Because of the staged entry of participants, we evaluated two time variables: calendar date (proxy for unmeasured confounders) and day of study for a specific subject (possibly related to fatigue).
BoneXpert automatically generates the following outcome variables: (calendar) age, bone age based on Greulich and Pyle, Z-scores of bone age (compared with a healthy reference population) [ 10], BHI and Z-scores of BHI [ 5].
Adjustment for heteroscedasticity of the variance function of both u and v (as provided for by the Stata™ module "frontier") was undertaken in model development with a combination of appropriate patient (gender, patient surgical status), treatment (ventilation status) and provider descriptor variables (calendar year, ICU level, annual patient admission number and geographical locality).
To control for changes in VA enrollment policy in 2003, we included indicator variables for calendar year.
Indicator variables for calendar month of blood draw, measurement occasion, treatment assignment, and measurement occasion by treatment assignment interactions were also included.
Time-based variables (e.g., time to response) were analyzed by creating a new column with the difference between two date variables (in calendar days) and then obtaining the mean for each year and for all six years.
Finally, seasonal variation in immunization scheduling could affect the timeliness of vaccination and was entered into the time spent underimmunized regression models using indicator variables for calendar month of birth.
First, the average spatial surface for each pollutant, 1985–2000, was generated in a GAM containing a bivariate thin-plate spline of the projected x- and y-coordinates of the monitoring locations and indicator variables for calendar year to adjust for temporal trends (Wood 2006).
Each potential covariate (or groups of covariates for distance to road, land use, and power plant distance/emissions) was first considered separately in models that included the bivariate spline for the 1985–2000 spatial surface and the indicator variables for calendar year.
†R = 0.31: calculated based on the multiple linear regression model using the 7 socioeconomic variables, 3 climate variables, and calendar months.
A variable for calendar year from 2000 to 2004 was included to assess time variability.
