Likelihood Inference in the Random Effects Logistic Regression Model with ‎Response Misclassification and Covariate Subject to Measurement Error‎

SUMMARYA survey of 788 pigs from 120 farms was conducted to determine the within-farm prevalence of pathogenicYersinia enterocoliticaand a questionnaire of management conditions was mailed to the farms afterwards. A univariate statistical analysis with carriage and shedding as outcomes was conducted with random-effects logistic regression with farm as a clustering factor. Variables with aPvalue <0·15 were included into the respective multivariate random-effects logistic regression model. The use of municipal water was discovered to be a protective factor against carriage and faecal shedding of the pathogen. Organic production and buying feed from a certain feed manufacturer were also protective against total carriage. Tonsillar carriage, a different feed manufacturer, fasting pigs before transport to the slaughterhouse, higher-level farm health classification, and snout contacts between pigs were risk factors for faecal shedding. We concluded that differences in management can explain different prevalences ofY. enterocoliticabetween farms.

Download Full-text

Comparing a multivariate response Bayesian random effects logistic regression model with a latent variable item response theory model for provider profiling on multiple binary indicators simultaneously

Statistics in Medicine ◽

10.1002/sim.8484 ◽

2020 ◽

Vol 39 (9) ◽

pp. 1390-1406

Author(s):

Peter C. Austin ◽

Douglas S. Lee ◽

George Leckie

Keyword(s):

Logistic Regression ◽

Item Response Theory ◽

Regression Model ◽

Item Response ◽

Random Effects ◽

Latent Variable ◽

Logistic Regression Model ◽

Theory Model ◽

Provider Profiling ◽

Multivariate Response

Download Full-text

A comparison of imputation strategies in cluster randomized trials with missing binary outcomes

Statistical Methods in Medical Research ◽

10.1177/0962280214530030 ◽

2016 ◽

Vol 25 (6) ◽

pp. 2650-2669 ◽

Cited By ~ 6

Author(s):

Agnès Caille ◽

Clémence Leyrat ◽

Bruno Giraudeau

Keyword(s):

Logistic Regression ◽

Regression Model ◽

Multiple Imputation ◽

Random Effects ◽

Logistic Regression Model ◽

Randomized Trials ◽

Cluster Randomized Trials ◽

Complete Case ◽

Intervention Effect ◽

Cluster Randomized

In cluster randomized trials, clusters of subjects are randomized rather than subjects themselves, and missing outcomes are a concern as in individual randomized trials. We assessed strategies for handling missing data when analysing cluster randomized trials with a binary outcome; strategies included complete case, adjusted complete case, and simple and multiple imputation approaches. We performed a simulation study to assess bias and coverage rate of the population-averaged intervention-effect estimate. Both multiple imputation with a random-effects logistic regression model or classical logistic regression provided unbiased estimates of the intervention effect. Both strategies also showed good coverage properties, even slightly better for multiple imputation with a random-effects logistic regression approach. Finally, this latter approach led to a slightly negatively biased intracluster correlation coefficient estimate but less than that with a classical logistic regression model strategy. We applied these strategies to a real trial randomizing households and comparing ivermectin and malathion to treat head lice.

Download Full-text