Design & Statistics

Guidance for Proxy Data and Dyadic Analysis

Posted on by Erin Luers

May 17, 2024

Authors: Katie Newkirk, PhD and Joan Monin, PhD

Description: This document provides guidance on what dyadic data is, how it is different from proxy data, and what information can be gained by collecting and analyzing dyadic data compared to proxy data when conducting embedded pragmatic clinical trials (ePCTs) that include people living with dementia and care partners. Topic covered:

  • Dyadic Data
  • Proxy Data
  • Constructs Often Measured for Both People Living with Dementia and Care Partners
  • Analyzing Dyadic Data
  • When Data Don’t Require Dyadic Analysis
  • Proxy Data
  • Do I Have Dyadic Data?
Citation: Newkirk K, Monin J. Guidance for Proxy Data and Dyadic Analysis. NIA IMPACT Collaboratory; 2024. doi: 10.58234/236781742
Click to view PDF of Guidance for Proxy Data and Dyadic Analysis

IMPACT Collaboratory announces two new resources to aid in the design of ePCTs

Posted on by Jeannine Sato

Two IMPACT Collaboratory cores have developed new tools to assist in the design of embedded pragmatic trials (ePCTs) for people living with dementia.

The Design and Statistics Core has developed statistical tools and novel methodology to aid in the design and analyses of ePCTs for people living with dementia. These methods, manuscripts, statistical programs, and interactive web applications are now available to help researchers calculate sample sizes, intra-cluster correlations, and power for stepped wedge and cluster randomized trials.

The content can be accessed in IMPACT’s new Statistical Tools web page. The tools will be updated as new statistical resources become available.

The Technical Data Core has generated prevalence estimates of Alzheimer’s disease and related dementias (ADRD) for the Medicare population by geographic regions (e.g., state, hospital referral regions) and settings of care (hospitals, emergency departments, skilled nursing facilities). These data include the total number of Medicare beneficiaries, total number of beneficiaries with ADRD, and key demographic characteristics (age, sex, race, dual eligibility). Data from 2020 and 2021 are available for Medicare Advantage and Traditional Medicare populations. The prevalence data are available with consultation to help investigators planning ePCTs for people living with dementia in these settings of care.

Learn more about United States Dementia Prevalence Estimates among Medicare Beneficiaries. Interested investigators may request a consultation with an expert.

 

 

Goldfeld develops a simple way to simulate setting sample size for variable cluster sizes in randomized trials

Posted on by Sonya Sutton

Keith Goldfeld, DrPH, MS, MPA, member of the executive committee of the IMPACT Design & Statistics Core, recently published a blog post inspired by discussions with collborators from the IMPACT Collaboratory.

Goldfeld discussed how the question of variable cluster sizes has come up a number of times in recent discussion with IMPACT Collaborators about setting the sample sizes for proposed cluster randomized trials, Goldfeld explains that when working with  a fixed overall sample size, it is generally better (in terms of statistical power) if the sample is equally distributed across the different clusters. Highly variable cluster sizes increase the standard errors of effect size estimates and reduce the ability to determine if an intervention or treatment is effective.

Goldfeld realized that there is no easy way to generate the desired variable cluster sizes while holding the total sample size constant using simstudy, his preferred simulation package. In response to this, he developed a simple solution that is available for download on the blog.

Read the full post here.