If one is to “look at the table” from a “different angle,” then the prevalence of cancer among women exposed to estrogen is 56/330 or 17% and among nonexposed women it is 18/408 or 4.4%.

In women without cancer, the prevalence was 41.3%. In women with cancer, the prevalence of estrogen exposure was 75.7% (see above). Prevalence difference and prevalence ratio All three indicators provide a “numerical value” of an association between estrogen exposure and cancer diagnosis. Using the data from the first study, one can calculate the prevalence ratio and prevalence difference, as well as odds ratio. The risk is, therefore, based on incidence (ie, it is determined prospectively), whereas in the first study we can calculate only prevalence. The risk (or probability) quantifies the share of “events” in the total number of “attempts,” eg, the risk of scoring a point (“event”) in a series of 10 throws of the ball (“attempts”) in basketball. Obviously, we cannot assess the risk of cancer based on such a study. In the first study, we look at the “existing situation,” or the so-called cross-sectional convenience sample of subjects at a gynecology ward. However, due to different study designs, conclusions that can be derived from each study are also different. The number of subjects in each category is the same for both studies.

In both studies, the results can be summarized as shown in Table 1. At the same time, 408 similar women are not exposed to estrogen and 18 of them are diagnosed with endometrial cancer – the incidence in this group is 18/408 or 0.044 or 4.4%. Second study: 330 women are exposed to pharmacological estrogens and over the next 5 years, 56 of them are diagnosed with endometrial cancer. At the same time, among the 664 women without cancer, 274 had been exposed to estrogens, giving a prevalence of 274/664 or 0.413 or 41.3%. The prevalence of estrogen exposure was therefore 56/74 or 0.757 or 75.7%. We found that 56 out 74 women with cancer had previously been exposed to estrogens. We may ask the question: “Is the exposure (to estrogens) associated with cancer diagnosis?” For this purpose, let us imagine two studies.įirst study: At a gynecology ward, we examined 74 patients with endometrial cancer and 664 patients without cancer. Let us assume that two variables of interest are “exposure of women to pharmacological estrogens” (yes/no) and “diagnosis of endometrial cancer” (yes/no).

54 odds