Commentary on “Methodologic limitations of prescription opioid safety research and recommendations for improving the evidence base”