Analysis of Quality - Adjusted Life Time
by
S. P. Mukherjee
Department of Statistics, Calcutta University, India

Traditional Reliability Analysis is based on time-to-failure ( survival time ) of an equipment, for which only two states viz., 'functioning' (up) and 'failed' ( down ) are recognized. Multi-state Reliability Studies, of course, take into account more than two states. However, little has so far been attempted to associate appropriate measures of efficiency ( quality of performance ) with these various states, to note durations of stay in each of these states and to define Quality - Adjusted Life (QAL) time as a weighted total of these durations. Recently some work in this direction has been reported in the context of clinical trials, where the concept of 'Time WIthout Symptoms and Toxicity' (TWIST) has been used to compare the effects of several competing treatment protocols.

While it is obvious that QAL will be stochastically dominated by life ( defined as the total duration of stay in all the non-failed states taken together ), a comprehensive analysis of QAL is still awaited.

The present Talk gives some results related to the distribution of QAL, assuming exponential duration of stay in each of the non-failed states. The same way one can develop parallels of some results known in relation to TWIST.

Date received: October 9, 2000