Multi-state Markov models in cancer screening evaluation: a brief review and case study

Uhry et al have a new paper in Statistical Methods in Medical Research. This looks at the use of multi-state Markov models for evaluating cancer screening regimes. Time homogeneous three-state and five-state models to describe progression through asymptomatic (but detectable by screening) stages of cancer. Difficulties with this type of modelling are that either screening sensitivity has to be considered to be 100% or else has to be accounted for in the likelihood. Moreover, the progression rates from asymptomatic to symptomatic disease can only be observed indirectly (i.e. via interval cancers) meaning estimates have substantial uncertainty. Estimates of mean sojourn times will be quite highly dependent on the time homogeneous Markov assumption made. The authors note that there is evidence the 5-state model is unrealistic as sojourn times in the preclinical states (node negative and node positive) are likely to be correlated, but generalising to non-homogeneous and/or non-Markov models would be difficult given the lack of data.

Nonparametric estimator for the survival function of quality adjusted lifetime (QAL) in a three-state illness–death model

Biswabrata and Dewanji have a new paper in the Journal of the Korean Statistical Society. This considers estimation of quality adjusted lifetime (QAL) in illness-death models under right censoring. In contrast to their previous work, estimation is non-parametric rather than parametric. A semi-Markov assumption, where sojourn times in each state are assumed independent, is taken.