In this cohort study, the standard outcome measure is relative risk. Cohort studies are a type of observational study that examines the outcomes of patients with different exposures or risks. This study investigated the outcomes (erectile function scores) of two groups of men who had received radiation therapy for prostate cancer, one of which was exposed to tadalafil and one of which was not.

Relative risk is the ratio of the probability of developing a specific outcome (in this case, erectile function score) in a group with an exposure or risk factor present (treatment with tadalafil) to the probability of developing the specific outcome in a group without the exposure or risk factor present (no treatment with tadalafil). A relative risk that is greater than or less than 1 indicates a more likely or less likely outcome, respectively, of the measured variable in the exposure or risk factor group compared with the group without the exposure or risk factor present.

Confidence intervals are a method for indicating the range in which a value derived from a study is likely to lie, with 95% being the usual calculated range. Whereas a P value is calculated to assess whether a trial result is likely to have occurred simply by chance, confidence intervals provide a range of possible effect size compatible with the data. Confidence intervals may be calculated for most statistical reporting measures but are not themselves used as an outcome measure for evaluating study results.

The number needed to treat (NNT) is an estimate of the number of patients who must receive an intervention to cause one patient to experience the beneficial outcome of interest. The NNT has become a standard reporting measure for randomized controlled trials. It is calculated by taking the inverse of the absolute risk change between study groups. The concept of NNT has been applied to cohort studies to suggest the magnitude of effect of an exposure (sometimes referred to as number needed to be exposed); however, because cohort studies are observational and not experimental trials, the calculation of similar measures for cohort studies is statistically more difficult than in experimental studies. Additionally, the accuracy of these measures for cohort studies is less clear. Therefore, relative risk remains the main outcome measure used for cohort studies.

The odds ratio is conceptually similar to relative risk, but rather than comparing the ratio of rates of occurrence of an event between study groups, the odds ratio compares the odds (the ratio of the probability that the event will happen to the probability that the event will not happen) between study groups. The odds ratio is required for studies using retrospective data, such as case-control studies in which the groups being compared may not have a similar risk for the condition. Case-control studies compare the outcomes of patients with a disease (cases) to those without a disease (controls). Because all of the patients in this study had prostate cancer treated with radiation therapy (a cohort) and a specific subsequent exposure or risk was evaluated, it is not a case-control study and an odds ratio would not be the standard outcome measure.