“…Thus, the more plausible interpretation of their data is that in an era where filter use was widespread, unstable patients with PE, whom physicians elected not to place a filter, were sicker. The study by Spencer et al 4 is an observational study, from which drawing conclusions regarding benefits is tenuous, though such studies are better at assessing harms. The highlighted difference in PE is likely due to ascertainment bias, since physicians were less likely to scan patients with a known filter.…”