Send to

Choose Destination
J Arthroplasty. 2017 Aug;32(8):2390-2394. doi: 10.1016/j.arth.2017.03.014. Epub 2017 Mar 16.

Opioid Use After Total Knee Arthroplasty: Trends and Risk Factors for Prolonged Use.

Author information

University of Iowa Hospitals and Clinics, Department of Orthopaedics, Iowa City, Iowa.



The United States is in the midst of an opioid epidemic. Little is known about perioperative opioid use for total knee arthroplasty (TKA). The purpose of this study was to identify rates of preoperative opioid use, evaluate postoperative trends and identify risk factors for prolonged use after TKA.


Patients who underwent primary TKA from 2007-2014 were identified within the Humana database. Postoperative opioid use was measured by monthly prescription refill rates. A preoperative opioid user (OU) was defined by history of opioid prescription within 3 months prior to TKA and a non-opioid user (NOU) was defined by no history of prior opioid use. Rates of opioid use were trended monthly for one year postoperatively for all cohorts.


73,959 TKA patients were analyzed and 23,532 patients (31.2%) were OU. OU increased from 30.1% in 2007 to 39.3% in 2014 (P < .001). Preoperative opioid use was the strongest predictor for prolonged opioid use following TKA, with OU filling significantly more opioid prescriptions than NOU at every time point analyzed. Younger age, female sex and other intrinsic factors were found to significantly increase the rate of opioid refilling following TKA throughout the postoperative year.


Approximately one-third of TKA patients use opioids within 3 months prior to surgery and this percentage has increased over 9% during the years included in this study. Preoperative opioid use was most predictive of increased refills of opioids following TKA. However, other intrinsic patient characteristics were also predictive of prolonged opioid use.


narcotic; opioid; risk factors; total knee arthroplasty; trends

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center