I've been perpetually surprised by how many common medical and/or psychological interventions have no experimental data backing them up once I go looking (not even no good data, no data at all.) (Sometimes it's because it would just be too hard to get the experimental data, but way too often it seems to just be, well, it made sense, and we have case studies, and we weren't required by law to do any testing, so we just went with it.)

That doesn't mean they're not valid! But it does make it hard to do things like measuring their influence compared to other things.