Grouping data


Data groups streamline the match process by enabling you to group similar incoming data items for comparison. You create these groups independently for HCP and HCO entities.

Data groups determine how data in the Network instance and from the incoming file is grouped for matching. This part of the process reduces the amount of data that needs to be compared, making the overall match process much more efficient. The match process applies to each like block, based on the data groups you define.

If the data groups are not defined effectively, matches might be missed as a result of records not being compared to one another that otherwise should have been compared. For example, blocking on first name and last name alone would result in missed matches because similar records would not exist in the same data group. The records John Smith and Jon Smith would end up in separate blocks and would not be compared. You should use multiple data group definitions ensure an effective match process.

Note: Data group definitions are critical to the match process; they must be defined effectively.

Grouping HCO data

The following data group definition is effective for blocking on HCOs:

  • Thoroughfare + Locality
  • Corporate Name + Thoroughfare + Locality
  • Corporate Name

Grouping HCP data

The following data group definition is effective for blocking on HCPs:

  • Thoroughfare + Locality
  • First Name + Last Name
  • NPI Number
  • ME Number

You can only group on data that exists in the incoming file. Errors occur if a record does not have values in all fields included in the data group definition when a source subscription is run. For example, if a data group definition is Thoroughfare + Locality and a record does not have any values in both fields, the record will be considered unmatched and will follow the configured merge behavior for unmatched records (by default, ADD).

Sample data groups

The following example shows how a selection of HCP records would be grouped if the data group definitions included both of the following definitions:

  • Definition 1: Thoroughfare + Locality
  • Definition 2: First Name + Last Name

The following example shows how a selection of HCP records would be grouped if the data group definitions included both the Thoroughfare + Locality data group definition and the First Name + Last Name data group definition.

Each rectangle identifies the contents of the data groups created by the first definition, while each colored circle identifies the contents of the data groups created using the second data group definition.

When grouping occurs, only the fields referenced in each data group definition are analyzed to create unique groups. Note how the fourth record, Sidney Burke, would not be added to the Sid Burke data group because the first names are not identical. A data group definition that included first_name__v, last_name__v, and locality__v would capture the Sid Burke and Sidney Burke records in the same group.

When you run a source subscription, you can choose to export the data group analysis to see the HCO and HCP records that were grouped.