Kan Z Gianattasio1, Adam Ciarleglio2, Melinda C Power1. 1. From the Department of Epidemiology, Milken Institute School of Public Health, George Washington University, Washington, DC. 2. Department of Biostatistics and Bioinformatics, Milken Institute School of Public Health, George Washington University, Washington, DC.
Abstract
BACKGROUND: Disparities research in dementia is limited by lack of large, diverse, and representative samples with systematic dementia ascertainment. Algorithmic diagnosis of dementia offers a cost-effective alternate approach. Prior work in the nationally representative Health and Retirement Study has demonstrated that existing algorithms are ill-suited for racial/ethnic disparities work given differences in sensitivity and specificity by race/ethnicity. METHODS: We implemented traditional and machine learning methods to identify an improved algorithm that: (1) had ≤5 percentage point difference in sensitivity and specificity across racial/ethnic groups; (2) achieved ≥80% overall accuracy across racial/ethnic groups; and (3) achieved ≥75% sensitivity and ≥90% specificity overall. Final recommendations were based on robustness, accuracy of estimated race/ethnicity-specific prevalence and prevalence ratios compared to those using in-person diagnoses, and ease of use. RESULTS: We identified six algorithms that met our prespecified criteria. Our three recommended algorithms achieved ≤3 percentage point difference in sensitivity and ≤5 percentage point difference in specificity across racial/ethnic groups, as well as 77%-83% sensitivity, 92%-94% specificity, and 90%-92% accuracy overall in analyses designed to emulate out-of-sample performance. Pairwise prevalence ratios between non-Hispanic whites, non-Hispanic blacks, and Hispanics estimated by application of these algorithms are within 1%-10% of prevalence ratios estimated based on in-person diagnoses. CONCLUSIONS: We believe these algorithms will be of immense value to dementia researchers interested in racial/ethnic disparities. Our process can be replicated to allow minimally biasing algorithmic classification of dementia for other purposes.
BACKGROUND: Disparities research in dementia is limited by lack of large, diverse, and representative samples with systematic dementia ascertainment. Algorithmic diagnosis of dementia offers a cost-effective alternate approach. Prior work in the nationally representative Health and Retirement Study has demonstrated that existing algorithms are ill-suited for racial/ethnic disparities work given differences in sensitivity and specificity by race/ethnicity. METHODS: We implemented traditional and machine learning methods to identify an improved algorithm that: (1) had ≤5 percentage point difference in sensitivity and specificity across racial/ethnic groups; (2) achieved ≥80% overall accuracy across racial/ethnic groups; and (3) achieved ≥75% sensitivity and ≥90% specificity overall. Final recommendations were based on robustness, accuracy of estimated race/ethnicity-specific prevalence and prevalence ratios compared to those using in-person diagnoses, and ease of use. RESULTS: We identified six algorithms that met our prespecified criteria. Our three recommended algorithms achieved ≤3 percentage point difference in sensitivity and ≤5 percentage point difference in specificity across racial/ethnic groups, as well as 77%-83% sensitivity, 92%-94% specificity, and 90%-92% accuracy overall in analyses designed to emulate out-of-sample performance. Pairwise prevalence ratios between non-Hispanic whites, non-Hispanic blacks, and Hispanics estimated by application of these algorithms are within 1%-10% of prevalence ratios estimated based on in-person diagnoses. CONCLUSIONS: We believe these algorithms will be of immense value to dementia researchers interested in racial/ethnic disparities. Our process can be replicated to allow minimally biasing algorithmic classification of dementia for other purposes.
Authors: Brenda L Plassman; Kenneth M Langa; Ryan J McCammon; Gwenith G Fisher; Guy G Potter; James R Burke; David C Steffens; Norman L Foster; Bruno Giordani; Frederick W Unverzagt; Kathleen A Welsh-Bohmer; Steven G Heeringa; David R Weir; Robert B Wallace Journal: Ann Neurol Date: 2011-03-18 Impact factor: 10.422
Authors: Martha L Daviglus; Carl C Bell; Wade Berrettini; Phyllis E Bowen; E Sander Connolly; Nancy Jean Cox; Jacqueline M Dunbar-Jacob; Evelyn C Granieri; Gail Hunt; Kathleen McGarry; Dinesh Patel; Arnold L Potosky; Elaine Sanders-Bush; Donald Silberberg; Maurizio Trevisan Journal: Ann Intern Med Date: 2010-06-14 Impact factor: 25.391
Authors: Brenda L Plassman; Kenneth M Langa; Gwenith G Fisher; Steven G Heeringa; David R Weir; Mary Beth Ofstedal; James R Burke; Michael D Hurd; Guy G Potter; Willard L Rodgers; David C Steffens; John J McArdle; Robert J Willis; Robert B Wallace Journal: Ann Intern Med Date: 2008-03-18 Impact factor: 25.391
Authors: Michael D Hurd; Paco Martorell; Adeline Delavande; Kathleen J Mullen; Kenneth M Langa Journal: N Engl J Med Date: 2013-04-04 Impact factor: 91.245
Authors: Ellen P McCarthy; Chiang-Hua Chang; Nicholas Tilton; Mohammed U Kabeto; Kenneth M Langa; Julie P W Bynum Journal: J Gerontol A Biol Sci Med Sci Date: 2022-06-01 Impact factor: 6.591
Authors: Elizabeth L Whitlock; L Grisell Diaz-Ramirez; Alexander K Smith; W John Boscardin; Kenneth E Covinsky; Michael S Avidan; M Maria Glymour Journal: JAMA Date: 2021-05-18 Impact factor: 56.272
Authors: Melinda C Power; Erin E Bennett; Robert W Turner; N Maritza Dowling; Adam Ciarleglio; M Maria Glymour; Kan Z Gianattasio Journal: JAMA Neurol Date: 2021-03-01 Impact factor: 18.302
Authors: Jing Li; Jorge J Llibre-Guerra; Amal Harrati; Jordan Weiss; Ivonne Z Jiménez-Velázquez; Daisy Acosta; Juan de Jesús Llibre-Rodriguez; Mao-Mei Liu; William H Dow Journal: Alzheimers Dement (N Y) Date: 2021-09-05