correcting for measurement error based on datasets with a surrogate measure (Q), a gold sandard measure (R) and a biomarker (M) available at tow points in time with no additional covariates: