Prescribing opioids to older persons in pain involves potential trade-offs between pain control and untoward negative health consequences, one of which may be injuries. Good data to guide clinical judgment about the absolute and relative risk of fatal and non-fatal injuries among opioid users are lacking. The proposed prospective cohort study will use a large health care data set and pharmacoepidemiologic methods to generate new, clinically relevant information about the risk of injury among elderly Americans using prescription opioids. The study population consist of low-moderate income elderly Americans who received prescription drug benefits through the states of Pennsylvania and New Jersey, 1994-2005. The study will compare 1) the rate of selected injuries among elderly persons taking opioids to those not, and 2) variation in the risk of injury among opioid users. Outcomes will include fatal and non-fatal injuries resulting from falls, motor vehicle collisions, unintentional overdoses, and suicide attempts. Basic tables will describe the study population, crude outcome rates, and adjusted associations between opioid use and injury outcomes. The time-varying nature of injuries after initiation of opioid treatment will be rendered by plotting Kaplan-Meier curves for event rates of select injuries among opioid users as a function of the duration of continuous opioid use. The relative risk of injuries by potency, duration of action, and duration of use will be estimated using Cox proportional hazard regression models with time-varying exposure, adjusting for baseline patient characteristics. Interaction terms will identify factors that exacerbate the risk of injury associated with opioid use. Factors to be explored include patient characteristics, specific preparations, and opioid use that appears problematic based on explicit prescribing guidelines. The proposed work is important because injury, injury deaths, and opioid use are common among elderly Americans. In 2004, for example, older Americans made nearly three million visits to emergency departments for injuries, 1.8 million resulted from falls (25% required hospitalization). An unknown fraction of these injuries occurred among the 5% of community dwelling elderly who use prescription opioids. Results from the proposed study will be relevant to clinicians treating elderly patients in pain and to public health practitioners, since even small increases in the relative risk of injury attributable to opioids could amount to thousands of injuries each year. Since many older Americans benefit from opioids, a null finding or a finding that risk is concentrated (e.g., in some but not other specific preparations) would also be clinically relevant. ? ? ?
| Miller, Matthew; Stürmer, Til; Azrael, Deborah et al. (2011) Opioid analgesics and the risk of fractures in older adults with arthritis. J Am Geriatr Soc 59:430-8 |
| Simard, Julia F; Stoll, Matthew L; Shadick, Nancy A et al. (2010) Validity of self-report of infections in a longitudinal cohort of patients with rheumatoid arthritis differs by source of report and infection severity. J Clin Epidemiol 63:1358-62 |
| Solomon, D H; Hochberg, M C; Mogun, H et al. (2009) The relation between bisphosphonate use and non-union of fractures of the humerus in older adults. Osteoporos Int 20:895-901 |
