HomeADScreen: Developing Alzheimer's disease and related dementia risk identification model in home healthcare.

Autor: Zolnoori M; Columbia University Irving Medical Center, New York, NY, USA; Center for Home Care Policy & Research, VNS Health, New York, NY, USA; School of Nursing, Columbia University, USA. Electronic address: mz2825@cumc.columbia.edu., Barrón Y; Center for Home Care Policy & Research, VNS Health, New York, NY, USA., Song J; School of Nursing, Columbia University, USA., Noble J; Columbia University Irving Medical Center, New York, NY, USA., Burgdorf J; Center for Home Care Policy & Research, VNS Health, New York, NY, USA., Ryvicker M; Center for Home Care Policy & Research, VNS Health, New York, NY, USA., Topaz M; Center for Home Care Policy & Research, VNS Health, New York, NY, USA; School of Nursing, Columbia University, USA.
Jazyk: angličtina
Zdroj: International journal of medical informatics [Int J Med Inform] 2023 Sep; Vol. 177, pp. 105146. Date of Electronic Publication: 2023 Jul 13.
DOI: 10.1016/j.ijmedinf.2023.105146
Abstrakt: Background: More than 50 % of patients with Alzheimer's disease and related dementia (ADRD) remain undiagnosed. This is specifically the case for home healthcare (HHC) patients.
Objectives: This study aimed at developing HomeADScreen, an ADRD risk screening model built on the combination of HHC patients' structured data and information extracted from HHC clinical notes.
Methods: The study's sample included 15,973 HHC patients with no diagnosis of ADRD and 8,901 patients diagnosed with ADRD across four follow-up time windows. First, we applied two natural language processing methods, Word2Vec and topic modeling methods, to extract ADRD risk factors from clinical notes. Next, we built the risk identification model on the combination of the Outcome and Assessment Information Set (OASIS-structured data collected in the HHC setting) and clinical notes-risk factors across the four-time windows.
Results: The top-performing machine learning algorithm attained an Area under the Curve = 0.76 for a four-year risk prediction time window. After optimizing the cut-off value for screening patients with ADRD (cut-off-value = 0.31), we achieved sensitivity = 0.75 and an F1-score = 0.63. For the first-year time window, adding clinical note-derived risk factors to OASIS data improved the overall performance of the risk identification model by 60 %. We observed a similar trend of increasing the model's overall performance across other time windows. Variables associated with increased risk of ADRD were "hearing impairment" and "impaired patient ability in the use of telephone." On the other hand, being "non-Hispanic White" and the "absence of impairment with prior daily functioning" were associated with a lower risk of ADRD.
Conclusion: HomeADScreen has a strong potential to be translated into clinical practice and assist HHC clinicians in assessing patients' cognitive function and referring them for further neurological assessment.
Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
(Copyright © 2023 Elsevier B.V. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: