If patients in clinical trials are not sufficiently similar to patients in the real world, then this leaves a critical knowledge gap about how best to treat patients in practice. A recently analysis in JAMA Oncology by Jeremy O’Connor and colleagues looked to see if there was such a gap for some of the new cancer immunotherapies. The "Population Space" tab below maps out what they found (plus some updated data on patient populations) as "heat maps," contrasting the age distribution in the patient populations from trials versus the age distribution of patients treated in practice.

categorical.py example

What does this show?

Summarizing population data with this kind of visual presentation allows us to immediately see how the distribution in trials is skewed more toward younger patients. The breakdown of patient populations into cells in the interactive presentation (which the user can then mouse-over to see more information) also suggests how there is a lot more data that could be added for each age group, such as safety and efficacy information.