Digital Tablet-Based Testings for the Screening of Neurocognitive Disorders in Older Adults: a Diagnosis Study. (Preprint)
BACKGROUND The prevalence of neurocognitive disorders (NCD) is increasing as the population ages. The diagnosis includes a time-consuming medical and neuropsychological battery of tests in specialized centers. The development of effective screening tools could benefit primary care professionals by improving referral practices. OBJECTIVE To explore the feasibility of a digital tablet-based investigation comprising a wide set of questions combined with cognitive and manual dexterity tasks, study their relevance for the diagnosis of NCD, and select the best combination of variables likely to help the screening. METHODS We conducted a prospective diagnosis study conducted in 2 memory clinics. We recruited older adults who underwent a comprehensive cognitive assessment with a specialized team leading to a reference diagnostic. This consisted in 3-category classification: no NCD, mild NCD (m-NCD) and major NCD (M-NCD). At the same time, they completed a variety of testings implemented on a touchscreen tablet which included a medical questionnaire, cognitive and manual dexterity tasks. We analyzed the dataset using several approaches to determine the most important variables for identifying M-NCD, or any NCD (either m-NCD or M-NCD). RESULTS We included 160 participants whose reference diagnosis were M-NCD for 58 of them, m-NCD for 52, and no NCD for 50. The feasibility of the index test battery was excellent; all participants answered the medical questionnaire, all completed the cognitive tasks, and there were 7 protocol stops during the manual dexterity tasks due to pain and poor hand detection. We achieved a final model with 9 variables for the detection of M-NCD and 5 for any NCD. The variables relevant for identifying M-NCD are those related to loss of independence, frailty, information processing, sleep, and executive functions related to manual dexterity (inhibition and mental flexibility). Sensitivity was 81% and specificity 86%, with an AUC of 0.90 for the screening of M-NCD. The variables relevant to identifying NCD are memory, categorization, processing speed, the presence of hypertension, and success in a motor task requiring mental flexibility. Sensitivity was 81% and specificity 82% with and AUC of 0.91 for the detection of any NCD. CONCLUSIONS We identified 2 combinations of variables that can distinguish any NCD from M-NCD with satisfactory performance for a screening. CLINICALTRIAL This study has been registered in clinical trial under the number NCT06032611. The protocol has been approved by the French national ethic committee (Comité de Protection des Personnes Est III) under the number IDRCB2022-A00609-34.
Authors
- Frédéric Davesne (ORCID: https://orcid.org/0000-0001-9100-7109)
- Påvel G. Lindberg (ORCID: https://orcid.org/0000-0001-6976-6387)
- Quentin Le Boterff (ORCID: https://orcid.org/0009-0003-9959-7475)
- Fadi Badra (ORCID: https://orcid.org/0000-0002-2437-8230)
- Guillaume Loup (ORCID: https://orcid.org/0000-0003-3476-583X)
- Florian Maronnat (ORCID: https://orcid.org/0000-0003-3766-5789)
- Nathavy Um Din (ORCID: https://orcid.org/0000-0002-1480-6414)
- Sylvie Pariel (ORCID: https://orcid.org/0000-0003-4520-8545)
- Samir Otmane (ORCID: https://orcid.org/0000-0003-2221-4264)
- Joël Belmin (ORCID: https://orcid.org/0000-0003-1630-9582)
- Maxime Térémetz (ORCID: https://orcid.org/0000-0003-1131-856X)
Publication Details
- Published
- 2026-11-15
- DOI
- https://doi.org/10.2196/preprints.114204
- Primary Topic
- Dementia and Cognitive Impairment Research
- Type
- preprint