4 ms·
Yes. You would have to have a placebo controlled test to confirm placebo vs "just happens to last 8 hours". This is what placebo controls are for.
by daveguy 5y ago
Yes. You would have to have a placebo controlled test to confirm placebo vs "just happens to last 8 hours". This is what placebo controls are for.
- giantg2 5y agoI generally agree. The secondary (and more interesting) question is whether placebo effect is applicable to infants. Based on the theory of how placebo works, it seems it might not be. It seems there are no studies on that. At least I couldn't find any.
- tsimionescu 5y agoPlacebo mostly just means "good effect unrelated to intended treatment (active substance, surgical procedure etc)". That can include psychological effects of treatment (such as in pain management, depression, heart rate and blood pressure, and a handful of others). More commonly though, the placebo effect is just natural progression of the disease, errors in data collection (such as patients trying to be helpful and confirming that vague symptoms have somewhat improved), or errors in data analysis (such as lab techs thinking they've seen something in the sample that they expected to see there). The psychological effects may or may not apply to infants - that's a somewhat separate question. The other effects definitely do - it's perfectly plausible that a lab tech looking for an improvement in an infant's blood sample after treatment "finds" it.
- giantg2 5y agoI would consider those other effects to be bias, controlled via some other mechanism like a study being double blind. For example, lab techs shouldn't even know whose sample they're testing, let alone that patient's treatment, whether they're involved in a study, or the expected outcome. So I do agree these other effects exist, but I don't think they exist under the classification of placebo effect. It's possible one of these is at play in this study.