Getting Started With Sun Master Patient Index

The Server Project

The server project defines an enterprise-wide master patient index that processes and stores information about the patients in a healthcare organization. The Sun Match Engine is used to standardize data and provide matching weights. The Connectivity Map defines the Sun Master Patient Index application and also defines the database connection pool using an Oracle Adapter, which is configured to connect to the database using a thin client JDBC connection, or a SQL Server Adapter (depending on which version of Sun Master Patient Index you installed). The server project includes a method OTD that contains the methods used in the client projects’ Java Collaboration and Business Process. It also includes an outbound OTD that you can use with a JMS Topic to broadcast messages processed by Sun Master Patient Index back out to external systems.

JMS Topics are not configured in the default Connectivity Maps and the process is not described in this tutorial. For more information and instructions, see Defining Connectivity Components for a Master Index (Repository) in Configuring Master Index (Repository) Connectivity and Environments.

The Object Structure

The object structure for the Sun Master Patient Index application is based on standard patient data and includes information essential to identifying a patient, such as names, date of birth, national identifiers, and so on. The Parent object, named Person, contains demographic and identifying information. Child objects contain address, phone, alias, auxiliary ID, and comment information, including parsed and phonetic versions of the name and street address fields.

Query Configuration

Four different types of queries are defined in the Candidate Select file: a basic alphanumeric query, a phonetic version of the alphanumeric query, a blocking query to use for potential duplicate processing, and a blocking query to use for phonetic searches from the EDM. The blocking queries contain the following data blocks, which you can modify.

Standardization and Matching Configuration

The server project is configured to parse, normalize, and phonetically encode person names and street address names. The fields that store the modified data are defined in the database, but do not appear on the EDM. The following sections provide a general overview of how standardization and matching are configured. You can modify the configuration to obtain different results.

Person Names

By default, each name stored in the master index is phonetically encoded, and a patient’s primary and alias names are normalized prior to phonetic encoding. Normalization is based on the United States domain. The following fields are included in the object structure to store the normalized data.

The following fields are included in the object structure to store the phonetic codes for each name.

Parsing Addresses

Addresses are parsed and standardized based on the United States domain. The default value, used when no domain is specified, is the United States. Addresses are parsed into the following components (with the actual field name in parentheses).

After being parsed, the street name is also phonetically encoded. The phonetic version is stored in the Address.StreetNamePhoneticCode field.

The Match String

The default match string consists of the following fields, which are considered standard for matching on person information.

Threshold Configuration

The following table lists the default configuration for the Threshold file of the sample server project. The query builder specified for matching is the query named “BLOCKER-SEARCH” in the Candidate Select file.

Table 1 Default Threshold Parameter Configuration

Parameter 

Default Configuration 

Update Mode 

Pessimistic 

Merged Record Update 

Disabled 

One Exact Match 

False 

Same System Match 

True 

Duplicate Threshold 

7.25 

Match Threshold 

29.0 

EUID Length 

10 

Checksum Length 

0 (checksum values are not used) 

Chunk Size 

1000 

Enterprise Data Manager Configuration

In addition to the standard EUID and System/Local ID lookups, three additional searches are defined for the EDM: an alphanumeric search, a phonetic search, and a Social Security Number lookup. The following criteria fields are defined for alphanumeric searches:

  • Last Name

  • First Name

  • Middle Name

  • Gender

  • DOB, as a range

  • Mother’s Maiden Name

  • SSN

  • Maiden Name

  • Category

  • Address Type

  • Address Line 1

  • City

  • State

  • Postal Code

The following fields are defined for phonetic searches:

  • Last Name

  • First Name

  • DOB, as a range

  • Gender

  • Mother’s Maiden Name

  • SSN

  • Address Line 1

  • Address Line 2

See Query Configuration for the possible combinations allowed for the phonetic search.

The DOB field is defined to support range searching and includes a “From” field and a “To” field on the EDM search pages, allowing you to supply the range of values on which to search.

In the default configuration, the Search page is the first to appear when you log on to the EDM, all standard reports are enabled, and the audit log is disabled. If you want to view the audit log for the sample project, be sure to enable it before generating the project. Field masking is enabled for the SSN, Maiden Name, VIP Flag fields as well as certain address and telephone fields.