Does testing for prostate-specific antigen contribute to declining prostate cancer mortality?

Key variables associated with prostate cancer mortality were examined using Canadian province level time-series data over the period 1979–1999 in a multiple regression framework. The key variables driving the prostate cancer mortality rate are per capita income, per capita number of family physicians, rate of prostate cancer incidence, real per capita total health spending, and a time trend. As well, provincial dummy variables show that regional differences exist with higher rates of mortality in eastern Canada. Econometric results show a positive and significant relationship between incidence and mortality, suggesting that the increased amount of prostate-specific antigen (PSA) testing may not have been responsible for mortality declines. However, there is also a downward trend in mortality from prostate cancer when all other factors are controlled for that can be attributed partly to the onset of PSA testing given that the benefits should emerge over time. A 1% increase in incidence is associated with a 0.2% increase in mortality. A 1% increase in real per capita income is associated with a 0.5% increase in mortality when real per capita income is below Canadian $20,054. A 1% increase in the per capita number of family physicians reduces the mortality rate by 0.5%. A 1% increase in real per capita total health expenditures is associated with a 0.7% decline in the mortality rate. Finally, by 1999 there were 4.74 fewer deaths per 100,000 population due to the effect of time after controlling for all other factors—a decrease of approx. 15%.