By Clare Gurton (

There is increasing pressure to describe the benefits of new treatments in terms of increased productivity in the work-place, and many clinical studies are now attempting to include some measurement of improved productivity as a secondary endpoint.

Whilst many trials have indeed shown that effective medical intervention can decrease illness-associated loss of work time, it does not necessarily follow that an effective treatment will always improve a standard measure of productivity.

Laura Pizzi and colleagues from Jefferson Medical College have constructed a simple model of employee behaviour based on economic theory which can predict the effect of a treatment on time worked.

Their model suggests that a medicine might only decrease loss of work time if its use is associated with increased job satisfaction but has little effect on non-work or leisure time.

In addition the model highlights the fact that work hours are influenced by several variables such as wages, the comparative cost of other goods, job characteristics and health status. These variables might work to make an effective therapy either increase or decrease productivity.

This work illustrates the need for a greater understanding of the effect of many different social and economic factors on productivity and the importance of considering these factors when designing and interpreting clinical trials.

Pizzi LT, and Gagne JJ. Pills and productivity: what economic theory tells us about employees work behaviours. ISPOR Connections, February 15, 2007, Vol.13, No.1