Intrigued by the study by Buckeridge and colleagues,1 we examined fall and opioid use data in 352 long-term care skilled nursing beds in three Eddy Foundation facilities in the Capital Region of New York State. We excluded a subacute care unit where patients commonly receive pain medication for postoperative pain. We extracted falling data for the second quarter of 2011. Of 222 falls, 126 (56.7%) occurred in facility residents who were using opioids, whereas only 27.5% of residents were receiving opioids (likelihood of falling 0.140 per patient per month for those receiving any opioid prescription vs 0.041 for those not receiving any opioid, relative risk = 3.45, 95% confidence interval = 2.74–4.34). Of the 126 falls, 23 (18.3%) were associated with use of a short-acting opioid within the prior 4 hours (five tramadol, 17 hydrocodone, one oxycodone). Only one (0.8%) was associated with use of a long-acting opioid within the prior 6 hours (MS-Contin, with a recent doubling of dosage in this individual). The finding in the Buckeridge study, that fall risk was more associated with the introduction of a new or increased hydrocodone dose than with more-potent opioids, seems counterintuitive. Nevertheless, our limited data are consistent with this finding, suggesting an explanation; as an “entry” drug, hydrocodone is commonly given to patients who are opioid-naive. This may result in greater susceptibility to the opioid's effects on level of consciousness or stability of gait, particularly with rapid onset of effect. We are now designing an interventional study to monitor fall rates as we further encourage reduced use of short-acting opioids and reduced use of as-needed orders. The available evidence appears to support such interventions. Conflict of Interest: There were no conflicts of interest for any of the authors. Author Contributions: M Wolff: concept, data analysis, and writing. R Kewley: concept, data extraction, and analysis. MC Hassett: data extraction and analysis. J Collins: concept, data analysis, and review. M Brodeur: concept and data review. S Nokes: data extraction. Sponsor's Role: There was no direct sponsorship of this study.
Read full abstract