Automated assertion reuse for improved record linkage in distributed & autonomous healthcare environments with heterogeneous trust models

Abstract

An assertion acceptance value matrix ( 300 ) indicates the reliability of assertions, particularly assertions or decisions whether records match or do not match, made by other medical institutions in a federation of medical institutions with different patient record systems and some common patients. Records from different institutions with a high likelihood of matching or not matching are automatically matched or not matched. Those that are ambiguous are manually reviewed. The assertion acceptance value matrix is used to reduce or expedite the manual review.

Having thus described the preferred embodiments, the invention is now claimed to be: 
     
         1 . An apparatus for generating reusable comparisons, comprising:
 an input which receives a record supplied by an outside party, the records being patient records of a federation of medical institutions, the institutions having different medical records systems, some patients having unmatched patient records in the medical systems of a plurality of the medical institutions;   a database which stores a plurality of records;   at least one computer processor which:   compares the received record with each record retrieved from the database in order to generate a likelihood ratio based on a probability that the compared records match;   compares the likelihood ratio to both an accept threshold and a reject threshold,   assigns a record that is between the accept threshold and the reject threshold to an exception list for a manual determination whether the records on the exception list match;   receives assertions made by at least one outside party wherein the assertion comprises whether the records on the exception list were accepted or rejected as matching;   compares the records on the exception list that were at least one of accepted and rejected both by the party receiving the record and each outside party in order to calculate an assertion acceptance value for each outside party,   creating a matrix of assertion acceptance values including at least one axis representing the party receiving assertions, another axis representing the party issuing assertions, one cell for each intersection between the receiving axis and the issuing axis, wherein each cell containing at least one acceptance value; and   records the assertion acceptance values in the matrix.   
     
     
         2 . The apparatus according to  claim 1 , wherein the matrix is accessible by each medical institution in the federation of medical institutions. 
     
     
         3 . The apparatus according to  claim 1 , wherein each cell in the matrix contains separate assertion acceptance values for a positive assertion and for a negative assertion. 
     
     
         4 . The apparatus of  claim 1 , where the probability value is within the in the range from 0 to 1 inclusive, where 0 represents no acceptance and 1 means the received assertion is the ground truth. 
     
     
         5 . The apparatus according to  claim 1 , wherein the at least one computer processor is further programmed to calculate each assertion acceptance value as a weighted ratio of prior common assertions determined by the receiving party and each other party. 
     
     
         6 . The method according to  claim 1 , wherein each cell in the matrix of assertion acceptance values may contain separate assertion acceptance values for a positive assertion and one for a negative assertion. 
     
     
         7 . The method according to  claim 6 , wherein the received record and an accepted record of one of the medical institutions in the federation of plurality of medical institutions are assigned a unique, federation wide medical record number. 
     
     
         8 . The method according to  claim 1 , wherein the acceptance value matrix is at least one of:
 used in the likelihood ratio generating step to increase or decrease the probability in accordance with the acceptance values; or   used in the step of determining whether records on the exception list should be accepted or rejected.   
     
     
         9 . A method of matching patient records for a plurality of medical institutions with different medical records systems, with some patients having records in a plurality of the medical records systems, the method comprising:
 comparing a selected patient medical record in a medical records system with a plurality of patient records from at least one other medical record system with at least one computer processor;   generating likelihood ratios indicative of a probability that the selected patient medical record matches each of the compared records with the at least one computer processor;   automatically (1) matching the selected patient medical record with one of the compared records if the likelihood ratio exceeds an accept threshold, (2) not matching the selected patient medical record with compared records that meet a reject threshold, and (3) if the selected patient medical record and one or more compared records do not meet either the accept or the reject threshold, assigning the selected patient medical record to an exception list for manual review with the at least one computer processor;   receiving an indication whether the selected patient medical record was matched to one of the compared records by another of the medical institutions with an input;   performing at least one of the manual review and the generating step in accordance with an assertion acceptance value matrix indicative of a reliability of matches made by other medical institutions with the at least one computer processor; and   utilizing the acceptance value matrix to automatically determine whether accepted or reject records on the exception list should be accepted or rejected.   
     
     
         10 . The method according to  claim 9 , where the matrix of assertion acceptance values is created by:
 placing the name of at least one other medical institutions which receives assertions on one axis with the at least one computer processor;   placing the name of at least one other medical institutions which issues the assertions on a second axis with the at least one computer processor; and   creating one cell for each intersection between the receiving axis and the issuing axis, wherein each cell contains at least one assertion acceptance value with the at least one computer processor.   
     
     
         11 . The method according to  claim 10 , wherein each assertion acceptance value is calculated as a weighted ratio of prior common assertions determined by the receiving party and each other party. 
     
     
         12 . The method according to  claim 10 , wherein each cell in the matrix of assertion acceptance values may contain separate assertion acceptance values for a positive assertion and one for a negative assertion. 
     
     
         13 . The method according to  claim 9 , wherein the acceptance value matrix is utilized to adjust the accept and reject thresholds. 
     
     
         14 . The method of  claim 15 , where the probability value is within the in the range from 0 to 1 inclusive, where 0 represents no acceptance and 1 means the received assertion is the ground truth. 
     
     
         15 . The method according to  claim 9 , wherein further comprising:
 assigning a unique, federation wide medical record number to the received record and an accepted record of each medical institution in the federation of medical institutions.   
     
     
         16 . The method according to  claim 9 , further including using the acceptance value matrix in at least one of:
 the likelihood ratio generating step to increase or decrease the probability in accordance with the acceptance values; or   the step of determining whether records on the exception list should be accepted or rejected.   
     
     
         17 . A non-transitory computer readable medium storing instructions executable by at least one electronic processor to perform a method of matching patient records for a plurality of medical institutions with different medical records systems, with some patients having records in a plurality of the medical records systems, the method comprising:
 comparing a selected patient medical record in a medical records system with a plurality of patient records from at least one other medical record system with at least one computer processor;   generating likelihood ratios indicative of a probability that the selected patient medical record matches each of the compared records with the at least one computer processor;   automatically (1) matching the selected patient medical record with one of the compared records if the likelihood ratio exceeds an accept threshold, (2) not matching the selected patient medical record with compared records that meet a reject threshold, and (3) if the selected patient medical record and one or more compared records do not meet either the accept or the reject threshold, assigning the selected patient medical record to an exception list for manual review with the at least one computer processor;   receiving an indication whether the selected patient medical record was matched to one of the compared records by another of the medical institutions with an input;   performing at least one of the manual review and the generating step in accordance with an assertion acceptance value matrix indicative of a reliability of matches made by other medical institutions with the at least one computer processor; and   utilizing the acceptance value matrix to automatically determine whether accepted or reject records on the exception list should be accepted or rejected.   
     
     
         18 . The non-transitory computer readable medium according to  claim 17 , where the matrix of assertion acceptance values is created by:
 placing the name of at least one other medical institutions which receives assertions on one axis with the at least one computer processor;   placing the name of at least one other medical institutions which issues the assertions on a second axis with the at least one computer processor; and   creating one cell for each intersection between the receiving axis and the issuing axis, wherein each cell contains at least one assertion acceptance value with the at least one computer processor.   
     
     
         19 . The non-transitory computer readable medium according to  claim 18 , wherein each assertion acceptance value is calculated as a weighted ratio of prior common assertions determined by the receiving party and each other party. 
     
     
         20 . The non-transitory computer readable medium according to  claim 17  wherein the acceptance value matrix is utilized to adjust the accept and reject thresholds.