Pseudo cluster randomization / George F. Borm.

Format:

Video

Author/Creator:

Borm, George F., author.

Series:

Henry Stewart talks.

Henry Stewart talks

Language:

English

Subjects (All):

Clinical trials.

Physical Description:

1 videorecording (41 min., 10 sec.) : sound, color.

Place of Publication:

London : Henry Stewart Talks, 2007.

System Details:

video file

Summary:

Audio-visual presentation : Individual randomization versus cluster randomization ; Contamination as a possible problem with individual randomization ; Possible problems with cluster randomization: selection bias and slow recruitment control arm ; Pseudo cluster randomization as an alternative method: less contamination, less selection bias, better recruitment and statistically efficient ; Formula to calculate power ; Analysis methods.

Contents:

Printable Handouts

Navigable Slide Index

Introduction

Trial design

Types of randomisation (1)

Types of randomisation (2)

From individual to cluster randomisation

Why cluster randomisation?

Contamination

Individual vs. cluster randomisation

EASYcare study - elderly with geriatric problems

EASYcare study - design

EASYcare study - individual randomisation?

EASYcare study - cluster randomisation? (1)

EASYcare study - cluster randomisation? (2)

Dilemma

Solution?

Pseudo cluster randomisation (rando) (1)

Pseudo cluster randomisation (rando) (2)

Addressing EASYcare problems by psedo cluster

Pseudo cluster rando: less contamination

Pseudo cluster rando: less selection bias

Pseudo cluster rando: better recruitment

Conclusion for EASYcare study

Efficiency and power of pseudo cluster rando

Estimators treatment results

Simple, unweighted mean

Minimal contamination mean

Compromise: weighted mean

Sample size cluster randomisation

Sample size pseudo cluster randomisation

D-pseudo cluster (minimal variance approach)

Sample size EASYcare study

Cluster randomised design

Pseudo cluster design

Individual randomisation

Sample size comparison

Efficiency: contamination 20%, cluster size 10

Efficiency: contamination 20%, cluster size 30

Efficiency: contamination 40%, cluster size 10

Efficiency: contamination 40%, cluster size 30

Conclusion

Analysis pseudo cluster trial

Inappropriate methods

Mixed models

EASYcare study: variables

EASYcare study: dataset

EASYcare study: analysis (1)

EASYcare study: analysis (2)

EASYcare study: analysis (3)

EASYcare study: specify fixed part of model (1)

EASYcare study: specify fixed part of model (2)

EASYcare study: specify random part of model (1)

EASYcare study: specify random part of model (2)

EASYcare study: request model statistics (1)

EASYcare study: request model statistics (2)

Analysis program - SPSS code

Output: independent variables

Output: test of fixed effects

Output: estimates fixed effects (1)

Output: within and between variance

Analysis program: SAS code

Literature.

Notes:

Description based on publisher supplied metadata and other sources.

Retrieved April 11, 2024, from https://hstalks.com/bs/540/.