>That's it. The patient's and family member's word. No testing of samples for asbestos contamination, no review of their home to see if there was asbestos. Just a sworn statement.
I just looked at the two studies referenced in the article - Moline 2019 and Emory, Kradin, and Maddox 2020.
One might assume that they looked at large samples of people with and without talc exposure and with and without a mesothelioma diagnosis to find a correlation. That assumption would be incorrect. Sample sizes were small, n=33 and n=75, respectively. In both studies, all subjects had a mesothelioma diagnosis. As you said, they relied on patient and family member statements that they had no asbestos exposure other than from cosmetic talc (something we now know to be false). They asked subjects when they started using talc, how long they used it, and to what degree they used it, and then calculated average latency to onset from that. They also looked at tissue biopsies in a subset of the subjects (e.g., 11 of the 75 in the later study) and confirmed the presence of asbestos fibers and made an attempt at calculating the fiber concentration in 9 of the 11. It is unclear exactly what these studies purport to demonstrate. There's really no conclusion to draw other than "these people all have mesothelioma, some of them have asbestos fibers in their lungs, they all claim to never have had any exposure to asbestos, and oh by the way they all also used cosmetic talc to varying degrees over the years".