Using surrogate markers in primary electronic patient record systems to confirm or refute the diagnosis of diabetes

Ashkan Bagheri, Ahmed Sadek, Tom Chan, Kamlesh Khunti, Simon de Lusignan

Abstract


Background UK primary care records are computerised and these records are used for both research and quality improvement. However, there is disparity in the prevalence of diabetes found in epidemiological studies compared with that reported through the UK's national quality improvement scheme.
Objective To investigate how non-diagnostic computer data could be used to identify, confirm or refute prevalent cases of people with diabetes.
Method We carried out a literature review to identify the most accurate non-diagnostic markers. For each type of diabetes we focused on four broad areas; demographic details, biochemical markers, clinical features and therapeutic strategies. Sample markers were tested by calculating their positive predictive value (PPV) and sensitivity (Sn) and their ability to differentiate between types of diabetes.
Results Biochemical markers were useful in identifying cases of diabetes but not in differentiating between types of diabetes as the same plasma glucose criterion is used to diagnose Type 1, Type 2, and 'other' types of diabetes; the lack of a 'fasting' qualifier blunts the use of this marker.
Auto-immune markerswere the most accurate in identifying Type 1 diabetes but are not recorded frequently in primary care.
Clinical features of diabetes and therapeutic strategies were of some use - however, without time sequence data are difficult to interpret.
Raised plasma glucose (PG), and glycated haemoglobin (HbA1c), had useful PPV but low Sn. When PG was more than 7.0 and less than 11.1 mmol/l, PPV equalled 77.8% and Sn 48%; and when PG was 11.1, PPV equalled 92% and Sn 17%. For an HbA1c of more than 6.5%, PPV was 89% and Sn 73.3%, and for anHbA1c of more than 8, PPV was 92% and Sn 26%.
A person with a combination of aged under 30 years and prescribed insulin has an 84% PPV of Type 1 diabetes; if they also have a BMI <30 kg/m2 the PPV increases to 88%. A person age over 45 years and with a BMI >30 kg/m2 has a 5.3% PPV of Type 2 diabetes; if they are also hypertensive the PPV is 30%; Asian ethnicity increases PPV to 44%.
Conclusion Non-diagnostic data has the potential to confirm or refute the diagnosis of diabetes and identify its type.

Keywords


algorithms; computer systems; diabetes mellitus; family practice; medical informatics; medical record systems; computerised; vocabulary; controlled

Full Text:

PDF


DOI: http://dx.doi.org/10.14236/jhi.v17i2.724

Refbacks

  • There are currently no refbacks.


This is an open access journal, which means that all content is freely available without charge to the user or their institution. Users are allowed to read, download, copy, distribute, print, search, or link to the full texts of the articles in this journal starting from Volume 21 without asking prior permission from the publisher or the author. This is in accordance with the BOAI definition of open accessFor permission regarding papers published in previous volumes, please contact us.

Privacy statement: The names and email addresses entered in this journal site will be used exclusively for the stated purposes of this journal and will not be made available for any other purpose or to any other party.

Online ISSN 2058-4563 - Print ISSN 2058-4555. Published by BCS, The Chartered Institute for IT