Recherche de facteurs associés à la maladie d’Alzheimer par réutilisation de base de données massives

Abstract : INTRODUCTION. Severe neurocognitive disorders or dementias are defined by ICD-10 and DSM-5. They encompass a broad nosographic framework: Alzheimer's dementia, vascular dementia, Lewy body dementia, frontal-temporal lobar degeneration, etc. Each type of dementia has its own diagnostic criteria and partially identified risk factors. Identifying cognitive disorders in large databases is a complex issue, which must take into account changes in knowledge. Our first objective was to describe the evolution of dementia coding in the national database of the Medicalization of Information Systems Program (PMSI) for short stays, as diagnostic criteria evolved. Our second objective was to summarize the main known associated factors of Alzheimer's disease. Our third objective was to determine the factors associated with the onset of Alzheimer's disease in the national database of the short stay PMSI.METHODS. For the first work, we used the main diagnoses on the ScanSanté site for the short stay PMSI from 2007 to 2017. For the second work, we synthesized the literature reviews and meta-analyses using the PubMed and LiSSa search engines. For the third work, we conducted an analytical study by data mining in the national database of the short stay PMSI for patients aged 55 years or older in 2014: we selected 137 potential explanatory variables in 2008; the dependant variable was Alzheimer's disease or dementia in 2014.RESULTS. Our first work on the identification of dementias shows a decrease in inpatient stays with a main diagnosis of Alzheimer's disease or dementia, with a shift towards other organic mental disorders; stability of inpatint stays with a main diagnosis of vascular dementia but with a modification of under-diagnosis (decrease in main diagnoses of multiple heart attacks and increase in all other subtypes); a significant increase in inpatient stays with a main diagnosis of dementia or other persistent or late cognitive disorders related to alcohol consumption; a homogeneous evolution throughout the French territory. These results support a coding that respects the evolution of the literature. Our next two studies on the identification of at-risk populations identify several factors associated with Alzheimer's disease or dementia, including age, gender, diabetes mellitus, depression, undernutrition, bipolar, psychotic and anxiety disorders, low education, excess alcohol, epilepsy, falls after age 75 and intracranial hypertension. These associated factors may be risk factors, early, revealing or precipitating symptoms.CONCLUSION. Identifying cognitive disorders in large databases requires a good understanding of the evolution of dementia coding, which seems to respect the evolution of knowledge. The identification of patients with factors associated with dementia allows a more focused early identification and then proper identification of the etiological diagnosis necessary for appropriate management.
Document type :
Theses
Complete list of metadatas

Cited literature [671 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02166113
Contributor : Abes Star <>
Submitted on : Wednesday, June 26, 2019 - 2:51:11 PM
Last modification on : Wednesday, July 10, 2019 - 10:59:38 AM

File

2019LILUS001.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02166113, version 1

Collections

Citation

Michaël Rochoy. Recherche de facteurs associés à la maladie d’Alzheimer par réutilisation de base de données massives. Médecine humaine et pathologie. Université du Droit et de la Santé - Lille II, 2019. Français. ⟨NNT : 2019LIL2S001⟩. ⟨tel-02166113⟩

Share

Metrics

Record views

96

Files downloads

176