TM Merge - The Merge Report

The Create Merge Report function produces an Excel workbook for you to analyse. The files are named with the date of the report creation and the version of the program used. The files should be retained as each one contains a snapshot of the algorithm settings used -- which you may find useful to refer back to.

Results Summary Worksheet

The first sheet you see is the Summary:

• The first set of data (Status) lists the count of patients for each merge/import status and allows you to see how many patients will be imported as new & whether there were any close matches, how many will be merged with patient records in Aquila and how many will not be imported and any errors. (Note, if there are any errors, there is an Exceptions worksheet).
• Below the status of patient records is a list of the algorithm settings used during the creation of the Merge Report. Items such as the confidence threshold, the various demographic element scores and the various modifier options like fuzzy matching etc. This information is potentially vital
• Below the algorithm summary is a note of the execution date and the program version used to create the report.

Full Patient Data Worksheet

This sheet lists every record in the TM Export data and the results of it's attempted match with Aquila data.

The columns, from the left are as follows:

• ORDER: this is the order in which the records were processed. The worksheet is sorted into Last Name order by default, but you can sort by the Order column to show the records in the order they were processed.
• CPI, NHSNUMBER....RECORD YEAR: These 11 columns are an output of the source Telemarque patient record
• EXPECTED MERGE ACTION: This is what the algorithm is expecting to do with this record - see below for a summary of the different values in this column.
• AQUILA PATIENT_ID, AQUILA TARGET PATIENT SUMMARY: this is the internal ID & a summary of the demographics of the Aquila record that was found. Items in italics do not meet the required Confidence Threshold and will not be merged with the Telemarque patient record, but are shown for convenience in case you think there should be a match. If the data should match, consider editing the TM or Aquila data first.
• SQL SUBSET COUNT: this is the number of records that was found in Aquila during the initial "broad" search. This number should be quite high and is the number of records that the algorithm has to process for each TM record.
• TOP MATCHES: this is the number of records that the Algorithm has found within the subset that have some matching characteristics (but not necessarily sufficiently matching).
• TOP SCORE: the score of the top matched records. Again this could actually be quite a low score and so not necessarily a very close match at all. The higher the score, the better the match.
• REQ'D SCORE: this is the threshold score above which records will be merged.
• ALGORITHM STEPS: this is a detailed explanation of the algorithmic process for that particular TM record. It details the matches, mismatches & characteristics of the TM record. Useful to refer to if you're confused as to why the various scores have been given and/or why there is/isn't a match.

Expected Merge Action

• Import as New: no match for the TM data found in Aquila, imports as a new record.
• Patient found but Below Confidence Threshold: a patient was found in Aquila, but it did not meet the required score to merge the two records. If this should match, consider editing the data in TM or Aquila before altering the algorithm.
• Matched Patient & Above Confidence Threshold: a patient was found in Aquila that meets the algorithms minimum score required to merge the two records. These two records will be merged into a single record. If the two records should not be merged, please alter the data in either TM or Aquila. If there are many false-positives, consider increasing the algorithms confidence threshold or the individual scores - but bare in mind that this will affect *every* record.
• Poor Quality Demographics, with Images: import as new: the TM data has poor quality demographics (that does not meet the minimum Source Patient Quality Level) and which cannot be used to attempt to match a patient. However, image data is present so the record will be imported as a new Aquila patient. Consider editing the TM data to improve data quality.
• Poor Quality Demographics, no images: assume empty, do not import: the TM data has poor quality demographics (that does not meet the minimum Source Patient Quality Level) which cannot be used to attempt to match a patient, and there are no images. This record will not be imported.

---

Back to Telemarque Aquila Merge.