It's an excellent and also a difficult question because we haven't done enough. One of the reasons we haven't done enough is that it's expensive to do what you're talking about.
There's an excellent report from the Canadian Academy of Health Sciences on measuring impacts of health research at every level, starting with the basic outcomes, the papers that are published, to the more complex impact on longevity, on morbidity, on improvements and outcomes in various diseases.
What we're finding at the CIHR is that we have to do better to track these impacts. We realize there's a limit to the money we can invest in producing metrics, so we have to choose the metrics very carefully. I think that as we are developing our new programs, we're being very careful to ensure that each specific objective of the program is matched with the very specific metric. The researcher knows what he or she will be evaluated on because they are aware of what the metric is. They are then aware that the metric is absolutely linked to the objective of the program. Therefore, if they don't meet the objective, they won't meet the metric.
It is expensive and time-consuming, and is a change in culture. It means that for every metric you need a baseline. When we're talking about health outcomes, the baselines often become a problem, and measuring outcomes across the country is a challenge. Don't forget that we're not talking about one health care system; we're talking about 13 health care systems.