The coexistence of femoral shaft fractures and ipsilateral femoral neck fractures (IFNFs) presents challenges for both diagnosis and management. Femoral neck fractures have been reported in up to 9% of femoral shaft fractures.1 During high-energy femoral shaft fractures, IFNFs are missed or have delayed diagnosis in 20 to 50% of cases.2–7

Optimal diagnostic methods for femoral neck fractures associated with femoral shaft fractures have been a subject of intense inquiry. O'Toole et al2 compared the diagnostic accuracy of plain radiography and CT in detecting femoral neck fractures. They discovered substantial rates of missed fractures for both imaging modalities, highlighting the significance of intraoperative and postoperative imaging for identifying these injuries. Tornetta et al3 also emphasized the role of fine-cut CT scans and dedicated internal rotation hip radiographs in enhancing the diagnostic capability for associated femoral neck fractures. Even with these improved diagnostic capabilities, Tornetta reported 9% of femoral neck fracture were missed, highlighting the need for further research and protocol improvement. Other diagnostic modalities used to identify IFNFs in the setting of femoral shaft fractures are preoperative MRI and the presence of capsular distention.8

IFNFs occurring alongside femoral shaft fractures are challenging injuries that require early diagnosis for optimal management. This retrospective cohort study seeks to ascertain the risk factors associated with IFNFs in patients with femoral shaft fractures. By delineating these risk factors and fracture characteristics, we hope that treating physicians can promptly identify and effectively treat these injuries.

Methods Patient Population and Outcome Measures

After IRB approval, a retrospective cohort study of femoral shaft fractures with IFNFs at three level I trauma centers over a 5-year period (2017 to 2022) was identified by their combined current procedural terminology codes (one of 27235 or 27236 and one of 27244, 27245, 27506, or 27507). Patient demographic data such as age; sex; fracture characteristics, including open or closed, fracture displacement on radiographs, and mechanism of injury (MOI); and Gustilo classification were recorded. All patients received plain radiograph studies, including AP pelvis, AP/lateral injured hip, AP/lateral injured femur, and AP/lateral injured knee, and CT scans with 2 mm cuts of the injured proximal femur and pelvis. Fracture displacement was measured on initial AP and lateral radiographs of the femur taken out of traction splints. Fracture displacement was measured as a function of femoral shaft diameter at the level of fracture. In addition, intraoperative fluoroscopy of the proximal femur and postoperative AP imaging of the ipsilateral hip were all obtained. Fall from height was defined as a fall greater than 10 feet. Fractures were classified by Orthopedic Trauma Association (OTA)/Arbeitsgemeinshaft fur Osteosynthesefragen (AO) classification and Winquist classification.4

Inclusion criteria consisted of age 18 years or older, isolated femoral shaft or combined femoral shaft and IFNFs, and the availability of all imaging studies listed above. Patients without all imaging studies were excluded. All patients with OTA/AO 31 and 33 fractures were excluded from the study. A comparison group of 280 isolated femoral shaft fractures was created from the same 5-year study period using their current procedural terminology codes (27506, 27507). Patients were added in consecutive order based on combined fractures, i.e., once a combined fracture was identified, the next two isolated femoral shaft fractures were consecutively added. We did this to ensure an even split of combined fractures and isolated fractures across the study period. All demographic information, fracture characteristics, and radiographic data were collected for the comparison group.

Statistical Analysis

Descriptive statistics were done to determine study population demographics (age, sex), fracture characteristics (laterality, open versus closed, displacement, and MOI), and fracture classification (Gustilo, Winquist, and OTA/AO). Mann-Whitney U tests were done to determine differences in age, fracture displacement, and Winquist classification. Pearson chi square was done to determine differences in sex, open versus closed fracture, Gustilo classification, MOI, and OTA/AO classification. Multivariate analysis was done with all notable variables on univariate analysis. Adjusted odds ratios (aORs) and 95% confidence intervals were reported for all variables. The variables with a significant P-value were included in the predicted probability model that was used to estimate the probability of an IFNF for each combination of predictors included. In addition, a receiver operating characteristics curve was generated using this model to assess the diagnostic performance of the group of multivariate predictors in diagnosing an IFNF. IBM SPSS Statistics v29.0 (Armonk, NY) was used for statistical analysis. Significance was defined as P-value < 0.05.

Results Demographics

A total of 140 patients were identified with femoral shaft fracture and IFNF. A control group of 280 patients with isolated femoral shaft fractures were identified during the same study period for comparison. The average age of patients was 38.1 years ± 18.4 years and 38.5 ± 14.3 years in the isolated femoral shaft fracture group and IFNF group, respectively (Table 1). Most of the patients in both groups were men (71.4% and 62.9%, P = 0.075).

Table 1 - Demographic Characteristics for Both Cohorts Demographics Isolated Femoral Shaft Fracture (n = 280) Ipsilateral Femoral Neck Fracture (n = 140) P Age, mean ± SD 38.1 ± 18.4 38.5 ± 14.3 0.115 Sex, n (%) 0.075  Male 200 (71.4%) 88 (62.9%)  Female 80 (28.6%) 52 (37.1%)
Fracture Characteristics

No significant difference was observed in open and closed fractures between groups (P = 0.46). No significant difference was observed in fracture displacement (85.4% versus 95.8%, P = 0.118), Winquist classification (2.06 versus 2.02, P = 0.631), or Gustilo classification (P = 0.507) between cohorts. A significant difference was observed in MOI between cohorts (P < 0.001). Motor vehicle crashes (MVCs) were the most common MOI in both the isolated femoral shaft and IFNF groups (43.7% versus 61.5%). A significant difference was observed in OTA/AO classification between both groups (P = 0.002). OTA/AO 32A fractures were the most common femoral shaft fracture (40.5%), followed by OTA/AO 32B (39.1%) and OTA/AO 32C (20.4%). OTA/AO 32A fractures were the most common IFNF (50.7%), followed by OTA/AO 32C (28.6%) and OTA/AO 32B (20.7%). A significant difference was observed in fracture location between both groups (P < 0.001). Isthmic fractures were the most common location for femoral shaft fractures (45.4%), followed by infraisthmic (38.2%) and proximal to the isthmus (16.4%). Isthmic fractures were the most common location for IFNFs (72.9%), followed by infraisthmic (17.1%) and proximal to the isthmus (10.0%; Table 2).

Table 2 - Fracture Characteristics for Both Cohorts Fracture Characteristics Isolated Femoral Shaft Fracture (n = 280) Ipsilateral Femoral Neck Fracture (n = 140) P Closed or open fracture, n (%) 0.460  Closed 213 (76.1%) 111 (79.3%)  Open 67 (23.9%) 29 (20.7%) Fracture displacement (%), mean ± SD 85.3 ± 52.3 95.8 ± 53.1 0.118 Winquist, mean +/SD 2.06 ± 1.30 2.02 ± 1.43 0.631 Gustilo classification 0.507  0 213 (76.1%) 111 (79.3%)  I 21 (7.5%) 8 (5.7%)  II 27 (9.6%) 8 (5.7%)  IIIa 5 (1.8%) 3 (2.1%)  IIIb 12 (4.3%) 9 (6.4%)  IIIc 2 (0.7%) 1 (0.8%) Mechanism, n (%) < 0.001  MVC 122 (43.7%) 86 (61.5%)  MCC 44 (15.7%) 40 (28.6%)  Fall from height 25 (8.9%) 9 (6.4%)  Pedestrian struck 2 (0.7%) 1 (0.7%)  Extreme sport 6 (2.1%) 0 (0%)  Sport 30 (10.7%) 2 (1.4%)  GSW 51 (18.2%) 2 (1.4%) OTA/AO classification segment 32, n (%) 0.002  32A 113 (40.5%) 71 (50.7%)  32B 109 (39.0%) 29 (20.7%)  32C 57 (20.5%) 40 (28.6%) Segment 32 fracture location, n (%) < 0.001  Proximal to isthmus 46 (16.4%) 14 (10.0%)  Transisthmic 127 (45.4%) 102 (72.9%)  Infraisthmic 107 (38.2%) 24 (17.1%)

GSW = Gunshot wound, MVC = motor vehicle crash, MCC = motorcycle crash.

Bolded items demonstrate statistical significance.

Multivariate Analysis

Multivariate analysis was done for all notable variables on univariate analysis. MVCs were more commonly associated with IFNFs when compared with sport (aOR = 0.12, P = 0.006) and gun shot wound (aOR = 0.06, P < 0.001). OTA/AO 32A fractures were more commonly associated with IFNFs when compared with OTA/AO 32B fractures (aOR = 0.36, P < 0.001). Fractures through the isthmus were more commonly associated with IFNFs than fractures more proximal (aOR = 2.52, P = 0.011). No significant difference was observed in IFNFs between OTA/AO 32C and 32A fractures (aOR = 0.85, P = 0.566). No significant difference was observed in IFNFs between fractures proximal to the isthmus and fractures infraisthmic (aOR = 0.74, P = 0.455; Table 3).

Table 3 - Multivariate Analysis of Risk Factors for Combined Femoral Shaft Fracture and Ipsilateral Femoral Neck Fracture Compared With Femoral Shaft Fracture Alone Risk Factor aOR (95% CI) P value Mechanism  MVC Reference N/A  MCC 1.40 (0.81–2.43) 0.225  Fall from height 0.63 (0.26–1.48) 0.286  Pedestrian struck 0.69 (0.06–8.49) 0.772  Extreme sport N/A N/A  Sport 0.12 (0.03–0.54) 0.006  GSW 0.06 (0.01–0.26) < 0.001 OTA/AO classification segment 32  32A Reference N/A  32B 0.35 (0.20–0.62) < 0.001  32C 0.85 (0.48–1.49) 0.566 Segment 32 fracture location  Proximal to isthmus Reference N/A  Transisthmic 2.52 (1.24–5.11) 0.011  Infraisthmic 0.74 (0.33–1.64) 0.455

GSW = Gunshot wound, MVC = motor vehicle crash, MCC = motorcycle crash

Bolded items demonstrate statistical significance.

Receiver Operator Curve

A receiver operator curve analysis was done for the three variables included in the multivariate analysis. Owing to the need for binary variables, the input for OTA/AO classification was 32A versus 32B and 32C, fracture location was transisthmic versus proximal to the isthmus and infraisthmic, and MOI was MVC versus all other mechanisms of injury (Table 4). Our results show if a patient in a MVC sustained an OTA/AO 32A femoral shaft fracture through the isthmus, their probability of having an IFNF was 59% (Figure 1).

Table 4 - Receiver Operator Curve Predicted Probability of Sustaining an Ipsilateral Femoral Neck Fracture in the Presence of a Femoral Shaft Fracture Multivariate Predictor Predicted OTA/AO 32A Transisthmic Fracture MVC Probability (%) Yes Yes Yes 59.41 Yes Yes No 44.24 Yes No Yes 31.31 Yes No No 19.82 No Yes Yes 45.16 No Yes No 30.86 No No Yes 20.41 No No No 12.21

MVC = motor vehicle crash

Figure 1:

Receiver operator curve derived from OTA/AO 32A classified fractures, fractures through the isthmus, and motor vehicle collision as the mechanism of injury.

Discussion

Femoral shaft fractures can have IFNFs that can often be occult. Early identification of IFNFs is critical to prevent potentially devastating complications. Current diagnostic methods have varying sensitivity in identification of IFNFs. The results of our study can be used to aid in the detection of IFNFs in the setting of femoral shaft fractures. Our study indicates that patients with femoral shaft fractures through the isthmus were 2.5 times more likely to have a concomitant IFNF compared with those with femoral shaft fractures proximal to the isthmus. Motorcycle crashes and MVCs were the most notable MOI associated with IFNFs, while falls from height, sport injury, and gunshot wounds were the least associated with IFNFs.

Fractures of the femoral shaft combined with IFNFs are commonly associated with high-energy traumatic events such as MVCs or falls from height.5,6,9,10 In our study, it was observed that 61.5% of these MCCs, whereas 28.6% were attributed to motorcycle collisions. Motorcycle collisions were the strongest predicting mechanism for IFNFs. Early detection of femoral neck fractures is essential because of complications, such as nonunion and osteonecrosis, that can occur when surgical treatment is not appropriately initiated. The incidence of nonunion after sustaining a combined femoral shaft and femoral neck fracture has been reported at 5%, while osteonecrosis rates range from 15% to 30%.11 Such complications are associated with notable morbidity.11,12

Several studies have created protocols to assess for IFNFs, but missed or delayed diagnosis is still reported in 20% to 50% of cases (Figure 2, Figure 3).2,3,5–7,13 This challenge has been noted in multiple studies and is largely due to nondisplaced or minimally displaced femoral neck fractures.7,13–15 Owing to this rate of missed injuries, Tornetta et al3 developed a protocol including 2-mm cut CT scan and internal rotation hip radiographs. This protocol effectively reduced the rates of delayed diagnosis by 91% compared with the previous year. However, even with such improvement, there is still a 9% missed diagnosis rate, indicating the need for additional study and protocol improvement. O'Toole et al and Yang et al also created protocols for the detection of IFNFs in the setting of femoral shaft fractures. However, O'Toole's protocol of hip radiographs and CT scan produced a sensitivity of < 65%, and Yang's protocol of CT scan yielded a 75% sensitivity.2,14 Park and Song et al developed an additional protocol of lipohemarthrosis detection on CT scan in the setting of high-injury femoral shaft fractures to indicate preoperative MRI or prophylactic femoral neck fixation.8 Lipohemarthosis was identified by anterior capsular distension > 1 mm on a soft-tissue window CT scan. In the results of the 29 patients with positive lipohemarthrosis on CT, five patients were identified to have IFNFs on preoperative MRI or future imaging studies postoperatively. The remaining 119 without positive lipohemarthrosis on CT had no IFNFs on future radiographs. Their prophylactic femoral neck fixation using CT positive lipohemarthrosis was successful in preventing missed or delayed complications of IFNFs.8

Figure 2:

Preoperative femoral shaft fracture radiograph (A) and CT (B) showing no signs of femoral neck fracture.

Figure 3:

Postoperative femoral shaft fracture radiograph (A) and CT (B) showing an ipsilateral femoral neck fracture.

Recently, MRI and intraoperative fluoroscopy have been used to aid in the diagnosis of IFNFs.16–18 Avilucea et al.16 reported a 92% sensitivity using dynamic stress fluoroscopy for the diagnosis of IFNFs. In addition, Rogers et al added rapid-sequence MRI to their protocol. They reported diagnosing IFNFs in the setting of high-energy femoral shaft fractures of six patients on plain radiographs, three additional patients on CT, and 10 additional patients on rapid-sequence MRI. No missed or delayed diagnoses were reported in their 121 femoral shaft fracture study group.17 Although these are promising findings, MRI availability, patient pain during transfer, and MRI cost are all problematic.

Our study showed a notable difference among OTA/AO classifications, with segment 32A fractures being more associated with IFNFs when compared with segment 32B fractures in multivariate analysis. In addition, segment 32 fractures through the isthmus were 2.5 times more likely to be associated with IFNFs than fractures proximal to the isthmus. Multiple studies have tried to assess fracture classification as a predictor of IFNF but have failed to reach consensus. Park et al and Randelli et al reported higher rates of IFNF in patients with more comminuted fracture patterns.12,19 However, Chen et al20 reported higher rates of IFNF in patients with less comminuted fractures. In addition, Rennard et al reported that AO/OTA 32-C fracture types had the highest association with ipsilateral AO/OTA 31 type fractures. However, their study included a small number of IFNFs (41) compared with isolated fractures (187).18 No statistics were done, so no statistical significance could be determined.

Some studies have supported the use of antegrade reconstruction nailing of all femoral shaft fractures to stabilize any occult femoral neck fractures.21 This technique has been found to possibly be cost-effective in light of the tremendous costs of missed femoral neck fractures.22 However, this cost analysis is dependent on the implant cost to each health system and does not take into account the feasibility of nailing every femoral shaft fracture antegrade based on fracture characteristics, patient body habitus, and other technical factors.22

This study is not without limitation. First, the relatively low number of femoral shaft fractures with IFNFs limits the statistical power of outcomes. Second, our findings may not be generalizable because of the subjective nature of OTA/AO and Winquist classifications. Furthermore, we only assessed differences in OTA/AO 32 fracture types. Finally, there is inherent bias because of the retrospective study design. To address these limitations, future studies should prioritize a larger patient cohort to gain a more comprehensive understanding of the relationship between IFNF and radiographic characteristics. The ability to identify such risk factors can assist in protocol formation for MRI utilization and prophylactic femoral neck fixation.

Conclusion

Accurate detection of IFNFs in the setting of femoral shaft fractures remains a diagnostic challenge. Although diagnosis remains difficult, injury due to MVC or motorcycle collision, AO/OTA type 32A fractures, and fractures through the isthmus all suggests an increased possibility of an IFNF. This knowledge can be valuable in guiding decisions regarding imaging protocols and fixation approaches for minimally displaced or occult fractures of the femoral neck associated with femoral shaft fractures.

Acknowledgement

We would like to express our sincere gratitude to Jiahao Peng, MD, MPH, for his contribution in conducting the statistical analyses for this research manuscript. His expertise and dedication greatly enhanced the rigor and quality of our study. Thank you for your support throughout this research endeavor.

References 1. Alho A: Concurrent ipsilateral fractures of the hip and shaft of the femur. A systematic review of 722 cases. Ann Chir Gynaecol 1997;86:326-336. 2. OʼToole RV, Dancy L, Dietz AR, et al.: Diagnosis of femoral neck fracture associated with femoral shaft fracture: Blinded comparison of computed tomography and plain radiography. J Orthop Trauma 2013;27:325-330. 3. Tornetta P 3rd, Kain MS, Creevy WR: Diagnosis of femoral neck fractures in patients with a femoral shaft fracture. Improvement with a standard protocol. J Bone Joint Surg Am 2007;89:39-43. 4. Meinberg EG, Agel J, Roberts CS, Karam MD, Kellam JF: Fracture and dislocation classification compendium – 2018. J Orthop Trauma 2018;32suppl 1:S1-S170. 5. Loizou CL, Parker MJ: Avascular necrosis after internal fixation of intracapsular hip fractures; a study of the outcome for 1023 patients. Injury 2009;40:1143-1146. 6. Riemer BL, Butterfield SL, Ray RL, Daffner RH: Clandestine femoral neck fractures with ipsilateral diaphyseal fractures J Orthop Trauma 1993;7:443-449. 7. Boulton CL, Pollak AN: Special topic: Ipsilateral femoral neck and shaft fractures – does evidence give us the answer? Injury 2015;46:478-483. 8. Ritchey SJ, Schonholtz GJ, Thompson MS: The dashboard femoral fracture: Pathomechanics, treatment, and prevention. The J bone joint Surg Am Volume 1958;40-A:1347-1358. 9. Wang C, Xu G, Han Z, et al.: Correlation between residual displacement and osteonecrosis of the femoral head following cannulated screw fixation of femoral neck fractures. Medicine 2015;94:e2139. 10. Jones CB, Walker JB: Diagnosis and management of ipsilateral femoral neck and shaft fractures. The J Am Acad Orthopaedic Surgeons 2018;26:e448-e454. 11. Park YC, Um KS, Hong SP, Oh CW, Kim S, Yang KH: Preoperative “Computed tomography capsular sign” for the detection of occult ipsilateral femoral neck fractures associated with femoral shaft fractures. Injury 2020;51:1051-1056. 12. Peljovich AE, Patterson BM: Ipsilateral femoral neck and shaft fractures. J Am Acad Orthop Surg 1998;6:106-113. 13. Yang KH, Han DY, Park HW, Kang HJ, Park JH: Fracture of the ipsilateral neck of the femur in shaft nailing. The role of CT in diagnosis. J Bone Joint Surg Br 1998;80:673-678. 14. Gümüştaş SA, Tosun HB, Isyar M, Serbest S, Oznam K, Bulut G: Femur neck fracture in young adults, is it really an urgent surgery indication: Retrospective clinical study. Pan Afr Med J 2018;30:112. 15. Park YC, Song HK, Yang KH: Prevention of unplanned surgery due to delayed diagnosis of occult ipsilateral femoral neck fractures associated with femoral shaft fractures: A study of the CT capsular sign with lipohemarthrosis. J Bone Joint Surg Am 2021;103:1431-1437. 16. Avilucea FR, Joyce D, Mir HR: Dynamic stress fluoroscopy for evaluation of the femoral neck after intramedullary nails: Improved sensitivity for identifying occult fractures. J Orthop Trauma 2019;33:88-91. 17. Rogers NB, Achor TS, Kumaravel M, et al.: Implementation of a novel MRI protocol for diagnosing femoral neck fractures in high energy femoral shaft fractures: One year results. Injury 2021;52:2390-2394. 18. Rennard J, Rogers N, Achor T, et al.: The impact of magnetic resonance imaging on the diagnosis of high-energy ipsilateral femoral neck and shaft fractures. J Orthop Trauma 2022;36:93-97. 19. Randelli P, Landi S, Fanton F, Hoover GK, Morandi M: Treatment of ipsilateral femoral neck and shaft fractures with the Russell-Taylor reconstructive nail. Orthopedics 1999;22:673-676. 20. Chen CH, Chen TB, Cheng YM, Chang JK, Lin SY, Hung SH: Ipsilateral fractures of the femoral neck and shaft. Injury 2000;31:719-722. 21. Collinge C, Liporace F, Koval K, Gilbert GT: Cephalomedullary screws as the standard proximal locking screws for nailing femoral shaft fractures. J Orthop Trauma 2010;24:717-722. 22. Faucett SC, Collinge CA, Koval KJ: Is reconstruction nailing of all femoral shaft fractures cost effective? A decision analysis. J Orthop Trauma 2012;26:624-632.