Baseline study factors the predict dropout in the UDS

 
   Lead Investigator:    Leslie E. Phillips
   Institution      :    NACC
   E-Mail           :    lesliep@u.washington.edu
   Proposal ID      :    50


   Publications:


1  Phillips LE, Koepsell TD and Kukull WA
   Refusal and Passive Loss-to-Follow-up in the NIA Alzheimer's Disease Program Uniform Data Set
   Abstract  Neurology,72:A479-A480,2009,

   PUBLISHED





   Proposal Description:
   Aim: To identify baseline study factors that predict subject dropout in the  Uniform Data Set (UDS)**Background and Significance**: Longitudinal studies  collect baseline as well as follow-up data on subjects over a designated length  of time.  Loss to follow-up unrelated to study disqualification may be  indicative of informative censoring and biased data collected from the subset  of those who continue on study(Hernan, Hernandez-Diaz et al. 2004).  Attempts  to analytically correct for loss-to-follow-up are best conducted by knowing  those characteristics that predict dropout.  Additionally, understanding the  predictors of study discontinuation may enable interventions which reduce the  frequency of study dropout.  We will examine study-related variables from the  UDS in order to identify predictors of UDS study dropout at follow-up.**Some  factors associated with study discontinuation are disease-specific while others  are associated with drop-out across various clinical areas.  Study subjects  lost to follow-up have been characterized as more likely to be of older  age(Korkeila, Suominen et al. 2001), male sex(Garcia, Fernandez et al. 2005  Kjoller and Thoning 2005), African-American race(Psaty, Cheadle et al. 1994),  have high comorbidity / hospital admissions and/or worse self-rated  health(Garcia, Fernandez et al. 2005 Kjoller and Thoning 2005), lower  education(Russell, Palmer et al. 2001) , and are more likely to be divorced,  unmarried or living alone(Korkeila, Suominen et al. 2001 Zunzunegui, Beland et  al. 2001) (Kjoller and Thoning 2005).  In a study of aging and dementia,  advanced age, living alone, and early study item non-response were all  associated with study discontinuation(Zunzunegui, Beland et al. 2001).   **Because persons with dementia frequently rely on caregivers to participate in  study-related and other medical appointments, a consideration of caregiver  characteristics and burden may help predict study dropout.  We hypothesize that  ca