs. not want to see? |
|
Tag(s) |
|
|
Panel |
Business Facing |
|
Define-AI ID |
22100008 |
|
Originator |
Nikita Consul |
| Lead | Nikita Consul |
|
Panel Chair |
Jeff Chang |
|
Panel Reviewers |
Business Facing Panel |
|
License |
|
| Status | Public Comment |
Patient access to their EPIC chart via MyChart allows them to keep up with all new test results and messages from their doctors. Radiologists dictating imaging reports try to keep this in mind and may sometimes use language that is understandable by patients with a non-medical background. However, patients may often receive imaging test results that reveal life-changing diagnoses, before having the opportunity to discuss this with their doctor at their upcoming scheduled appointment. An algorithm that can recognize certain keywords within more sensitive imaging report results, and accordingly temporarily shield certain types of results until the doctor releases them to the patient (where permitted by law), may help to prevent these uncomfortable situations for patients.
A middle-aged female patient with breast cancer in remission for the past 5 years receives an alert regarding new results in her electronic chart and logs onto her electronic chart to view the results of her restaging scan prior to her appointment with the doctor the next day.
An initial imaging scan is taken (CT cancer staging scans) and sent to PACS. The radiologist accesses the imaging scan and dictates a report in their dictation software. The report is saved within the dictation software and auto-forwarded to PACS as the official report for the imaging scan and also sent to the AI engine. From the AI engine, the algorithm is triggered to temporarily shield the report from appearing in the patient’s version of the chart prior to their upcoming encounter with the ordering provider (where permitted by law), if specific types of results in the report would cause undue stress to the patient without the appropriate clinical context. Once the follow-up encounter with the ordering provider has been completed, the patient will have full access to the report.
|
Procedure |
CT |
|
View(s) |
N/A |
|
Age |
All |
Dictated Report
|
Procedure |
CT chest |
|
Views |
axial, coronal, sagittal |
|
Data Type |
Impression |
|
Modality |
CT |
|
Body Region |
chest |
|
Anatomic Focus |
Any |
Dictated Report
|
Procedure |
CT abdomen |
|
Views |
axial, coronal, sagittal |
|
Data Type |
Impression |
|
Modality |
CT |
|
Body Region |
abdomen |
|
Anatomic Focus |
Any; these exams may be organ-specific, for example CT Abdomen with/without contrast with Liver Mass Protocol (multi-phase dynamic enhanced contrast) |
Dictated Report
|
Procedure |
CT abdomen and pelvis |
|
Views |
axial, coronal, sagittal |
|
Data Type |
Impression |
|
Modality |
CT |
|
Body Region |
abdomen and pelvis |
|
Anatomic Focus |
Any |
Encounters
|
Definition |
Date of encounter with ordering provider whether encounter has been completed or not |
|
Data Type |
Numeric |
|
Value Set |
N/A |
|
Units |
Date/Time |
Imaging Report Field in Patient-Facing Chart
|
Definition |
Report show at all times or report does not show until encounter is complete |
|
Data Type |
Text |
|
Value Set |
N/A |
|
Units |
N/A |
Being able to gather feedback from patients’ ordering providers as to whether the algorithm appropriately or did not appropriately shield specific imaging reports may provide a data feed to better assess which inputs should lead to desired outputs.
Over time, it would be helpful for the algorithm to be able to tailor itself to different ordering providers and different patients, based on the way they would prefer to communicate (i.e., how ordering clinicians prefer to break news to their patients, how patients prefer to receive information, whether this passage of information should occur before the next clinic visit or not).