Automated medical coding
Inventors
Assignees
Publication Number
US-12272445-B1
Publication Date
2025-04-08
Expiration Date
2040-12-04
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
Systems and methods for improving reimbursement rates from payors for services or procedures provided by a medical care provider are disclosed herein. In some examples, reimbursement can be improved by improving the accuracy of medical coding of services and diagnosis codes through the use of machine learning techniques. In some examples, the system can suggest codes more likely to get reimbursed by a payor through analysis of patient data and/or insurance information. In some examples, the system can determine a likelihood of reimbursement through analysis of patient data and/or insurance information.
Core Innovation
The invention provides systems and methods that improve the accuracy of medical coding and increase reimbursement rates from payors for services or procedures provided by medical care providers. This is achieved by utilizing machine learning classifiers trained with historical reimbursement data to analyze patient data and insurance information, thereby identifying service codes and diagnosis codes most likely to be reimbursed by a payor. The system automates the process of evaluating likelihood of reimbursement for specific codes, suggesting those that have a higher probability of being accepted.
The problem addressed is that medical coding is prone to errors due to clinician mistakes, incomplete documentation, and the need for specialized knowledge of complex and varying payor rules. Manual review and cleaning of data by coders is often insufficient, leading to suboptimal reimbursements or denied claims. The invention solves this by reducing user error and supporting clinicians and coders with automated, data-driven code selection and verification tools.
The system integrates patient data, clinician inputs, sensor data, and insurance details and processes this information through multiple machine learning classifiers. One classifier suggests the optimal service code with a threshold probability of reimbursement based on historical outcomes, while another analyzes the accuracy of the diagnosis code. The system presents candidate codes and associated confidence scores to clinicians, allowing them to select the code(s) most appropriate for the patient's situation and maximize chances of reimbursement.
Claims Coverage
The patent claims cover two main inventive features pertaining to automated medical coding using machine learning classifiers for code accuracy and reimbursement optimization.
Dual machine learning classifier system for code selection and accuracy
A medical coding system includes a non-transitory memory storing multiple machine learning classifiers: - One classifier identifies a service code associated with a threshold probability of reimbursement by a payor, trained using historical reimbursement data specific to the payor. - A second classifier determines the likelihood that a diagnosis code corresponding to the service code is accurate, trained using diagnostic data associated with rejected and accepted medical service codes. The system combines patient monitoring devices, clinician input, insurance information, and electronic medical record integration. It receives patient symptom and diagnosis information, assesses diagnosis code accuracy, proposes candidate diagnosis codes, and outputs service codes with confidence scores to guide clinician treatment and billing decisions.
Classifier-based system for diagnosis code accuracy and candidate service code output
A medical coding system comprising: - At least one classifier trained using machine learning with diagnostic data from both rejected and accepted service codes to determine the likelihood that a diagnosis code corresponding to a service code is accurate. - Functionality to receive patient, symptom, and diagnosis data from clinician inputs or devices, propose updated diagnosis codes from suggested candidates, and use the updated code to determine a candidate service code. The candidate service code is output to a clinician device and is associated with a threshold reimbursement probability, serving as a basis for clinician treatment plan selection.
In summary, the claims protect systems using machine learning classifiers for automated, data-driven selection and verification of diagnosis and service codes, enhancing accuracy and reimbursement likelihood based on historical billing and medical record data.
Stated Advantages
Improves the accuracy of medical coding by analyzing patient and insurance data using machine learning classifiers.
Reduces user error and administrative inefficiencies in medical billing and coding.
Enhances reimbursement rates from payors by suggesting codes with higher probabilities of acceptance.
Allows clinicians to make more informed treatment decisions by providing coded suggestions with associated confidence scores.
Documented Applications
Use in hospitals, clinics, and healthcare facilities to automate and optimize medical coding for insurance reimbursement purposes.
Integration with electronic medical record systems to process patient data and improve clinical billing workflows.
Application in patient monitoring environments where sensor and physiological data is part of the coding process.
Providing code selection support and verification for clinicians, coders, and other healthcare billing professionals.
Interested in licensing this patent?