Avoca Systems Ltd, Identity Management

Problem Overview

NHS information teams have an ongoing challenge to ensure that patient records are not duplicated or fragmented in their various PAS and clinical systems. The scale of this problem is formidable as legacy systems continue to create new records, and the current processes to address this problem struggle to keep pace. Failure to address the challenge however can result in important information being entered on one occurrence of a patient’s record but missed from another, thus creating a risk that important patient information can be inadvertently scattered on multiple incidences of a patients record, and that on retrieving any one of these fragments, a clinician may falsely believe it to be the current and complete record. Therefore, managing multiple and fragmented patient records is important not only for information governance but for clinical risk.

The National Program for IT (NPfIT) places a renewed focus on the problem as the centralised care records services (CRS) being deployed as part of NPfIT will receive migrated patient data from the legacy systems of multiple, geographically adjacent health providers. Many of the legacy systems will hold information about the same patients, and when records belonging to patients that already have data on the central system are migrated to the CRS it is essential they be properly merged to avoid duplicate/fragmented records being created.

Typically, the CRS being deployed, such as Cerners’ Millennium system, are not designed to identify this type of duplicate record at upload time. Therefore the data would have to be first uploaded, and then the duplicates would need to be subsequently identified and merged. This approach has been deemed by the NHS to be an unacceptable solution as it would pose a potentially high clinical governance risk as the live CRS would hold multiple instances of a patient record during the time it takes to identify and merge the records.

Thus the challenge is now threefold, requiring a process to handle:

Resolving multiple patient records locally, prior to migration, and

Dealing with migrating patients who already have a record in the central shared system (e.g. Millennium)

Coordinating these activities with neighbouring health providers with common patients and similar go live dates

All of this, of course, needs to be achieved within the often aggressive deployment timescales.

IQAP have published guidelines for the merging of patient records; document reference NPFIT-FNT-TO-DQM-0071.06 - Merging Trust Master Patient Indexes (MPIs) into Shared Instances, and Avoca have been leading the way in providing an integrated solution to all of the challenges.

Solution – Identity Management

Avoca’s identity management tool, Data Synthesis, identifies fragmented or duplicate records within a single MPI, or across multiple MPIs. It is used by the NHS team to resolve these issues within their local data and across large scale shared domains to maintain a single patient record, for the migration, and the already live system MPI.

Patient identity management is extremely sensitive and the need for accuracy and information governance is critical. No other company has as much knowledge of the complex processes involved, or can offer the wealth of expertise Avoca has gained in this mission critical area. Our pioneering work for Fujitsu Services and Connecting for Health in the delivery of a cluster wide solution for intra domain de-duping of patient data is a first for any company operating within the NHS.

Local identity management

When an individual health organisation is planning to migrate their data to a central solution (such as Cerner’s Millennium), they must first ensure that they have properly resolved all fragmented records within their source data (this is also an NHS (IQAP) mandate). Data Synthesis has been used extensively on NPfIT data migration projects to enable the NHS to create a clean MPI ready to migrate. NPfIT projects that have used Avoca’s Data Synthesis to clean up their MPI data ready to migrate to Millennium include:

Barnet & Chase Farm NHS Trust

Milton Keynes Hospital NHS Foundation Trust

Queen Mary’s Sidcup NHS Trust

Surrey and Sussex Healthcare NHS Trust

Weston Area Health NHS Trust

Winchester and Eastleigh Healthcare NHS Trust

On several of these projects a single MPI was created by combining and deduping multiple MPIs from separate systems. Individual or multiple MPI (master patient index) are uploaded into Data Synthesis with no limit on the number of records (typically an MPI will be in excess of 1.5 million patient demographic records, but the software will support many times this number). Once loaded into the Data Synthesis database, the software executes a sequence of over 50 matching rules to identify fragmented/duplicate records. These rules can utilise key demographic fields from both source data and the National Tracing Service (NSTS). Many of the rules exploit soundex and phoneme logic. Each rule is uniquely identified and carries a score relating to the degree of match. Only the highest scoring match is recorded to avoid multiple counting and a detailed matching statistics report is produced detailing the number of records trapped by each matching rule.

The user is able to review the matching results in a variety of different ways, so that the review and resolution task can be distributed to a number of staff, with each user being given a different rule-set or category to review. The user can select to bulk link large numbers of high-ranking matches or can review individual records before confirming a link. Once the match is confirmed by the user linking the records, the software will execute the merge and produce a report and merged output file.

Intra Domain Identity management

In 2006, the success of Avoca’s identity management solution at a local level, led Fujitsu and the NHS to commission Avoca to deliver a centrally managed service for intra domain de-duping. Avoca provide the solution as a managed service on behalf of Fujitsu, with the software hosted centrally at the Fujitsu NPfIT data centre. The solution is used by organisations in the Southern Cluster to merge their migrating records with matching records already present on the central system, prior to the upload. To allow sufficient time for the NHS team to investigate potential matches the work needs to commence well in advance of the planned go live, the Avoca solution accommodates this by providing functionality to receive regular refreshes of source and Millennium data and tracks any changes.

The success of this service, known to Fujitsu as the Intra Domain De-duping Tool (IDDT), has led CFH to commission extensions in order to address the challenge of a split hospital trust migration. Avoca’s identity management solution helps resolve the patient numbering and case note tracking problem that staggering the trust migration into central Master Patient Index creates.

The latest version allows the domain deduping to be synchronised with local ‘source system’ deduping. Data Synthesis creates the appropriate insert and update files to upload to the target system as a result of the users actions. The merges are fully audited and any records that are merged can be unmerged. As part of the record merge output, Data Synthesis can generate HL7 merge messages to feed back to the originating system or systems so that where Data Synthesis merges records from the same source system, the original source data can be merged.