Do Nonsteroidal Anti-Inflammatory or COX-2 Inhibitor Drugs Increase the Nonunion or Delayed Union Rates After Fracture Surgery?: A Propensity-Score-Matched Study

J Bone Joint Surg Am. 2021 Aug 4;103(15):1402-1410. doi: 10.2106/JBJS.20.01663.


Background: The effects of nonsteroidal anti-inflammatory drugs (NSAIDs)/cyclooxygenase (COX)-2 inhibitors on postoperative fracture-healing are controversial. Thus, we investigated the association between NSAID/COX-2 inhibitor administration and postoperative nonunion or delayed union of fractures. We aimed to determine the effects of NSAID/COX-2 inhibitor administration on postoperative fracture-healing with use of a common data model.

Methods: Patients who underwent operative treatment of a fracture between 1998 and 2018 were included. To determine the effects of NSAID/COX-2 inhibitor administration on fracture-healing, postoperative NSAID/COX-2 inhibitor users were compared and 1:1 matched to nonusers, with 3,264 patients matched. The effect of each agent on bone-healing was determined on the basis of the primary outcome (nonunion/delayed union), defined as having a diagnosis code for nonunion or delayed union ≥6 months after surgery. The secondary outcome was reoperation for nonunion/delayed union. To examine the effect of NSAIDs/COX-2 inhibitors on bone union according to medication duration, a Kaplan-Meier survival analysis was performed.

Results: Of the 8,693 patients who were included in the analysis, 208 had nonunion (178 patients; 2.05%) or delayed union (30 patients; 0.35%). Sixty-four (30.8%) of those 208 patients had a reoperation for nonunion or delayed union. NSAID users showed a significantly lower hazard of nonunion compared with the matched cohort of nonusers (hazard ratio, 0.69 [95% confidence interval, 0.48 to 0.98]; p = 0.040) but did not show a significant difference in the other matched comparison for any other outcomes. Kaplan-Meier survival analysis revealed significantly lower and higher nonunion/delayed union rates when the medication durations were ≤3 and >3 weeks, respectively (p = 0.001). For COX-2 inhibitors, the survival curve according to the medication duration showed no significant difference among the groups (p = 0.9).

Conclusions: Our study demonstrated no short-term impact of NSAIDs/COX-2 inhibitors on long-bone fracture-healing. However, continued use of these medications for a period of >3 weeks may be associated with higher rates of nonunion or delayed union.

Level of evidence: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

MeSH terms

  • Adult
  • Aged
  • Anti-Inflammatory Agents, Non-Steroidal / administration & dosage
  • Anti-Inflammatory Agents, Non-Steroidal / adverse effects*
  • Cyclooxygenase 2 Inhibitors / administration & dosage
  • Cyclooxygenase 2 Inhibitors / adverse effects*
  • Female
  • Fracture Fixation / adverse effects
  • Fracture Healing / drug effects*
  • Fractures, Ununited / chemically induced
  • Fractures, Ununited / epidemiology*
  • Fractures, Ununited / surgery
  • Humans
  • Male
  • Middle Aged
  • Pain, Postoperative / drug therapy*
  • Pain, Postoperative / etiology
  • Reoperation / statistics & numerical data
  • Retrospective Studies
  • Risk Factors
  • Time Factors


  • Anti-Inflammatory Agents, Non-Steroidal
  • Cyclooxygenase 2 Inhibitors