ENTITY COHORT DISCOVERY AND ENTITY PROFILING
Disclosed are systems and techniques for providing a master health entity index and a data analysis mechanism designed for entity cohort discovery and entity profiling. The entity can be a healthcare facility that diagnoses or treats health conditions and diseases (e.g., hospital, clinic), individuals (e.g., providers, patients, caregivers), healthcare data (e.g., medical conditions, treatments, diagnostic studies, health outcomes), etc. For example, a data analysis mechanism may identify distinctive patient cohorts based on what happened to patients in a hospital and why the occurrence happened, reconstruct timelines of healthcare events from fragmented medical data, and leverage the existing electronic health data to generate comprehensive profiles of healthcare entities.
1 . A method performed by a computer comprising a processor and a memory of recommending customized healthcare treatment, comprising:
retrieving, by the computer, healthcare data related to a group of patients, a group of healthcare providers, and a group of procedures performed by the group of healthcare providers on the group of patients,
wherein the healthcare data related to a procedure includes timing information associated with the procedure;
receiving, by the computer, one or more healthcare criteria;
determining, by the computer, a set of patients who satisfy the one or more healthcare criteria from the group of patents based on the healthcare data;
identifying, by the computer, a set of procedures performed on the set of patients from the group of procedures based on the healthcare data;
constructing, by the computer, at least one sequence of multiple procedures from the set of procedures based on the timing information associated with the set of procedures;
receiving, by the computer, a query regarding one of the at least one sequence of procedures from a user device of a user over a communication network; and
sending, by the computer, a response to the query to the user device.
2 . The method of claim 1 , wherein the healthcare data includes insurance claims.
3 . The method of claim 1 ,
wherein the constructing includes clustering the set of procedures, and
wherein the multiple procedures belong to different clusters.
4 . The method of claim 1 , further comprising:
determining a second set of patients that do not satisfy the set of healthcare criteria from the group of patients based on the healthcare data; and
identifying a second set of procedures performed on the second set of patients from the group of procedures based on the healthcare data,
wherein identifying the set of procedures excludes any procedure in the second set.
5 . The method of claim 1 ,
wherein each of the multiple procedures in the one sequence has a common plurality of attributes, and
wherein the healthcare data related to the group of procedures includes at least one value for one of the plurality of attributes of one of the multiple procedures in one performance of the one procedure on one of the set of patients.
6 . The method of claim 5 , wherein the plurality of attributes includes a time period, a cost, or an amount of risk associated with a procedure.
7 . The method of claim 5 , further comprising computing an aggregate of values for one of the plurality of attributes of a specific one of the multiple procedures in the one sequence in all performances on the set of patients on which the multiple procedures in the one sequence have been performed from the healthcare data.
8 . The method of claim 7 ,
wherein the plurality of attributes includes a time period and a cost,
wherein the query requests an annual cost of the one sequence, and
wherein the response indicates a quotient of the aggregate cost for the one sequence and the aggregate time period in years.
9 . The method of claim 7 ,
wherein one of the plurality of attributes is a success status, and
wherein the aggregate for the success status is the success rate over all the performances.
10 . The method of claim 7 ,
wherein the query requests a worst portion of the one sequence for a specific one of the plurality of attributes, and
wherein the response indicates one of the multiple procedures in the one sequence that has the highest aggregate for the specific one attribute.
11 . The method of claim 7 , further comprising, when the query requests an estimate of an overall value of at least a part of the multiple procedures in the one sequence for a specific one of the plurality of attributes, the response includes a sum or product of the aggregates for the specific one attribute over the at least part of the multiple procedures.
12 . The method of claim 7 ,
wherein the query requests the best of the at least one sequence for a specific one of the plurality of attributes,
further comprising:
computing a sum or product of the aggregates for the specific one attribute over the multiple procedures in each of the at least one sequence; and
sending to the user device a description of one of the at least one sequence with the smallest aggregate.
13 . The method of 7 ,
wherein the healthcare data includes an assessment of current health condition for each of the group of patients,
further comprising
for each of the at least one sequence, computing an aggregate assessment of current health condition over the set of patients on which the multiple procedures in the one sequence have been performed,
where the query requests one of the at least one sequence leading to the best health condition,
wherein the response includes a description of one of the at least one sequence having the best aggregate assessment of current health condition.
14 . The computer-implemented method of claim 7 , further comprising:
for a specific one of the plurality of attributes of a specific one of the multiple procedures in the one sequence, identifying a performance of the specific one procedure on one of the set of patients where a value of the specific one attribute is greater than the aggregate for the one attribute by at least a certain amount; and
generating a report of the identified performances.
15 . The method of claim 5 ,
wherein the timing information includes a start time in a performance of one of the multiple procedures in the one sequence, and
wherein one of the plurality of attributes is a duration between a start time of the one procedure in the one sequence and a start time of the next procedure in the one sequence.
16 . The method of claim 1 , wherein the timing information includes a start time or an end time of the one procedure or a total amount of time taken by the one procedure.
17 . The method of claim 1 ,
wherein the query includes a specific sequence of procedures, and
wherein the response indicates a difference between the specific sequence and the one sequence.
18 . The method of claim 15 , wherein the difference includes a procedure in the one sequence but not in the specific sequence.
19 . The method of claim 1 , wherein the set of one or more healthcare criteria includes having a medical condition, symptom, or disease.
20 . The method of claim 1 , wherein the constructing includes identifying the at least one sequence of the multiple procedures as the sequence that most frequently occurs to the set of patients from the healthcare data.
21 . The method of claim 1 , wherein the one or more healthcare criteria are received from the user device.