Welsh informatics service plans patient index upgrade

Plans for new version to be operational by spring 2018 with emphasis on further reduction of duplicated data

NHS Wales Informatics Service is planning to launch a new version of its Enterprise Master Patient Index (EMPI) by the spring.

Binary data abstractNicola Turner, senior products specialist for the service, said it will begin the upgrade from version 10.1 to v11.6 in the second week of January, with the aim of having it in place by the end of March.

She was speaking at the Single View of the Citizen and Data Wise event staged by Camden Council in London this week, outlining the role of the EMPI in providing a unified view of patients treated by NHS Wales.

The move marks the next step in the development of the EMPI, which links records from several systems for an individual patient and provides a ‘gold standard’ identity record. It was originally introduced to improve patient safety at the time of a rationalisation of the NHS in Wales – the original version 9.2 went into operation in 2010 – and is aimed at reducing errors and the number of duplicate records held for each patient.

It also contributes to the Welsh Clinical Portal, which integrates patient records from several sources.

Turner said one of the key elements of the effort with the MPI will be to achieve a further reduction in the number of duplicate records.

Data quality

“You must have good data quality,” she said. “We have just over 3 million registered patients, but in the MPI there are 25 million records because of all the other systems linking in. We have nearly 1 million duplicate records, which in my eyes is a problem.”

She added that NHS Wales Informatics Service is also working on the deduplication of data for the Welsh Community Care Information System, an integrated health and social care system that is being rolled out to local authorities and health boards.

“The system will have 22 councils and seven health boards adding to the data,” she said. “There will be duplicates between sources of data, and we’re doing a big piece of work to identify where they are.”

 Image: W.Rebel under GNU Free Documentation Licence through Wikimedia