Appendix 6 Including nonhealth outcomes in a supplementary analysis

Page last updated: September 2016

Presenting nonhealth outcomes

Occasionally, listing a proposed medicine may have direct patient benefits that are not health outcomes – for example, providing a more convenient form of administration to the patient.

Supplementary methods to estimate the monetary (or other) value of the nonhealth benefit may include a conjoint analysis or a discrete choice experiment that includes a monetary attribute, an attribute reflecting a range of options for each of the nonhealth outcomes of interest, and/or other attributes.

Where there are no other substantive changes in health outcomes between the proposed medicine and its main comparator, this estimate (eg willingness to pay) can be included in a supplementary cost-benefit analysis. Where this cost-benefit analysis results in a consumer surplus, nominate a suitable basis for sharing this consumer surplus between the sponsor and the taxpayer.

Production changes

In the context of health economics analyses, a production change is a change in total output value across society of productive work in the economy. Productivity is a function of output units (eg days of work) multiplied by their value (eg an appropriate daily wage as a proxy for the value of each day of work).

Health interventions may claim to result in a change in production across society associated with patients gaining or losing working time as a result of changes in their health and consequent capacity to work. Less commonly, a health intervention may claim that workers’ efficiency will be affected, such that the value of their work output is changed on a per-unit basis (ie it can be represented by a higher or lower wage).

Changes in production as an outcome of therapy may be included in supplementary analyses in submissions to the PBAC, but do not include them in the base-case analysis. This separation allows the PBAC to consider the impact of including production changes on the direction and extent of change on the base case. Including production gains favours interventions that improve the health of people who are able, and choose, to return to contributing to societal production and, hence, there are equity implications of including productivity changes in the base case.

If presenting productivity claims associated with a proposed medicine, there are several difficulties in estimating the net present value of production changes. From a societal perspective, the productivity of an individual worker cannot be considered in isolation, but should be considered in the context of a workplace, a workforce and society. The following three underpinning assumptions should be incorporated into all productivity analyses:

• For short-term absence, production will be made up on return to work.
• Employers usually have excess capacity in the labour force to cover absenteeism.
• For long-term absence, production will be made up by a replacement worker who would otherwise be unemployed.

When presenting estimates of the marginal increase in society’s production because of the return of healthy workers:

• provide details of the method used and its assumptions
• discount appropriately any productivity changes anticipated beyond one year
• address each of the assumptions listed above when estimating production changes from the potential working time gained or lost (reported in time units).

For example, the claim that there has been a recovery of production lost because of returning to health from an episode of illness depends on demonstrating the following three factors:

• The worker returns to work and the worker is productive.
• The production lost is not made up elsewhere by others in the company or the same worker following return to work.
• No temporary replacement has been employed.

Address each of these three factors to provide robust evidence in support of estimates.

Ensure that estimates of the proportion of people who choose to return to work account for those who would choose not to return (and instead use their time gain on other activities that will have been captured by a gain in utility weights), as well as the influence of incentives provided through sickness benefits, which may operate differently across jurisdictions.

The approach above may be adapted to other contexts, such as a medicine that prevents future episodes of illness, or a medicine that might improve production capacity in individuals who, without the proposed medicine, would otherwise stay at work, although unwell, and therefore function at less than full production capacity.

When the economic approach is a cost-utility analysis, discuss how the method of estimating productivity changes avoids double-counting the estimates of health-related quality-of-life changes. The utility weights in this analysis already capture these health-related changes because they incorporate the utility impacts of productive capacity for the individual receiving the proposed medicine. These health-related changes are therefore already appropriately included in the denominator of the cost-utility ratio.

Strongly justify any production changes that are combined with surrogate outcome indicators in an economic evaluation, because this combination is generally associated with inappropriately high levels of uncertainty.