Furthermore, it has been recently shown that patients meeting current American-European Consensus Conference ARDS criteria may have highly variable levels of lung injury and outcomes [ 2].
In these analyses, control variables, baseline levels of outcome variables, and school clustering was taken into account.
The associations between the preoperative factors and the postoperative outcomes were analyzed using the linear regression for the continuous outcome variables (the WOMAC pain and function scores) and the logistic regression for the categorical outcome variable (level of patient satisfaction).
First, the study is not experimental nor does it control for prior levels of outcome variables; therefore, results are correlational rather than causal.
In order to assess the difference in levels of outcome variables and covariates at follow-up, we compared data in the intervention and control groups using regression models and adjusting for the level of the corresponding variable at baseline.
In order to determine the most pertinent predictors of the two outcome variables: level of awareness and level of knowledge of the depression/physical health link, univariate analyses were conducted.
Regression analyses were used to test for associations between capillarization variables and other primary and secondary variables; multivariable regression accounting for age, sex, and/or baseline levels of primary outcome variables was conducted where indicated.
Planned contrasts were used in comparing mean levels of each outcome variable between the predictor variable levels and their interactions against the reference category.
Including the random intercept in the model enables variation in individual levels of the outcome variable.
By contrast, Figures 1b and 2b show variation within specific levels of the outcome variable across different amounts of CRP.
The levels of the outcome variable were: no alcohol use in the past three months, current non-hazardous use, and current hazardous use.
