Sequencing medical codes methods and apparatus
Some aspect include a system for automatically processing text comprising information regarding a patient encounter to prioritize medical billing codes derived from the text. The system comprises at least one storage medium storing processor-executable instructions, and at least one processor configured to execute the processor-executable instructions to analyze the text to extract a plurality of facts from the text, assign a plurality of medical billing codes to the text based at least in part on the plurality of facts, using a model trained at least in part on feedback from a user, order the plurality of medical billing codes in a sequence beginning with a primary medical billing code corresponding to a primary diagnosis associated with the text, and present the ordered sequence of medical billing codes to the user for review.
1. A system for automatically processing text comprising information regarding a patient encounter to prioritize medical billing codes derived from the text, the system comprising:
at least one storage medium storing processor-executable instructions; and
at least one processor configured to execute the processor-executable instructions to:
generate an initial ordered sequence of a plurality of medical billing codes at least in part by:
analyzing the text to extract a plurality of facts from the text; and
assigning the plurality of medical billing codes to the plurality of facts;
generate a revised ordered sequence of the plurality of medical billing codes by using a model trained at least in part on feedback from a user, wherein generating the revised ordered sequence comprises using the model to change the initial ordered sequence of the plurality of medical billing codes, the revised ordered sequence comprising a sequence beginning with a primary medical billing code corresponding to a primary diagnosis associated with the text; and
output for presentation, in a user interface, the plurality of medical billing codes in the revised ordered sequence for review.
2. The system of claim 1 , wherein the model is configured to provide a likelihood that a corresponding medical billing code is within a predetermined portion of the revised ordered sequence.
3. The system of claim 2 , wherein the model is configured to provide a likelihood that the corresponding medical billing code is within a top half of the revised ordered sequence of the plurality of medical billing codes.
4. The system of claim 2 , wherein the model is configured to provide a likelihood that the corresponding medical billing code is first in the revised ordered sequence of the plurality of medical billing codes.
5. The system of claim 1 , wherein the model is configured to provide a likelihood that each medical billing code corresponds to the primary medical billing code corresponding to the principal diagnosis.
6. The system of claim 1 , wherein the model is configured to provide a likelihood that each medical billing code corresponds to a secondary diagnosis.
7. The system of claim 1 , wherein the model is trained using training data from which a plurality of features are extracted.
8. The system of claim 7 , wherein the plurality of features includes at least one feature indicating a section from which a corresponding medical billing code was derived.
9. The system of claim 8 , wherein the plurality of features includes at least one feature indicating a position within the section from which the corresponding medical billing code was derived.
10. The system of claim 9 , wherein the at least one feature indicating the position indicates whether the corresponding medical billing code was extracted from a beginning, middle or end of the section from which the corresponding medical billing code was derived.
11. The system of claim 7 , wherein the plurality of features includes at least one feature indicating whether a corresponding medical billing code is in a major complications list.
12. The system of claim 7 , wherein the plurality of features includes at least one feature indicating whether a corresponding medical billing code is in a comorbidity list.
13. The system of claim 7 , wherein the plurality of features includes at least one feature indicating whether a corresponding medical billing code is in a principal diagnosis list.
14. The system of claim 7 , wherein the plurality of features includes at least one feature indicating historical information about a corresponding medical billing code.
15. The system of claim 14 , wherein the historical information comprises a frequency of which the corresponding medical billing code was the primary medical billing code in the past.
16. The system of claim 7 , wherein the plurality of features includes at least one feature indicating a type of document from which a corresponding medical billing code was derived.
17. The system of claim 7 , wherein the plurality of features includes at least one feature indicating whether a corresponding medical billing code is one of a co-occurrence pair.
18. The system of claim 1 , wherein the at least one processor is configured to execute the processor-executable instructions to receive the feedback from the user, the feedback indicating that the revised ordered sequence of the plurality of medical billing codes was modified, and wherein the at least one processor is configured to execute the processor-executable instructions to further train the model based at least in part on the feedback.
19. At least one non-transitory computer readable medium storing instructions that, when executed by at least one processor, perform a method for automatically processing text comprising information regarding a patient encounter to assign a sequence to medical billing codes derived from the text, the method comprising:
generating an initial ordered sequence of a plurality of medical billing codes at least in part by:
analyzing the text to extract a plurality of facts from the text; and
assigning a plurality of medical billing codes to the plurality of facts;
generating a revised ordered sequence of the plurality of medical billing codes by using a model trained at least in part on feedback from a user, wherein generating the revised ordered sequence comprises using the model to change the initial ordered sequence of the plurality of medical billing codes, the revised ordered sequence comprising a sequence beginning with a first medical billing code corresponding to a primary diagnosis associated with the text; and
outputting for presentation, in a user interface, the plurality of medical billing codes in the revised ordered sequence for review.
20. A method for automatically processing text comprising information regarding a patient encounter to assign a sequence to medical billing codes derived from the text, the method comprising:
generating an initial ordered sequence of a plurality of medical billing codes at least in part by:
analyzing, by at least one processor, the text to extract a plurality of facts from the text; and
assigning, by the at least one processor, a plurality of medical billing codes to the plurality of facts;
generating a revised ordered sequence of the plurality of medical billing codes by using a model trained at least in part on feedback from a user, wherein generating the revised ordered sequence comprises using the model to change the initial ordered sequence of the plurality of medical billing codes, the revised ordered sequence comprising a sequence beginning with a first medical billing code corresponding to a primary diagnosis associated with the text; and
outputting for presentation, in a user interface, the plurality of medical billing codes in the revised ordered sequence for review.