durham_ve: Estimate VE using method from Durham et al. 1988 (AJE)

In fluvee/wave: Waning Vaccine Effectiveness

Estimate VE using method from Durham et al. 1988 (AJE)

Description

This method is based on smoothing scaled residuals from the Cox proportional hazard regression model. It consists of four steps. First, an ordinary proportional hazard model is fitted using a partial likelihood function. Second, Schoenfeld residuals are calculated. These residuals are used to test the independence between residuals and time. Third, the residuals are scaled and added to the coefficient from the ordinary proportional hazard regression model. Fourth, after smoothing we can get the estimated hazard ratios as function of time by recovering the time-varying regression coefficient. This allows testing the hypothesis of no VE waning and the estimation of TVE at each time point.

Usage

durham_ve(x, df = 2, n_days, n_periods, n_days_period, var)

Arguments

x	survival object
df	degrees of freedom
n_days	number of days in the study
n_periods	number of periods in the study
n_days_period	number of days per period
var	name of variable to assess

Value

tibble of VE estimates for each period (the estimate for each period is the average VE over the days within each period.

fluvee/wave documentation built on Nov. 9, 2023, 12:31 p.m.