Sentinel 5.17.5 User Guide

Deduplication Reports

Deduplication refers to the process of determining if a patient or case already exists in the application. Once you run the Dedup Report, the potential duplicate records display on a separate Deduplication Summary page. The deduplication process consists of two sub-processes:

The first process is Patient Deduplication. This stage relies on Master Patient Index (MPI) and focuses on patient demographic information. Depending on whether the patient has been determined to be a duplicate and is merged (automatically or manually), determines whether the case deduplication process begins
The second process is Case Deduplication. This process is optional and it occurs only for the case deduplication. It relies on the functionality available in the Sentinel application itself. It is triggered when certain case deduplication criteria are met. The second state is initiated only if the first stage is completed successfully. The deduplication criteria that affect the case deduplication are the same Reportable Condition and Look-back Intervals (which vary depending on the Reportable Condition)

The end result for each case of those processes is to determine whether the patient or case is a definite match, not a definite match, or is a probable match.

The settings for MPI Patient Deduplication are done within the MPI while the Sentinel Case Deduplication settings are done within Sentinel's General Configuration menu using the Disease Management option. See Disease Management.

When you enter and save a new case, Sentinel informs MPI about the patient record.

If a match is not found, MPI returns a new MPI ID to Sentinel
If the patient already exists in MPI, then MPI returns the existing patient's MPI ID information with the new information
If MPI does not have enough information to determine if the patient is a match, it defers the record to an administrator in the MPI Possible Merge Notifications to resolve

Duplicate patients are identified by comparing the following field values:

Patient First Name - Approximate string comparator (inexact matching algorithm) and nickname searches
Patient Last Name - Approximate string comparator (inexact matching algorithm)
Patient Date of Birth
Patient Sex

As each field is compared, a value is applied to the degree that it matches. Once a value is applied to each field, the values are weighed and the Deduplication Rank is calculated. The Deduplication Rank determines whether a record is an exact match of an existing record. Ranks above a certain threshold are automatically merged, while ranks below another threshold are considered non-matches. Ranks in between the two thresholds are sent to manual review in the MPI as possible matches.

NOTE: The thresholds are set in the Deduplication configuration file. The deduplication process in the MPI handles merging of the patient demographic information. Sentinel stores the information and a copy of the demographics data within its own database. After a merge is performed in the MPI, a notification is added to the Pending Work Queue in Sentinel. This allows an administrator to review the data stored in Sentinel and update it accordingly.

To run deduplication reports, select the ELR - CDC Reporting > Dedup Reports menu link. The Dedup Messages for a Given Time Period page opens. Enter the report parameter options and click View HTML Report. The report displays in a new browser tab or window. To reset all of the options to their default values and start over, click Reset.

The available fields and options for this report are as follows:

Field/Option	Description
Time Period - From Date / To Date	Enter the From and To dates using the format mm/dd/yyyy. These fields are required.
Advance Query
Message Sender	Enter the name of the entity sending the message.
Description	Enter the description to search for, such as Deduplication or Lab Message Received.
String to search in Message Text	Enter the text string to search for in the message.