The study abstract: https://jamanetwork.com/journals/jama/article-abstract/2661581 https://jamanetwork.com/journals/jama/article-abstract/26615... . This study

The study abstract: https://jamanetwork.com/journals/jama/article-abstract/2661581 https://jamanetwork.com/journals/jama/article-abstract/26615... . This study appears to be too new for sci-hub to have it; it is extremely unlikely that the Ars Technica writer read the study at all. The headline here says the study "finds opioids no better", but what actually happened is that the study was too small to reach statistical significance on any of its endpoints. More importantly, the study chose the top of the recommended dose range for the ibuprofen+acetaminophen group but the bottom of the recommended dose ranges for the oxycodone+acetaminophen, hydrocodone+acetaminophen, and codeine+acetaminophen groups. This strongly suggests that the study designers knew what outcome they wanted, and arranged subtle details of the study (which are hidden behind a paywall!) to make sure they got it.