Skip to main content
Fig. 1 | BMC Medical Research Methodology

Fig. 1

From: A new likelihood model for analyses of pharmacoepidemiologic case–control studies which avoids decision rules for determining latent exposure status

Fig. 1

Four typical persons (horizontal lines) representing the different types of contributions to the likelihood with respect to exposure status on the index date indicated by squares (filled: treated, unfilled: untreated). Black bullets represent prescription redemptions included in the likelihood, grey bullets are redemptions not included in the likelihood. The thicker black lines represent periods of treatment, whereas dashed lines are periods without treatment. \(T\) is duration of the prescription and \(R\) is time from the last prescription redemption before the index date \({t}_{0}\) until the index date. \(X\) indicates whether a patient continues treatment, \(Z\) exposure status on the index date, and \(V\) whether a prescription is redeemed in the interval \((\delta ;{t}_{0})\). For simplicity, index dates have been aligned on the time scale. Type refers to patient type, see text for details

Back to article page