USING NHS PRIMARY CARE DATA TO IDENTIFY UNDIAGNOSED DEMENTIA

E Jammeh; C Carroll; S Pearson; J Escudero; A Anastasiou; J Zajicek; E Ifeachor

doi:10.1136/jnnp-2015-312379.44

USING NHS PRIMARY CARE DATA TO IDENTIFY UNDIAGNOSED DEMENTIA

E Jammeh, C Carroll, S Pearson, J Escudero, A Anastasiou, J Zajicek, E Ifeachor

Caribbean Research and Management of Biodiversity Foundation

Research output: Contribution to journal › Conference article › peer-review

Abstract

BackgroundUp to 50% of patients with dementia may not receive a formal diagnosis, limiting access to appropriate services. It may be possible to build a picture of ‘underlying undiagnosed dementia’ from a profile of symptoms recorded in routine clinical practice.AimTo develop a machine learning tool to identify patients who may have underlying dementia but have not yet received formal diagnosis from analysis of routinely collected NHS data.MethodRoutinely collected NHS READ-encoded data were obtained from 18 consenting GP surgeries across Devon, UK, totalling 26,483 patient records of those aged >65 years. 539 Patients were identified as having dementia within the 2 year study period (June 2010 to June 2012). We determined other codes assigned to these patients that may contribute to dementia risk. The dataset was used to train a supervised classifier (Naives Bayes) to discriminate between patients with underlying dementia and healthy controls using a ten-fold cross-validation approach.ResultsThe model obtained a sensitivity of 72.31% and a specificity of 83.06% for identifying dementia.ConclusionRoutinely collected NHS data can be used to identify patients who are likely to have undiagnosed dementia. This type of methodology is promising for increasing dementia diagnosis within primary care.

Original language	English
Pages (from-to)	e4.134-e4
Number of pages	0
Journal	Journal of Neurology, Neurosurgery & Psychiatry
Volume	86
Issue number	11
Early online date	14 Oct 2015
DOIs	https://doi.org/10.1136/jnnp-2015-312379.44
Publication status	Published - Nov 2015

Access to Document

10.1136/jnnp-2015-312379.44

Cite this

@article{6e9fcb031add4ab3a8d1e743c461ab64,

title = "USING NHS PRIMARY CARE DATA TO IDENTIFY UNDIAGNOSED DEMENTIA",

abstract = "BackgroundUp to 50% of patients with dementia may not receive a formal diagnosis, limiting access to appropriate services. It may be possible to build a picture of {\textquoteleft}underlying undiagnosed dementia{\textquoteright} from a profile of symptoms recorded in routine clinical practice.AimTo develop a machine learning tool to identify patients who may have underlying dementia but have not yet received formal diagnosis from analysis of routinely collected NHS data.MethodRoutinely collected NHS READ-encoded data were obtained from 18 consenting GP surgeries across Devon, UK, totalling 26,483 patient records of those aged >65 years. 539 Patients were identified as having dementia within the 2 year study period (June 2010 to June 2012). We determined other codes assigned to these patients that may contribute to dementia risk. The dataset was used to train a supervised classifier (Naives Bayes) to discriminate between patients with underlying dementia and healthy controls using a ten-fold cross-validation approach.ResultsThe model obtained a sensitivity of 72.31% and a specificity of 83.06% for identifying dementia.ConclusionRoutinely collected NHS data can be used to identify patients who are likely to have undiagnosed dementia. This type of methodology is promising for increasing dementia diagnosis within primary care.",

author = "E Jammeh and C Carroll and S Pearson and J Escudero and A Anastasiou and J Zajicek and E Ifeachor",

year = "2015",

month = nov,

doi = "10.1136/jnnp-2015-312379.44",

language = "English",

volume = "86",

pages = "e4.134--e4",

journal = "Journal of Neurology, Neurosurgery & Psychiatry",

issn = "0022-3050",

number = "11",

}

TY - JOUR

T1 - USING NHS PRIMARY CARE DATA TO IDENTIFY UNDIAGNOSED DEMENTIA

AU - Jammeh, E

AU - Carroll, C

AU - Pearson, S

AU - Escudero, J

AU - Anastasiou, A

AU - Zajicek, J

AU - Ifeachor, E

PY - 2015/11

Y1 - 2015/11

N2 - BackgroundUp to 50% of patients with dementia may not receive a formal diagnosis, limiting access to appropriate services. It may be possible to build a picture of ‘underlying undiagnosed dementia’ from a profile of symptoms recorded in routine clinical practice.AimTo develop a machine learning tool to identify patients who may have underlying dementia but have not yet received formal diagnosis from analysis of routinely collected NHS data.MethodRoutinely collected NHS READ-encoded data were obtained from 18 consenting GP surgeries across Devon, UK, totalling 26,483 patient records of those aged >65 years. 539 Patients were identified as having dementia within the 2 year study period (June 2010 to June 2012). We determined other codes assigned to these patients that may contribute to dementia risk. The dataset was used to train a supervised classifier (Naives Bayes) to discriminate between patients with underlying dementia and healthy controls using a ten-fold cross-validation approach.ResultsThe model obtained a sensitivity of 72.31% and a specificity of 83.06% for identifying dementia.ConclusionRoutinely collected NHS data can be used to identify patients who are likely to have undiagnosed dementia. This type of methodology is promising for increasing dementia diagnosis within primary care.

AB - BackgroundUp to 50% of patients with dementia may not receive a formal diagnosis, limiting access to appropriate services. It may be possible to build a picture of ‘underlying undiagnosed dementia’ from a profile of symptoms recorded in routine clinical practice.AimTo develop a machine learning tool to identify patients who may have underlying dementia but have not yet received formal diagnosis from analysis of routinely collected NHS data.MethodRoutinely collected NHS READ-encoded data were obtained from 18 consenting GP surgeries across Devon, UK, totalling 26,483 patient records of those aged >65 years. 539 Patients were identified as having dementia within the 2 year study period (June 2010 to June 2012). We determined other codes assigned to these patients that may contribute to dementia risk. The dataset was used to train a supervised classifier (Naives Bayes) to discriminate between patients with underlying dementia and healthy controls using a ten-fold cross-validation approach.ResultsThe model obtained a sensitivity of 72.31% and a specificity of 83.06% for identifying dementia.ConclusionRoutinely collected NHS data can be used to identify patients who are likely to have undiagnosed dementia. This type of methodology is promising for increasing dementia diagnosis within primary care.

U2 - 10.1136/jnnp-2015-312379.44

DO - 10.1136/jnnp-2015-312379.44

M3 - Conference article

SN - 0022-3050

VL - 86

SP - e4.134-e4

JO - Journal of Neurology, Neurosurgery & Psychiatry

JF - Journal of Neurology, Neurosurgery & Psychiatry

IS - 11

ER -

USING NHS PRIMARY CARE DATA TO IDENTIFY UNDIAGNOSED DEMENTIA

Abstract

Access to Document

Fingerprint

Cite this