Just to be clear - the cohort in his paper is not all opioid users. It is the subset most likely to become addicts, for the following reasons: 1. The meta-anal

Just to be clear - the cohort in his paper is not all opioid users. It is the subset most likely to become addicts, for the following reasons: 1. The meta-analysis occurred during the period of time that pill mills and other unscrupulous prescribing was rampant. See that there was no note regarding patients' (legal) polydrug use or by morphine equivalent dosing, nor was there any modification to the analysis of the data to see how this period in the legal drug market impacted patient outcomes. 2. This analysis selects for chronic pain patients, which tends to mean the long term users of opioids while excluding the acute patients who, particularly among the in-patient subset, pretty much never become addicts during their course of use - which is what we'd expect if these compounds were as addictive as claimed. 3. These cohort selections often do not include some or all palliative care patients. I won't accuse them of doing this to sensationalize their numbers but it does have a significant impact on results, as again palliative care patients very rarely cease their use and carry forward afterwards an opioid misuse disorder. 4. They do not control for or look at the specific impact of the main problematic legal opioid - Purdue Pharma's original OxyContin formula. Because this drug did not have the pharmacodynamic profile claimed by the firm and represented in the safety studies, it was considerably more likely to cause opioid use disorder, as patients would be much more likely to take extra doses to cover the shortfall in the dosing window. This would lead to patients who otherwise would exist on a steady dose asking for increasing doses - or, worse, turning to the street to acquire diverted doses.