To Fix or Revise

Differences in Periprosthetic Distal Femur Fracture Management Between Trauma and Arthroplasty Surgeons

Noelle L. Van Rysselberghe, MD; Sean T. Campbell, MD; L. Henry Goodnough, MD, PhD; Derek F. Amanatullah, MD, PhD; Michael J. Gardner, MD; Julius A. Bishop, MD


J Am Acad Orthop Surg. 2022;30(1):e17-e24. 

In This Article


DFR and surgical fixation are both viable treatment options for periprosthetic distal femur fractures above total knee arthroplasties, but limited evidence exists to guide treatment decisions. The purpose of this study was to use a case-based survey to evaluate for differences in the likelihood of recommending fixation versus revision to DFR between surgeons who had or had not completed a trauma fellowship. The most important finding of this study was that surgeons who completed a trauma fellowship were markedly more likely to choose fracture fixation compared with arthroplasty surgeons, particularly for comminuted fractures proximal to the anterior flange and for any fracture distal to the anterior flange. When excluding simple proximal fractures, completion of a trauma fellowship was associated with a 2.7 times greater likelihood of recommending fracture fixation over reconstructive arthroplasty compared with arthroplasty surgeons (OR 2.78, 95% CI 2.11 to 3.65, P < 0.0001). In addition, high-volume surgeons who treat more than three periprosthetic distal femur fractures monthly were significantly more likely to recommend fixation than lower volume surgeons. This effect of surgeon experience on the decision to fix or revise was seen for all fracture patterns other than simple proximal fractures (where both groups unanimously recommended fixation).

As discussed previously, in the absence of a loose total knee arthroplasty, there are no widely accepted indications for revision to DFR over fracture fixation. Previous studies have attempted to address this question but are limited by their nonrandomized and retrospective nature, different complication profiles between treatments, and a tendency to have a longer follow-up after revision arthroplasty than fracture surgery.[9,18,19] In the absence of evidence-based surgical indications, choice of treatment relies on surgeon discretion, introducing the possibility of multiple types of bias.

This study demonstrates that trauma fellowship training is associated with a preference toward more frequent surgical fixation of periprosthetic distal femur fractures, as compared with surgeons with an arthroplasty background. In addition, arthroplasty fellowship training may be associated with a decreased preference for surgical fixation because surgeons who completed fellowships in both arthroplasty and trauma had a significantly lower preference for fixation over revision arthroplasty, as compared with surgeons who had completed a trauma fellowship alone.

With no obviously correct treatment choice, explanations for this discrepancy can include multiple types of cognitive bias. Availability bias involves misattributing the probability of an event occurring to its availability in one's memory; in other words, overvaluing one's anecdotal experience.[20] For example, arthroplasty surgeons may overestimate the rate of nonunion after periprosthetic distal femur fracture fixation because they are more likely to see in their own practice patients who failed an attempt at fixation and are now being referred for DFR. By contrast, patients who go on to heal do not typically re-present to the arthroplasty clinic. Observational selection bias may also play a role because surgeons' personal expertise can affect the details extracted from a situation. In the case of periprosthetic distal femur fractures, arthroplasty surgeons may appreciate more subtle problems with the initial total knee arthroplasty that may push them toward revision. Furthermore, whether a fracture is "too distal to fix" is directly related to distal screw options in an intramedullary nail or lateral locking plate, which have been increased in modern implants. Because trauma surgeons have greater familiarity with the newer versions of these implants, this may directly affect their estimation of fixability, whereas arthroplasty surgeons may remember older versions of these implants where distal fixation options were more limited. This may also explain why the high-volume surgeons were more likely to recommend fixation for fracture patterns distal to the anterior flange compared with lower volume surgeons. Ego bias refers to distorting information in a way that serves one's ego. This has been shown to be widespread throughout surgical decision-making because surgeons across specialties tend to underestimate their own complication rates.[20] Regarding the major complications after periprosthetic distal femur fractures, this means that trauma surgeons are likely to underestimate their personal rates of nonunion, whereas arthroplasty surgeons are likely to underestimate their rates of periprosthetic infection after DFR. Finally, there are multiple explanations for the differing ways in which trauma surgeons and arthroplasty surgeons value the potential outcomes of each procedure, which are beyond the scope of this study. The significance of preserving bone stock and/or avoiding infection of a megaprosthesis may be interpreted differently by trauma versus arthroplasty surgeons.

This study found that trauma and arthroplasty surgeons disagreed most frequently in the management of comminuted fractures proximal to the anterior flange, as well as all fracture patterns distal to the anterior flange. Reasons for this are unclear, but indicate that these are challenging fracture patterns for which optimal management is not obvious, and individual surgeon therefore play a bigger role.

The results of this study should be interpreted in the context of multiple limitations. First, because the survey was anonymous and distributed through a website link, the true response rate cannot be calculated. Because the study was posted on the OTA website and circulated between sites for the Insall Knee Society Fellowship, this did bias the respondent population toward academic practice (69%). We attempted to increase representation by forwarding the link to additional institutions, but more surgeons in academic practices ultimately completed the survey. Although the number of responses represents a fraction of the total OTA membership and practicing arthroplasty surgeons, the sample is still large enough to draw notable conclusions about treatment preferences between groups. The sample size is also consistent with other published survey studies evaluating trends in orthopaedic surgery.[21–23]

In conclusion, this study demonstrated notable variation in treatment recommendations for periprosthetic distal femur fractures above total knee arthroplasties between surgeons who had and had not completed a trauma fellowship, as well as between surgeons who have a high versus low volume of these injuries in their practice. These results highlight the need for more objective studies comparing outcomes after surgical fixation versus DFR so as to decrease the role of subjective bias in this important treatment decision. By describing clinical outcomes and complication rates after each procedure, surgical indications may be clarified and quality of care increased.