Regression model of symptom frequency and delay from first identifying a problem to ultimate diagnosis and the total cost including covariates.

aThe p-value was obtained from a regression model using the square root of the total cost as dependent variable (because of skewness in this variable).
bThis is the CI for the difference in cost between early and late.
cThis is the amount by which the cost increases (dollars) per unit increase on the symptom score.
Regression model of symptom frequency and delay from first identifying a problem to ultimate diagnosis and the total cost including covariates.