About Matching Patients for Prior Studies

A patient in IntelePACS may have more than one Medical Record Number (MRN) or patient ID. To mitigate the risk of patient priors not appearing in InteleViewer for a particular patient, IntelePACS uses a probability-based calculation to search external servers that are connected to IntelePACS for patients who are a potential match to the current patient.

Patients that are found to be potential matches to the current patient are referred to as candidates and appear in the Patient Record tab in InteleViewer.

What constitutes a candidate?

IntelePACS uses a combination of confidence-level values and a threshold value to find and show potential candidate patients.

The confidence-level values (matching algorithm) and threshold value are configured separately on your IntelePACS by Intelerad personnel. Contact your PACS administrator if you want to change the matching algorithm or threshold value.

Element	Description
Confidence-Level Values	Values that are assigned to various patient information attributes, such as last name and date of birth. The values are set by way of a candidate-matching algorithm wherein the patient-attribute values are defined. Each value in the algorithm is a percentage (0-99) that denotes the weight or importance that is given to that attribute in the event of a match.
Threshold Value	Defines the minimum confidence value (percentage) that a candidate patient must reach in order to be considered a possible match. Only candidates that meet the threshold appear in InteleViewer as a potential patient match. For example, if the threshold value that is configured on your IntelePACS is 50%, a patient appears as a candidate patient in InteleViewer if the matching patient attributes add up to 50% or higher. If the matching patient attributes add up to less than 50%, the patient is considered to be a non-match and will not appear in InteleViewer.

Element

Description

Confidence-Level Values

Values that are assigned to various patient information attributes, such as last name and date of birth. The values are set by way of a candidate-matching algorithm wherein the patient-attribute values are defined. Each value in the algorithm is a percentage (0-99) that denotes the weight or importance that is given to that attribute in the event of a match.

Threshold Value

Defines the minimum confidence value (percentage) that a candidate patient must reach in order to be considered a possible match. Only candidates that meet the threshold appear in InteleViewer as a potential patient match.

For example, if the threshold value that is configured on your IntelePACS is 50%, a patient appears as a candidate patient in InteleViewer if the matching patient attributes add up to 50% or higher. If the matching patient attributes add up to less than 50%, the patient is considered to be a non-match and will not appear in InteleViewer.

Candidate-Matching Algorithms

To accommodate regional and workflow differences, IntelePACS offers a variety of candidate-matching algorithms. All candidate-matching algorithms match against the following patient attributes, however the values (weight) given to the attributes differ for each algorithm:

First name
Last name
Date of birth
Gender

In addition to the above attributes, some candidate-matching algorithms also match against a regional Medical Record Number (MRN), such as RAMQ, ULI, and NHS.

Each algorithm has an initial-match attribute that sets the baseline for candidates. If a match is not found for the initial-match attribute, then no candidates will appear in InteleViewer. If the initial-match attribute yields results, the list of potential candidates is reduced further based on the remaining attributes.

The confidence-level values for the different attributes are predefined within a given candidate-matching algorithm. However, IntelePACS also offers a customizable candidate-matching algorithm for maximum flexibility.

Intelerad personnel can configure your IntelePACS to use one of the following candidate-matching algorithms:

Algorithm	Description
Last Name	Initial match based on patient last name.
Date of Birth	Initial match based on patient date of birth.
RAMQ or Last Name	Initial match based on patient RAMQ number or last name.
ULI or Last Name	Initial match based on patient ULI number or last name.
NHS or Last Name	Initial match based on patient NHS number or last name.
Generic (MRN or Last Name or Date of Birth)	An algorithm that allows for custom confidence-level values per attribute. The Initial match is based on a custom patient MRN (patient ID) number, last name, or date of birth. This algorithm is the only one that allows for a separate confidence-level value for a patient's middle name. All other matching algorithms use the value for the first name as the value for patient middle names.