You have been introduced to relative and absolute measures of association already. These include risk, rates, ratios, and differences. However, often there is some confusion around when to use each of these over the other. To help you understand when to use each, I'll discuss the key differences between risks and rates, and ratios and differences. In this part of the lecture, I will focus on risk and rates, which are often confused. Risk is based on a proportion of persons with disease or outcome of interest, as expressed as a percentage. It is also known as cumulative incidence because it refers to the occurs of disease in a group studied over time. Therefore, it is calculated by taking the total number of new cases and dividing it by the population at risk at the beginning of the observation period. However, there are a couple of practical difficulties in calculating the cumulative incidents. First, everyone being studied has to be followed up for the complete duration of the study. But unfortunately, some may die from some other cause or be lost to follow-up, which makes the resulting calculation uncertain. For example, if someone dies, can you be sure if that they would not have got the disease if they had lived? Second, many diseases can occur more than once and we have to decide how to handle reocurrences. If you include them, the incidence proportion could exceed one. While if you accept only first diagnosis, you may underestimate the true burden of disease. An alternative is to express incidents as a rate, which is the number of new cases divided by the total person time at risk for the population. As you have already learned, person time is calculated by the sum total of time all individuals remain in the study without developing the outcome of interest. Like cumulative instance or risk, incidence rates also measure the frequency of new cases of disease in a population. But take into account the sum of the time that each participant remained under observation and at risk of developing the outcome under investigation. As a result, the numerator is the same for both cumulative incidence and incidence rate, but the denominators are very different. For cumulative incidence, the denominator is the total number of at risk subjects being followed. For incidence rate, the denominator is the total amount of time at risk for all subjects who were being followed. Therefore, you can only calculate an incidence rate if you have periodic follow-up information on each subject. Including not only if they develop the outcome, but also when they developed it. Cumulative incidence and incidence rates also differ on the range of values they can take. Risk must be a proportion. Therefore, it must be between 0 and 1 and reported as a percentage. Rates, on the other hand, are not restricted between 0 and 1. To sum up, cumulative incidence is useful when you want to describe the incidence of disease in a country, but do not have detailed information on each and every member of the population. For example, if you were looking at incidence of breast cancer in the UK in 2017. But based on the data you have, you couldn't take into account when exactly people developed breast cancer, or if they left the country. If you did have this information, however, calculating incidence rates would take these factors into account. This will provide a more accurate estimate of the rate at which breast cancer develops. Incidence rates are particularly useful when trying to measure incidence in studies with dynamic populations and in studies with fixed populations with relatively long follow-up time. Hopefully, you now have a better understanding of when to use risk and rates. In the next part of this lecture, I'll discuss the differences between ratios and differences.