|
Purpose |
Develop a system that can automatically add appropriate codes for billing (both CPT & ICD-10) and provide real-time notifications for any missing key elements required for billing. |
|
Tag(s) |
Non-Interpretative |
|
Panel |
Business-facing subpanel |
|
Define-AI ID |
19100002 |
|
Originator |
Rich Heller, Vikram Krishnasetty |
|
Panel Chair |
Rich Heller |
| Non-Interpretive Panel Chairs |
Alexander J. Towbin, Adam Prater |
|
Panel Reviewers |
Business-facing subpanel |
|
License |
|
|
Status |
Clinical Implementation
Value Proposition
High-quality coding is necessary for the economic success of a practice or department. An advanced, computer-driven approach would improve both efficacy and efficiency by automatic coding and alerting radiologists to coding problems with the report at the time of dictation, leading to faster time to claim submission, reduced denials, and better payments.
Narrative(s)
Example 1: A radiologist dictates a CTA of the chest and abdomen. After signing the report, the computer reviews the report, adds in the appropriate ICD-10 and CPT codes, and ensures that no details are missing. In this case, the radiologist did not mention anything about 3-D imaging in the technique section. The computer immediately flags the case and asks the radiologist to resolve the problem. In this case, MIPS and 3-D reconstructions were performed, which the radiologist adds to the report technique section, satisfying the billing requirements.
Example 2: A radiologist dictates a 2 view radiographic exam of the chest. After signing the report, the computer reviews the dictated report, adds in the appropriate ICD-10 codes(s) and reviews the CPT code(s), and ensures that no details are missing. In this case, the radiologist mentioned the frontal view but omitted the presence of the lateral radiograph. The computer immediately flags the case and asks the user to resolve the problem. Since documentation that 2 views were obtained is necessary for the CPT code for the ordered exam, they must either acknowledge that a 1 view study was done (and thus alter the CPT code) or add in the presence of a 2nd, lateral, view.
Example 3: A radiologist dictates a complete ultrasound of the abdomen with the history “rule out gallstones.” After signing the report, the computer reviews the report, adds in the appropriate ICD-10 and CPT codes, and ensures that no details are missing. In this case, the history was insufficient. Also, mentioning the pancreas is required for the ordered exam, a complete abdominal ultrasound. The computer immediately flags the case and asks the radiologist to resolve the problems. They add in that the patient had acute right upper quadrant pain for two days, and the pancreas appears normal, satisfying the requirements.
Workflow Description
The system works with the voice recognition software to “read” the diagnostic imaging exam report immediately after the radiologist signs it before it leaves their computer screen. It auto-codes for ICD-10 and CPT and flags any problems for immediate reconciliation. These flags would be pushed as close to real time as possible.
Considerations for Dataset Development
|
Procedure(s) |
All modalities: plain film, fluoroscopy, nuclear medicine, CT, MRI, and ultrasound. Interventional radiology coding would be seperated |
Technical Specifications
Inputs
Study Information (from RIS)
|
Definition |
Procedures including all related codes, add on codes. Views, Modality, Body region, laterality |
|
Potential Features |
Exam description that will be analyzed for discrepancies |
Information on ICD-10 coding
|
Definition |
Information on ICD-10 coding, codes updated annually |
|
Potential Features |
ICD-10 coding is based on clinical indication, clinical history, findings & impression. |
Information on CPT coding
|
Definition |
Information on CPT coding, codes updated annually |
|
Potential Features |
The CPT code is based on the type of exam ordered & obtained from the RIS. The computer should confirm that this matches with the dictated report. |
Dictated Report
|
Definition |
Dictated report with information in history, technique, findings (including laterality) and impression sections. |
Demographic Information
|
Definition |
Demographic information from the report including patient age and sex. |
Primary Outputs
ICD-10 Coding Assignment
| RadElement ID | N/A |
|
Definition |
Based on information in the dictated report, assign an appropriate ICD-10 code. |
|
Data Type |
Categorical |
|
Value Set |
ICD-10 codes |
|
Units |
N/A |
CPT Coding Assignment
|
RadElement ID |
N/A |
|
Definition |
Based on information in the RIS & the dictated report, verify an appropriate CPT code. |
|
Data Type |
Categorical |
|
Value Set |
CPT Codes |
|
Units |
N/A |
Secondary Outputs
Reason for code assignment
|
RadElement ID |
N/A |
|
Definition |
Additional Information Required for Coding If not flagged for additional information, then would remain null. If flagged for additional information: Reason for info (ICD-10 or CPT), Type (History, missing elements, etc.) Specific information required |
|
Data Type |
Text |
|
Value Set |
N/A |
|
Units |
NA |
Future Development Ideas
Ideally, the system has a feedback loop such that data on the success or limitations of the codes it used for each report are analyzed for ongoing learning. Also, it would be useful to have information from the EMR used for history generation. Finally, while this system is optimized for diagnostic imaging, a similar program could be developed focused on IR exams.
