Medical record linkage in health information systems by approximate string matching and clustering
Medical record linkage in health information systems by approximate string matching and clustering
This interesting paper by 3 French researchers has just been published in the open access Biomed Central journal Medical Informatics and Decision Making.
The paper highlights important issues which arise where health records from several sources are being combined into single databases for communication & record sharing purposes. This is a major headache currently for NHS Connecting for Health & the UK's National Porgramme for IT which is trying to match records from GPs, hospitals etc where each patient may have multiple records each with different hospital numbers.
The study analysed 300,000 records & found over 10% had 2 or more records for the same patient - I suspect this ratio may be even higher in the UK - and examines tools to enable accurate matching and aggregation of these records.
Technorati Tags: health records, identity, research
This interesting paper by 3 French researchers has just been published in the open access Biomed Central journal Medical Informatics and Decision Making.
The paper highlights important issues which arise where health records from several sources are being combined into single databases for communication & record sharing purposes. This is a major headache currently for NHS Connecting for Health & the UK's National Porgramme for IT which is trying to match records from GPs, hospitals etc where each patient may have multiple records each with different hospital numbers.
The study analysed 300,000 records & found over 10% had 2 or more records for the same patient - I suspect this ratio may be even higher in the UK - and examines tools to enable accurate matching and aggregation of these records.
Technorati Tags: health records, identity, research
0 Comments:
Post a Comment
<< Home