A review of 14 studies and 17 different prediction models looked at how well these tools could identify the risk of pancreatic cancer. The study focused specifically on people who were newly diagnosed with diabetes. These models are intended to help doctors understand the timing of potential risks over time.
The analysis found that the models showed a moderate level of accuracy in distinguishing between patients at different risk levels. However, there was a lot of variation between the different studies included in the review. Additionally, many of the models were not tested enough in independent groups to be fully certain of their performance in everyday clinical settings.
Because these results come from a broad collection of varied data, they are currently used as tools for risk tracking rather than a way to prove a direct cause between diabetes and cancer. The findings are still early and have limitations regarding how well they can be used in clinics right now. Patients should discuss these findings with their doctors to understand what this means for their specific health situation.