Injury 55 (2024) 111463

Contents lists available at ScienceDirect

Injury
journal homepage: www.elsevier.com/locate/injury

No difference in clinical and hospital quality outcomes in treatment of

reverse obliquity intertrochanteric Hip fractures (AO/OTA 31.A3) based on
Cephalomedullary nail length
Sanjit R. Konda a, b, Lauren A. Merrell a, Garrett W. Esper a, Kester Gibbons a, Rachel Ranson a,
Abhishek Ganta a, b, Kenneth A. Egol a, b, *
a
Division of Orthopedic Trauma Surgery, Department of Orthopedic Surgery, NYU Langone Health, NYU Langone Orthopedic Hospital, New York, NY, USA
b
Department of Orthopedic Surgery, Jamaica Hospital Medical Center, Richmond Hill, NY, USA

A R T I C L E I N F O A B S T R A C T

Level of Evidence: III Introduction: The purpose of this study was to evaluate outcomes following reverse obliquity (RO) inter­
Declarations trochanteric hip fractures based on the use of short cephalomedullary nails (CMNs) compared to long CMNs for
fixation.
Key Words: Methods: An IRB-approved prospectively collected hip fracture registry at an urban academic medical center was
Hip fracture
queried for all AO/OTA 31A3.1–3 reverse obliquity intertrochanteric (RO) fractures. One hundred and seventy
Reverse obliquity
patients with age > 55 years old and minimum 6-month follow-up were identified for analysis. Data was
Cephalomedullary nail
Time to Healing collected for patient demographics, injury details, intraoperative radiographic parameters, perioperative phys­
Outcomes iologic parameters, hospital quality measures, and outcomes including radiographic time to healing, need for
reoperation, nonunion, and mortality. Comparative analyses were conducted between cohorts. Additional
multivariable binary logistic and linear regression analyses were performed to evaluate for factors independently
associated with short and long nail usage.
Results: The mean age of the entire cohort was 80.91±10.09 years: 103 patients had a long CMN implanted, and
67 patients had a short CMN implanted. There were no demographic differences or differences in radiographic
time to healing, rates of mortality, readmission, nonunion, and need for reoperation. Univariable analysis
revealed that short CMN had lower intraoperative blood loss (111.19±83.97 mL vs 176.72±161.45 mL, p =
0.002), decreased need for transfusion (37% vs. 55 %, p = 0.022), and shorter procedures (118.67±57.87 min vs.
148.95±77.83 min, p = 0.002. Multivariable analysis revealed that short nail usage was associated with
decreased intraoperative blood loss, decreased need for transfusion, and shorter operative times.
Conclusion: Nail length does not affect healing or hospital quality outcomes in the treatment of RO hip fractures.
The use of short CMNs for these fractures did correlate with lower intraoperative blood loss, operative time, and
need for blood transfusion, with non-inferior outcomes and similar hospital quality measures when compared to
long CMNs.

Introduction proximal femur, in a superolateral to inferomedial direction along the

intertrochanteric line. However, reverse obliquity (RO) IT fractures have
Intertrochanteric (IT) hip fractures comprise over half of all fractures a primary fracture line that extends in the opposite direction with the
that occur about the hip [1]. These injuries carry with them significant fracture generally exiting laterally, distal to vastus ridge. As such, these
morbidity and mortality and require surgical intervention to provide fracture patterns are biomechanically more unstable than other IT
immediate stability and allow for early mobilization to improve out­ fractures [3]. Although RO fractures are less common than other IT
comes [2]. Commonly, the major fracture line seen in IT fractures extend fractures, with an incidence in the literature reported between 4 and 15
obliquely from the greater trochanter to the lesser trochanter of the %, these are unstable patterns that behave similar to fractures with

* Corresponding author: Joseph E. Milgram Professor of Orthopedic Surgery, Department of Orthopedic Surgery at NYU Grossman School of Medicine, Vice Chair,
Academic Affairs, Orthopedic Surgery, Chief, Division of Orthopedic Trauma, 301 E. 17th Street, 14th Floor, New York NY 10003, USA.
E-mail address: [email protected] (K.A. Egol).

https://doi.org/10.1016/j.injury.2024.111463
Accepted 25 February 2024
Available online 2 March 2024
0020-1383/© 2024 Elsevier Ltd. All rights reserved.
S.R. Konda et al. Injury 55 (2024) 111463

Fig. 1. CONSORT Diagram.

comminution or disruption of the posteromedial calcar [4–7]. consecutive hip fractures was queried for any patient >55 years old
Stable IT fracture patterns may be treated with a sliding hip screw admitted for a reverse obliquity IT hip fracture [AO/OTA 31A3.1,
(SHS) or cephalomedullary nail (CMN) based on surgeon preference and 31A3.2, and 31A3.3] between October 1, 2014 and October 1, 2021.
additional fracture characteristics [8]. However, for mechanically un­ Additional inclusion criteria were treatment with a short or long CMN
stable patterns including RO fractures, CMNs have been shown to have a and a minimum of 6-months follow-up. 1382 consecutive inter­
higher load to failure and improved mechanical stability [9,10]. CMNs, trochanteric hip fractures were treated during the study period. A total
by nature of their position within the canal may also provide more of 179 patients (13 %) were classified as sustaining a reverse obliquity
medial support to mitigate medialization of the shaft [11]. There are intertrochanteric hip fracture. Of these patients, 170 met our inclusion
several studies that compare short versus long CMNs for stable and criteria. Sixty-five patients (38 %) had complete radiographic follow-up
unstable intertrochanteric hip fracture types, however, there is a paucity to assess time to complete radiographic healing. The long CMN group
of data and little consensus reported for the preferred nail length for consisted of 103 patients (44 [42.7 %] with complete radiographic
treatment of the RO fracture pattern. follow-up) and the short CMN group consisted of 67 patients (21 [31.3
The purpose of this study is to evaluate a large series of RO hip %] with complete radiographic follow-up) (Fig. 1).
fractures and determine the effect, if any of CMN length on radiographic, All patients were treated by orthopedic trainees of various levels
clinical and hospital quality outcomes. We hypothesize that nail length under the direct supervision of orthopedic faculty at one of three hos­
will have no effect on clinical outcomes. pital sites within an academic medical center.
The electronic medical records (EMR) were reviewed for de­
Methods mographic and baseline health data including age at time of admission,
race, sex, marital status, body mass index (BMI), Charlson Comorbidity
An Institutional Review Board-approved trauma database of Index (CCI) without age adjustment, need for baseline assistive device

2
S.R. Konda et al. Injury 55 (2024) 111463

Fig. 2. A) Preoperative anteroposterior, B) traction internal rotation, and C) cross-table lateral injury radiographs of a reverse obliquity intertrochanteric hip
fracture (31A3.2).

such as a cane or walker, and pre-injury ambulatory status defined as to the system of the Orthopedic Trauma Association by 1 of 3 fellowship-
either community ambulator (patients who ambulate outside of their trained orthopedic trauma surgeons who are also study authors (SK, AG,
home), household ambulator (patients who ambulate within their KE) [19]. The amount of distal fracture extension was measured on the
household), or non-ambulatory (patients who require use of a wheel­ injury AP hip radiograph from the most prominent aspect of the lesser
chair or are only able to perform transfers). Injury and treatment in­ trochanter to the distal fracture line as outlined by Shannon et al. [20].
formation collected from the hospital electronic medical record (EMR) All patients underwent surgical repair in an expeditious manner on a
included Glasgow Coma Scale (GCS), injury mechanism categorized as: traction fracture table using closed or percutaneous reduction tech­
low energy (falls from heights less than or equal to 2 stairs) or high niques and percutaneous implant placement. Implant selection was
energy (falls from heights greater than 2 stairs, motor vehicle accidents, made by the attending surgeon based upon their preference. Anesthesia
pedestrian struck by motor vehicles), American Society of Anesthesi­ techniques varied and included: general, spinal, regional, and local
ology (ASA) score class, type of procedure performed, and abbreviated infiltration techniques. Regional anesthesia utilized the lateral femoral
injury scores for head/neck (AIS-HN), chest (AIS-C), and extremity/ cutaneous and over the hip (LOH) block as described by Deemer et al.
pelvis (AIS-EXT). All patients were risk stratified utilizing the Score for [21]. The local infiltration technique utilized the monitored anesthesia
Trauma Triage in the Geriatric and Middle Aged (STTGMA) [12–18]. care and soft-tissue infiltration with local anesthetic (MAC-STILA)
Plain radiographs including an anteroposterior (AP), cross table approach as described by Konda et al. [22,23].
lateral, and an AP traction internal rotation film, were obtained at the The surgical technique for short cephalomedullary nails was stan­
time of injury in all patients. Fracture patterns were classified according dardized per institutional protocol. Either a 125-degree 10 mm x 170

3
S.R. Konda et al. Injury 55 (2024) 111463

mm nail or 11 mm x 180 mm short nail was used based on intra­

medullary canal diameter. For all nails, the target tip-apex distance was
less than 25 mm, all lag screws were statically locked, and all nails were
statically locked distally with a 5.0 mm screw using a targeting jig. The
surgical technique for the long cephalomedullary nail was similar with a
few exceptions. Nail length varied based on intraoperative measurement
with the goal of the distal end of the nail to be distal to the superior pole
of the patella. Two static distal interlocking bolts were placed using
either a freehand “perfect circle” technique or using a distal targeting
system based on surgeon preference. Postoperatively, all patients were
prescribed full weight bearing as tolerated immediately following sur­
gical fixation and were prescribed chemical and mechanical venous
thromboembolism prophylaxis.
Recorded perioperative parameters included operative time (mi­
nutes), preoperative and postoperative hemoglobin (g/dL)/hematocrit
(%) (H/H), intraoperative blood loss (milliliters [ml]), the need for
inpatient blood transfusion (packed red blood cells [pRBC]), and the
quantity of transfusion required (number of pRBC). Operative time was
defined as time of patient entry into the operating room until time of
patient exit. Preoperative hemoglobin/hematocrit (H/H) was the first
lab value drawn during their hospitalization which was drawn in the
emergency department. Postoperative hemoglobin/hematocrit (H/H)
was the first lab drawn after their operation and was drawn within 12 h
of the completion of surgery. Intraoperative blood loss was assessed by
the anesthesiologist and surgeon based upon collection of blood in the
suction and estimated loss in the drapes and on the floor during the
operation and recorded in the operative note.
Radiographic parameters that were recorded included the tip-to-
apex distance (TAD, millimeters [mm]) and nail diameter [mm]/canal
diameter [mm] ratio (N/C ratio) and both were measured from intra­
operative fluoroscopy per instructions by Baumgaertner et al. and
George et al. respectively [24,25]. Of note, the N/C ratio was only
measured for short nails, and comparison to established thresholds
published by George et al. was performed.
Hospital quality measures and outcomes were recorded and included Fig. 3. A) Postoperative anteroposterior and B) Cross table lateral radiographs
the following: length of stay (days), postoperative ambulation distance demonstrating a short cephalomedullary nail transfixing the fracture with in­
(feet) with physical therapy prior to discharge, inpatient mortality, terval healing.
inpatient complications, and discharge disposition. Minor inpatient
complications included: acute kidney injury, urinary tract infection, and reported as a reference cost (X for total cost and Y for procedural cost
acute blood loss anemia. Major inpatient complications included: sepsis with standard deviations) per institution requirements and a desire to
or septic shock, pneumonia, acute respiratory failure, stroke, myocardial make the costs more generalizable to other currencies.
infarction, cardiac arrest and deep vein thrombosis or pulmonary Patients were divided into cohorts based on the implant used at the
embolism. Additional outcomes included 30- and 90-day readmission, time of surgery: short CMN (Fig. 2,3,4) or long CMN (Fig. 5,6,7).
mortality at 30-days and one year, fixation failure requiring a reopera­ Comparative analyses were conducted between cohorts. Mann-
tion, radiographic healing, and clinical healing. Post-operative radio­ Whitney U tests, Chi-Square Tests, Independent Sample T-tests were
graphs were reviewed to determine healing rate, time to radiographic used as appropriate. Additional analysis included multivariable linear
healing, and development of fracture nonunion. regression for the following dependent variables: need for transfusion,
A radiographically healed fracture demonstrated callus formation on intraoperative blood loss, operative time, procedural and total admis­
3 of 4 cortices on AP and lateral plain radiographs. Post operative CT sion costs. Independent variables for each regression included age, sex,
scans were only attained in the setting of lack of progressive healing BMI, CCI, ambulatory status, the STTGMA tool risk score, AO/OTA
with continued pain on ambulation. A nonunion was defined as an fracture classification, anesthesia type, and whether treatment was
absence of callus formation appreciated on 3 consecutive plain radio­ provided by a short or long CMN. Statistics were calculated with IBM
graphs performed 6 weeks apart, confirmation on CT or evidence of SPSS data software, Version 25. Significance was defined with an alpha
hardware failure and loss of reduction at the fracture site. A clinically of 0.05.
healed fracture was defined based on an electronic medical record
(EMR) review demonstrating the patient was able to ambulate or bear Results
weight on the extremity without pain or complaint at fracture site at any
outpatient visit greater than 6 months from the time of surgery, The mean follow-up for the overall cohort was 430 days. De­
regardless of radiographic appearance. Those who expired before 6 mographics for the overall cohort included an average age of 81 years,
months without evidence of radiographic healing were excluded from mean CCI of 1.24, mean ASA of 2.92, mean BMI of 25.11 kg/m2, and
analysis involving clinical healing. mean STTGMA score of 1.35 %. The majority of patients were white (68
Total hospital and procedure costs were reported for patients at one %), female (75 %), and a community ambulator (64 %). There was no
of the three hospital sites where patients were treated during this study difference in demographics including age, CCI, sex, and fracture pattern
representing 40 % of included patients. The total costs and procedural between the long and short CMN cohorts. The majority of patients sus­
costs of these hospitalizations were obtained from the hospital finance tained a 31A3.3 AO/OTA classification (57 %) injury via a low-energy
department using their cost accounting system (EPSI, NY). Costs were

4
S.R. Konda et al. Injury 55 (2024) 111463

Fig. 4. A) Postoperative anteroposterior and B) Frog Leg Lateral and C) Cross table lateral radiographs demonstrating a short cephalomedullary nail transfixing the
healed fracture.

(95 %) mechanism. (Table 1). more surgeons then preferring long nails.
Patients treated with a short CMN had a shorter operative time in Six patients required a reoperation (3 Short CMN, 3 Long CMN). The
minutes (118.67±57.87 vs. 148.95±77.83, p = 0.002), a lower intra­ 3 reoperations in the short CMN group included 2 conversion total hip
operative blood loss in milliliters (111.19±83.97 mL vs. 176.72±161.45 arthroplasties (THA) due to intra-articular penetration of the femoral
mL, p = 0.002), and a lower rate of transfusion need (37% vs. 55 %, p = head lag screw and 1 conversion THA due to a nonunion with a broken
0.022). There were no other differences seen in perioperative parame­ nail. For the reoperations and revisions of the short nail cohort, the
ters (Table 2). radiographic parameters were as follows: distal fracture extension mean
Patients treated with a short CMN had less costly admissions (X ± was 2.71±1.83 cm TAD mean was 8.60±6.77 mm, N/C ratio mean was
0.57X vs. 1.23X±0.63X, p = 0.018) and less costly procedures (Z ± 0.73±0.42 mm.
0.62Z vs. 1.21Z±0.65Z, p = 0.031). There were no other differences seen The 3 reoperations in the long CMN group includes 2 conversion
in hospital quality measures (Table 3). The average radiographic time to THA due to fracture nonunion and 1 exchange nail due to fracture
healing for our cohort of patients with complete radiographic follow-up nonunion and broken hardware. For the reoperations and revisions of
was by 233 days, with a median of by 197 days. Patients with complete the long nail cohort, the radiographic parameters were as follows: distal
radiographic follow-up experienced a similar rate of bony healing be­ fracture extension mean was 2.95± 0.69 cm, TAD mean was 17.40
tween groups (62% vs. 70 %, p = 0.491). Patients without complete ±10.02 mm.
radiographic follow-up experienced a similar rate of clinical healing Multivariable linear regression demonstrated correlation between
regardless of implant type. There were no differences seen in rates of short CMN use and a shorter operative time by 29 min (p = 0.004) and
implant failure (screw cut out or intra-articular penetration, and hard­ lower intraoperative blood loss by 60 mL (p = 0.008)(Table 5). How­
ware breakage), nonunion, or need for reoperation between cohorts ever, the correlations between short CMN use and lower procedural
(Table 4). costs (p = 0.209) and total admission costs (p = 0.081) were not found to
When stratifying by the amount of distal fracture extension as be significant when controlling for other confounding variables. Multi­
measured on AP radiographs, there was a breaking point >1.85 cm with variable binary logistic regression demonstrated a 50 % less likely need

5
S.R. Konda et al. Injury 55 (2024) 111463

Fig. 5. A) Preoperative anteroposterior, B) traction internal rotation, and C) cross-table lateral injury radiographs of a reverse obliquity intertrochanteric hip
fracture (31A3.1).

for transfusion associated with short CMN use (p = 0.029) and a 50 % Discussion
less likely need for transfusion associated with more frequent use of
general anesthesia as compared to soft-tissue infiltration with local We found minimal differences between a short or long CMN for the
anesthesia and sedation (p = 0.023) (Table 6). treatment of a reverse obliquity intertrochanteric hip fractures (AO/
A sub-analysis comparing patients with and without complete OTA 31A3.1, 31A3.2, and 31A3.3) with regards to operative parame­
radiographic follow-up demonstrated patients without complete radio­ ters, hospital quality measure, radiographic healing and clinical out­
graphic follow-up had a lower BMI and a higher rate of 1-year mortality. comes. The study cohorts possessed similar demographics and data
There were no other differences in demographics, injury characteristics, analysis demonstrated that patients who were treated with a short CMN
or outcome measures in this sub-analysis comparison (See Supplemental had shorter operative times, less intraoperative blood loss, and a lower
Tables 1 and 2). Additional sub-analyses compared those who had rate of transfusion need as compared to those treated with a long CMN.
complete follow-up and those who did not within the short nail cohort We found no other differences in perioperative parameters, hospital
(Supplemental Table 3) and within the long nail cohort (Supplemental quality measures, and outcomes including length of stay, discharge
Table 4). disposition, nonunion rate, and radiographic time to healing.
A biomechanical study by Blum et al. utilized 4th generation syn­
thetic femurs to simulate both a comminuted (31A1) and a reverse
obliquity (31A3) fracture pattern. Both long and short nails were tested

6
S.R. Konda et al. Injury 55 (2024) 111463

Fig. 6. A) Postoperative anteroposterior and B) Frog Leg Lateral radiographs demonstrating a long cephalomedullary nail transfixing the fracture with inter­
val healing.

Fig. 7. A) Postoperative anteroposterior and B) Frog Leg Lateral radiographs demonstrating a long cephalomedullary nail transfixing the healed fracture.

to assess the respective load to failure and stiffness for each fracture fixation methods in unstable IT fracture patterns such as reverse obliq­
pattern. The authors demonstrated no difference in axial load to failure uity fractures. [26] Another biomechanical study by Linhart et al. uti­
between nail lengths, suggesting both nail lengths are appropriate lized synthetic bones for initial testing, followed by human cadaver for

7
S.R. Konda et al. Injury 55 (2024) 111463

Table 1 Table 2
Demographics and Injury Details. Perioperative Parameters.
Demographics Short Long Total P-Value Perioperative Parameters Short Long Total P-Value
n (%) n (%) n (%) n (%) n (%) n (%)

N 67 103 170 N 67 103 170

Age (years, mean ± std) 82.07 ± 80.15 ± 80.91 ± 0.228 Time to Surgery (days) 1.06 ± 1.40 ± 2.51 1.27 ± 2.03 0.182
10.24 9.96 10.09 0.87
Gender 0.196 Preoperative 11.62 ± 11.70 ± 11.68 ± 0.845
Male 13 (19.40 29 (28.16 42 (24.71 Hemoglobin (g/dL) 5.73 6.00 1.73
%) %) %) Preoperative Hematocrit 35.56 ± 35.23 ± 35.34 ± 0.793
Female 54 (80.60 74 (71.84 128 (75.29 17.50 18.05 5.01
%) %) %) Intraoperative Blood 111.19 ± 176.72 ± 150.74 ± 0.002
Charlson Comorbidity 1.40 ± 1.68 1.13 ± 1.24 ± 1.60 0.271 Loss (mL) 83.97 161.45 139.45
Index 1.54 Postoperative 9.29 ± 8.91 ± 4.63 9.04 ± 1.55 0.160
ASA Score 2.97 ± 0.78 2.89 ± 2.92 ± 0.69 0.478 Hemoglobin 4.59
0.62 Postoperative 28.41 ± 26.96 ± 27.45 ± 0.110
Body Mass Index 24.13 ± 25.74 ± 25.11 ± 0.103 Hematocrit 14.01 14.00 4.78
6.27 6.24 6.28 Need for Transfusion 25 (37.31 57 (55.34 82 (48.24 0.022
STTGMA Score 1.63 ± 1.75 1.33 ± 1.45 ± 1.50 0.212 %) %) %)
1.30 Units Tranfused 0.67 ± 1.02 ± 1.15 0.88 ± 1.11 0.046
Ambulatory Status 0.125 1.02
Community Ambulator 39 (58.21 70 (67.96 109 (64.12 Anesthesia Type 0.086
%) %) %) General 39 (58.21 67 (65.05 106 (62.35
Household Ambulator 26 (38.81 26 (25.24 52 (30.59 %) %) %)
%) %) %) Spinal 22 (32.84 31 (30.10 53 (31.18
Non-Ambulatory 2 (2.99 %) 7 (6.80 %) 9 (5.29 %) %) %) %)
Assistive Device 43 (64.18 59 (57.28 102 (60.00 0.110 LOH Block 0 (0.00 %) 3 (2.91 %) 3 (1.76 %)
%) %) %) MAC STILA 6 (8.96 %) 2 (1.94 %) 8 (4.71 %)
Operative Time (minutes, 118.67 ± 148.95 ± 139.02 ± 0.002
Injury Details
mean ± std) 57.87 77.83 38.44
Glasgow Coma Scale 14.91 ± 14.91 ± 14.91 ± 0.968 Tip to Apex Distance 13.15 ± 13.44 ± 13.33 ± 0.423
0.38 0.32 0.34 (mm) 7.64 10.23 9.25
AIS Head/Neck 0.03 ± 0.06 ± 0.05 ± 0.396 TAD > 25 mm 1 (1.49 %) 8 (7.77 %) 9 (5.29 %) 0.074
0.17 0.24 0.21 AP Nail/Canal Ratio 0.76 ± – – –
AIS Chest 0.00 ± 0.01 ± 0.01 ± 0.422 0.38
0.00 0.10 0.08 Lateral Nail/Canal Ratio 0.58 ± – – –
AIS Extremity/Pelvis 2.96 ± 3.00 ± 2.98 ± 0.216 0.27
0.37 0.00 0.23
Mechanism of Injury 0.393
High Energy 2 (2.99 %) 6 (5.83 %) 8 (4.71 %) that RO fractures can be treated with either short or long cepha­
Lower Energy 65 (97.01 97 (94.17 162 (95.29 lomedullary nails [28]. In one clinical study by Parola et al. assessing
%) %) %)
AO/OTA Classification 0.109
unstable multifragmentary pertrochanteric hip fractures (AO/OTA
31A3.1 24 (35.82 35 (33.98 59 (34.71 31A2.2 and 31A2.3), the authors found no difference between short and
%) %) %) long CMN treatment groups regarding long term outcomes including
31A3.2 9 (13.43 5 (4.85 %) 14 (8.24 mortality, radiographic time to healing, nonunion and reoperation rates
%) %)
[29]. This paper however, did not address the 31A3 patterns evaluated
31A3.3 34 (50.75 63 (61.17 97 (57.06
%) %) %) in this current study. A randomized controlled trial by Shannon et al.
Distal Fracture Extension 2.3 3.6 3.2 0.069 found comparable functional outcomes and implant failure rates for
(cm, median) both short and long CMNs in the treatment of pertrochanteric hip frac­
ASA=American Society of Anesthesiology; AIS=Abbreviated Injury Score;. tures [20].
Therefore, our data aligns with and confirms previous existing
literature demonstrating that patients and surgeons may expect similar
final testing. Reverse obliquity (31A3) patterns were created and fixed
outcomes regardless of long or short CMN usage in RO fracture patterns,
with either a short or long nail in order to test implant stiffness and the
except in cases where the distal fracture extension length is greater than
amount of displacement of the fracture gap following implantation.
1.85 cm. In cases of RO fractures where the distal fracture extension
They found no difference in axial stiffness with a slight increase in
median was >1.85 cm, surgeons at our institution preferred the use of
fracture gap displacement following short nail insertion. The authors
long CMNs. Surgical planning may then be more focused on surgeon
defined a safe range for fracture extension above the short CMN locking
implant preference, or additional fracture and patient characteristics.
screw (2 centimeter proximal to the distal locking screw), indicating
Our study demonstrated a similar rate of clinical healing in our
certain RO patterns would receive equivalent biomechanical stability
cohort of patients that did not have complete radiographic follow-up. All
from both a short and long nail [27]. However, the authors recognized
170 patients had at least a minimum of 6-month follow-up (either with
the need for additional clinical studies with which to correlate these
or without radiographs), excluding those who expired in the post-
biomechanical studies. A randomized trial by Okcu et al. sought to
operative period. Study data demonstrated a similar rate of clinical
compare the performance of short and long nails for the treatment of
healing between cohorts with a comparable rate of reoperation as only 3
reverse obliquity (31A3) fractures in regard to the rates of failure
% of patients in each cohort required further intervention at a mean of
requiring reoperation, 1-year mortality, bony union, and long-term
14 months post-operatively. It is important to acknowledge that our
function. Their prospective study included 40 patients divided into
study’s mean radiographic time to healing is longer than the mean cited
short or long nail treatment groups and followed for an average of 14
in the current literature [6,30]. While our study’s average time to
months. Their study found similar failure/reoperation rates in addition
healing being just shy of 8 months may be longer than expected, this is
to comparable rates of 1-year mortality and bony union. Both treatment
due to the fact that some patients missed the 3- or 6-month follow-up
groups demonstrated similar function and mobility scores at time of
time point. As a result, their most recent radiograph demonstrating
follow-up. As a result, the preliminary data from the authors suggested

8
S.R. Konda et al. Injury 55 (2024) 111463

Table 3 Table 5
Hospital Quality Measures. Multivariable Linear Regression Analyses.
Hospital Quality Measures Short Long Total P-Value Multivariable Linear Regression Analyses
n (%) n (%) n (%)
Operative Time* Beta 95 % Confidence Interval P Value
N 67 103 170
Short Nail Usage − 29.517 − 49.243- − 9.790 0.004
Length of Stay (days, 6.79 ± 6.84 ± 4.21 6.82 ± 4.18 0.935
Intraoperative Blood Loss* Beta 95 % Confidence Interval P Value
mean ± std) 4.18
Need for ICU 12 (17.91 22 (21.36 34 (20.00 0.583 Short Nail Usage − 60.516 − 104.801- − 16.231 0.008
%) %) %) *
Major Complications 8 (11.94 %) 6 (5.83 %) 14 (8.24 %) 0.156 Nonsignificant Variables=Age; Body Masss Index; Female Gender; Charlson
Sepsis/Septic Shock 0 (0.00 %) 1 (0.97 %) 1 (0.59 %) 0.419 Comorbidity Index; Ambulatory Status; STTGMA Score; AO/OTA Classification;
Pneumonia 3 (4.48 %) 1 (0.97 %) 4 (2.35 %) 0.140 Anesthesia Type.
DVT/PE 0 (0.00 %) 0 (0.00 %) 0 (0.00 %) –
Myocardial Infarction 1 (1.49 %) 2 (1.94 %) 3 (1.76 %) 0.828
Stroke 1 (1.49 %) 1 (0.97 %) 2 (1.18 %) 0.758
Table 6
Acute Respiratory Failure 2 (2.99 %) 3 (2.91 %) 5 (2.94 %) 0.978
Cardiac Arrest 2 (2.99 %) 1 (0.97 %) 3 (1.76 %) 0.330 Multivariable Binary Logistic Regression Analyses.
Minor Complications 29 (43.28 46 (44.66 75 (44.12 0.702 Multivariable Binary Logistic Regression Analysis
%) %) %)
Urinary Tract Infection 3 (4.48 %) 6 (5.83 %) 9 (5.29 %) 0.701 Transfusion Odds Standard 95 % Confidence P
Acute Kidney Injury 3 (4.48 %) 6 (5.83 %) 9 (5.29 %) 0.701 Need* Ratio Error Interval Value
Anemia 26 (38.81 39 (37.86 65 (38.24 0.902 Short Nail Usage 0.465 0.351 0.234–0.925 0.029
%) %) %) Anesthesia 0.567 0.250 0.347–0.925 0.023
Post-Operative 33.50 ± 42.08 ± 39.26 ± 0.520 Type**
Ambulation (feet) 32.75 43.54 53.04
*
Discharge Location 0.657 Nonsignificant Variables=Age; Body Masss Index; Female Gender; Charlson
Skilled Nursing Facility 43 (64.18 65 (63.11 108 (63.53 Comorbidity Index; Ambulatory Status; STTGMA Score; AO/OTA Classification.
%) %) %) **
Anesthesia type was coded categorically with general anesthesia as the
Acute Rehabilitation 9 (13.43 %) 12 (11.65 21 (12.35 control (1) and more localized anesthesia with higher categorical numbers
Facility %) %)
(2=spinal, 3=regional, 4=MAC-STILA (Monitored Anesthesia Care and Soft-
Home 3 (4.48 %) 10 (9.71 %) 13 (7.65 %)
Tissue Infiltration with Local Anesthesia).
Home with Health Services 6 (8.96 %) 9 (8.74 %) 15 (8.82 %)
Deceased 2 (2.99 %) 0 (0.00 %) 2 (1.18 %)
Total Cost of Admission X ± 0.57X 1.23X ± 1.15X ± 0.018 healing was closer to 1 year when they came in for their routine 1 year
(USD) 0.63X 0.97X follow-up appointment, thereby skewing the overall radiographic time
Procedural Costs (USD) Z ± 0.62Z 1.21Z ± 1.13Z ± 0.031
0.65Z 1.08Z
to healing reported in this study.
This study found that patients treated with a long CMN have longer
X = Reference Cost Amount for Total Cost. operative times than those treated with a short CMN. The reasons for this
Z = Reference Cost Amount for Procedural Cost.
may include the need for intramedullary reaming, the need for percu­
taneous blocking wires/screws to minimize the risk of anterior cortical
perforation, the use of freehand distal targeting for screw placement
Table 4 with a “perfect circles” technique, and the placement of 2 distal locking
Outcomes.
bolts instead of 1 as was done with all short nails. It is well established in
Outcomes Short Long Total P-Value the literature that treatment with a long CMN for most types of IT hip
n (%) n (%) n (%)
fractures leads to a longer operative time. A retrospective study by
N 67 103 170 Parola et al. found patients who sustained a multifragmentary per­
Mortality trochanteric hip fracture and were treated with a long CMN on average
Inpatient 2 (2.99 %) 1 (0.97 %) 3 (1.76 %) 0.330 had a 23 min longer operative time than those treated with a short CMN
30-Day 4 (5.97 %) 4 (3.88 %) 8 (4.71 %) 0.530 [29]. Comparative studies by Boone et al. and Hou et al. found similar
1- Year 7 (10.45 10 (9.71 %) 17 (10.00 0.875
results as IT hip fractures treated with long CMNs often required longer
%) %)
Readmissions operative times than those treated with short CMNs despite similar
30-Day 6 (8.96 %) 8 (7.77 %) 14 (8.24 0.783 outcomes [31,32]. Data from Okcu et al. highlighted similar findings
%) with RO hip fractures treated with a long CMN requiring a longer
90-Day 6 (8.96 %) 13 (12.62 19 (11.18 0.458 operative time than those treated with a short CMN [28]. Therefore, it is
%) %)
*Time to Healing (days, 208 ± 96 244 ± 143 233 ± 149 0.272
reasonable to believe that these longer operative times are a function a
mean ± std) function of the implant itself. Our multivariable analysis confirms this as
*Time to Healing (days, 192 204 197 0.724 long CMN usage was the only variable found to be significantly associ­
median) ated with operative time when controlling for other variables.
**Nonunion 1 (4.76 %) 2 (4.55 %) 3 (1.76 %) 0.984
Our study also found that patients treated with a long CMN have a
Hardware Breakage 1 (1.49 %) 1 (0.97 %) 2 (1.18 %) 0.646
Screw Cut Out 0 (0.00 %) 0 (0.00 %) 0 (0.00 %) – larger intraoperative blood loss and higher need for postoperative
Intra-Articular Screw 2 (2.99 %) 0 (0.00 %) 2 (1.18 %) 0.078 transfusion. As demonstrated by our regression analysis for transfusion
Penetration need, these findings are likely related to differences in operative re­
Need For Reoperation 3 (4.48 %) 3 (2.91 %) 6 (3.53 %) 0.589 quirements/technique seen between short and long CMN usage. The
***Patients with Clinical 42 (91.30 53 (100.00 95 (96.94 0.870
Healing %) %) %)
need for intramedullary reaming, need for percutaneous blocking wires/
screws, and placement of 2 distal locking bolts may all contribute to the
*
Data Complete for 65 Patients. higher blood loss and need for transfusion, with the longer operative
**
% Out of 65 Patients (21 and 44).
*** time perhaps playing a role as the time to closure may be extended. It is
% Out of Remaining 105 Patients Without Complete Radiographic Data
well established in the literature that long CMN usage is associated with
that Did Not Pass Away <30 Days.
a higher rate of transfusion need as shown by multiple authors [29,33].
Again, our multivariable analysis confirms long CMN use to be

9
S.R. Konda et al. Injury 55 (2024) 111463

associated with a 2x higher need for transfusion when controlling for Declaration of competing interest
other patient variables.
Our study also found a mean N/C ratio for short CMN of 0.76. This All authors report no conflict of interest or financial disclosures.
demonstrated sufficient fill of the canal as our mean surpassed the rec­
ommended ratio (>0.74 on AP view) for minimal nail toggle suggested
in the literature by George et al. [25]. As a result, our cohort of short Funding
CMN are appropriate for generalization to other instances of short CMN
fixation in regards to N/C ratio. This study did not receive any specific grant from funding agencies in
There are several limitations to our study. First, the retrospective the public, commercial, or not-for-profit sectors.
nature of this study introduces inherent biases associated with this study
type, however, since the demographics and injury characteristics be­
Ethical Approval
tween groups were similar it makes our comparison more valid. A pri­
mary limitation was some RO fracture patients did not have complete
This retrospective chart review study involving human participants
radiographic follow-up through the point of healing. Therefore, even
was in accordance with the ethical standards of the institutional and
though all 170 patients had at least a minimum of 6-month follow-up
national research committee and with the 1964 Helsinki Declaration
(either with or without radiographs) our time to radiographic healing
and its later amendments or comparable ethical standards. The NYU
data only includes patients with complete radiographic follow-up. These
Langone Health Office of Science and Research Institutional Review
patients without complete radiographic follow-up may skew our results
Board (IRB) Langone approved this study. IRB approval was completed
as we do not have their endpoints and outcomes to include in our
before this particular study/manuscript started.
analysis. To mitigate this, we performed a sub-analysis of patients with
and without radiographic follow-up and demonstrated largely no dif­
ference in demographics, injury characteristics as well as outcome Supplementary materials
measures apart from body mass index and 1-year mortality rate. It is
possible this higher 1-year mortality rate limited the ability to get Supplementary material associated with this article can be found, in
complete radiographic follow-up as they passed away. However, upon the online version, at doi:10.1016/j.injury.2024.111463.
review of their last orthopedic appointment prior to death, all patients
were doing well without issue regarding their hip fracture. (See Sup­ References
plemental Tables 1 and 2 for these analyses). Additionally, treatment
options were not randomized thereby relying on surgeon preference [1] Brunner LC, Eshilian-Oates L, Kuo TY. Hip fractures in adults. Am Fam Phys 2003;
67:537–42.
which introduces bias. However, as there were no differences in baseline [2] Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip
demographics or injury details (notably classification) between cohorts, fractures in the United States. JAMA 2009;302:1573–9. https://doi.org/10.1001/
this bias may be minimal. While previous work by Okcu et al. utilized a jama.2009.1462.
[3] Haidukewych GJ, Israel TA, Berry DJ. Reverse obliquity fractures of the
randomized control trial to reduce these biases, their work did not intertrochanteric region of the femur. J Bone Joint Surg Am 2001;83:643–50.
include a cost analysis of the treatment cohorts [28]. Therefore, future https://doi.org/10.2106/00004623-200105000-00001.
study utilizing a randomized control trial may include cost analysis to [4] Tawari AA, Kempegowda H, Suk M, Horwitz DS. What makes an intertrochanteric
fracture unstable in 2015? Does the lateral wall play a role in the decision matrix?
better understand the value aspects associated with implant choice. J Orthop Trauma 2015;29(4):S4–9. https://doi.org/10.1097/
Another limitation is that while this study speculates on the reasons BOT.0000000000000284. Suppl.
for longer operative times associated with the long CMN cohort, we [5] Babhulkar S. Unstable trochanteric fractures: issues and avoiding pitfalls. Injury
2017;48:803–18. https://doi.org/10.1016/j.injury.2017.02.022.
cannot rule out the impact of surgeon variability and the presence of
[6] Min WK, Kim SY, Kim TK, Lee KB, Cho MR, Ha YC, et al. Proximal femoral nail for
trainees. Finally, our medical center’s location in an urban metropolitan the treatment of reverse obliquity intertrochanteric fractures compared with
center may limit the generalizability of our study findings to other gamma nail. J Trauma 2007;63:1054–60. https://doi.org/10.1097/01.
ta.0000240455.06842.a0.
populations; however, the limitation is likely minimal as our served
[7] Park SY, Yang KH, Yoo JH, Yoon HK, Park HW. The treatment of reverse obliquity
population is very diverse with regard to patient demographics and intertrochanteric fractures with the intramedullary hip nail. J Trauma 2008;65:
mechanisms of injury. 852–7. https://doi.org/10.1097/TA.0b013e31802b9559.
[8] Baumgaertner MR, Curtin SL, Lindskog DM. Intramedullary versus extramedullary
fixation for the treatment of intertrochanteric hip fractures. Clin Orthop Relat Res
Conclusion 1998:87–94.
[9] Sadowski C, Lübbeke A, Saudan M, Riand N, Stern R, Hoffmeyer P. Treatment of
When using short CMNs for reverse obliquity fractures with fracture reverse oblique and transverse intertrochanteric fractures with use of an
intramedullary nail or a 95 degrees screw-plate: a prospective, randomized study.
extension distal to the lesser trochanter of less than 1.85 cm, clinical and J Bone Joint Surg Am 2002;84:372–81.
radiographic fracture healing is equivalent to long CMNs. Short CMNs [10] Mahomed N, Harrington I, Kellam J, Maistrelli G, Hearn T, Vroemen J.
correlate with lower intraoperative blood loss, operative time, and need Biomechanical analysis of the Gamma nail and sliding hip screw. Clin Orthop Relat
Res 1994:280–8.
for transfusion, with non-inferior outcomes and similar hospital quality [11] Kregor PJ, Obremskey WT, Kreder HJ, Swiontkowski MF. Evidence-based
measures when compared to long CMNs. orthopaedic trauma working group. Unstable pertrochanteric femoral fractures
J Orthop Trauma 2005;19:63–6. https://doi.org/10.1097/00005131-200501000-
00014.
CRediT authorship contribution statement [12] Konda SR, Seymour R, Manoli A, Gales J, Karunakar MA. Carolinas trauma
network research group. development of a middle-age and geriatric trauma
Sanjit R. Konda: Writing – review & editing, Supervision, Investi­ mortality risk score a tool to guide palliative care consultations. Bull Hosp Jt Dis
(2013) 2016;74:298–305.
gation, Conceptualization. Lauren A. Merrell: Writing – review & [13] Konda SR, Lott A, Saleh H, Lyon T, Egol KA. Using trauma triage score to risk-
editing, Writing – original draft, Methodology, Formal analysis, Data stratify inpatient triage, hospital quality measures, and cost in middle-aged and
curation, Conceptualization. Garrett W. Esper: Writing – review & geriatric orthopaedic trauma patients. J Orthop Trauma 2019;33:525–30. https://
doi.org/10.1097/BOT.0000000000001561.
editing, Writing – original draft, Methodology, Investigation, Formal
[14] Konda SR, Ranson RA, Solasz SJ, Dedhia N, Lott A, Bird ML, et al. Modification of a
analysis, Data curation, Conceptualization. Kester Gibbons: Writing – validated risk stratification tool to characterize geriatric hip fracture outcomes and
original draft, Data curation. Rachel Ranson: Data curation, Concep­ optimize care in a post-COVID-19 world. J Orthop Trauma 2020;34:e317–24.
tualization. Abhishek Ganta: Writing – review & editing, Supervision. https://doi.org/10.1097/BOT.0000000000001895.
[15] Konda SR, Lott A, Egol KA. Development of a value-based algorithm for inpatient
Kenneth A. Egol: Writing – original draft, Writing – review & editing, triage of elderly hip fracture patients. J Am Acad Orthop Surg 2020;28:e566–72.
Supervision, Conceptualization. https://doi.org/10.5435/JAAOS-d-18-00400.

10
S.R. Konda et al. Injury 55 (2024) 111463

[16] Esper GW, Meltzer-Bruhn AT, Ganta A, Egol KA, Konda SR. Seasonality affects [25] George AV, Bober K, Eller EB, Hakeos WM, Hoegler J, Jawad AH, et al. Short
elderly hip fracture mortality risk during the COVID-19 pandemic. Cureus 2022;14: cephalomedullary nail toggle: a closer examination. OTA Int 2022;5:e185. https://
e26530. https://doi.org/10.7759/cureus.26530. doi.org/10.1097/OI9.0000000000000185.
[17] Meltzer-Bruhn AT, Esper GW, Herbosa CG, Ganta A, Egol KA, Konda SR. The role of [26] Blum LE, Yee MA, Mauffrey C, Goulet JA, Perdue AM, Hake ME. Comparison of
smoking and body mass index in mortality risk assessment for geriatric hip fracture reamed long and short intramedullary nail constructs in unstable intertrochanteric
patients. Cureus 2022;14:e26666. https://doi.org/10.7759/cureus.26666. femur fractures: a biomechanical study. OTA Int 2020;3:e075. https://doi.org/
[18] Esper GW, Meltzer-Bruhn AT, Ganta A, Egol KA, Konda SR. Adaptive risk modeling: 10.1097/OI9.0000000000000075.
improving risk assessment of geriatric hip fracture patients throughout their [27] Linhart C, Kistler M, Kussmaul AC, Woiczinski M, Böcker W, Ehrnthaller C.
hospitalization. Injury 2022:S0020138322008622. https://doi.org/10.1016/j. Biomechanical stability of short versus long proximal femoral nails in osteoporotic
injury.2022.11.032. subtrochanteric A3 reverse-oblique femoral fractures: a cadaveric study. Arch
[19] Meinberg EG, Agel J, Roberts CS, Karam MD, Kellam JF. Fracture and dislocation Orthop Trauma Surg 2022. https://doi.org/10.1007/s00402-022-04345-0.
classification compendium-2018. J Orthop Trauma 2018;32(1):S1–170. https:// [28] Okcu G, Ozkayin N, Okta C, Topcu I, Aktuglu K. Which implant is better for
doi.org/10.1097/BOT.0000000000001063. Suppl. treating reverse obliquity fractures of the proximal femur: a standard or long nail?
[20] Shannon SF, Yuan BJ, Cross WW, Barlow JD, Torchia ME, Holte PK, et al. Short Clin Orthop Relat Res 2013;471:2768–75. https://doi.org/10.1007/s11999-013-
versus long cephalomedullary nails for pertrochanteric hip fractures: a randomized 2948-0.
prospective study. J Orthop Trauma 2019;33:480–6. https://doi.org/10.1097/ [29] Parola R, Maseda M, Herbosa CG, Konda SR, Ganta A, Egol KA. Quality differences
BOT.0000000000001553. in multifragmentary pertrochanteric fractures [OTA 31A2.2 and 31A2.3] treated
[21] Deemer AR, Furgiuele DL, Ganta A, Leucht P, Konda S, Tejwani NC, et al. The with short and long cephalomedullary nails. Injury 2022;53:2600–4. https://doi.
lateral femoral cutaneous and over the hip (LOH) block for the surgical org/10.1016/j.injury.2022.05.036.
management of hip fractures: a safe and effective anesthetic strategy. J Orthop [30] Ozkan K, Eceviz E, Unay K, Tasyikan L, Akman B, Eren A. Treatment of reverse
Trauma 2022. https://doi.org/10.1097/BOT.0000000000002508. oblique trochanteric femoral fractures with proximal femoral nail. Int Orthop
[22] Konda SR, Ranson RA, Dedhia N, Tong Y, Saint-Cyrus E, Ganta A, et al. Monitored 2011;35:595–8. https://doi.org/10.1007/s00264-010-1002-z.
anesthesia care and soft-tissue infiltration with local anesthesia: an anesthetic [31] Boone C, Carlberg KN, Koueiter DM, Baker KC, Sadowski J, Wiater PJ, et al. Short
option for high-risk patients with hip fractures. J Orthop Trauma 2021;35:542–9. versus long intramedullary nails for treatment of intertrochanteric femur fractures
https://doi.org/10.1097/BOT.0000000000002062. (OTA 31-A1 and A2). J Orthop Trauma 2014;28:e96–100. https://doi.org/
[23] Bi AS, Fisher ND, Ganta A, Konda SR. Monitored anesthesia care and soft-tissue 10.1097/BOT.0b013e3182a7131c.
infiltration with local anesthesia for short cephalomedullary nailing in medically [32] Hou Z, Bowen TR, Irgit KS, Matzko ME, Andreychik CM, Horwitz DS, et al.
complex patients: a technique guide. Cureus 2021;13:e20624. https://doi.org/ Treatment of pertrochanteric fractures (OTA 31-A1 and A2): long versus short
10.7759/cureus.20624. cephalomedullary nailing. J Orthop Trauma 2013;27:318–24. https://doi.org/
[24] Baumgaertner MR, Curtin SL, Lindskog DM, Keggi JM. The value of the tip-apex 10.1097/BOT.0b013e31826fc11f.
distance in predicting failure of fixation of peritrochanteric fractures of the hip. [33] Dunn J, Kusnezov N, Bader J, Waterman BR, Orr J, Belmont PJ. Long versus short
J Bone Joint Surg Am 1995;77:1058–64. https://doi.org/10.2106/00004623- cephalomedullary nail for trochanteric femur fractures (OTA 31-A1, A2 and A3): a
199507000-00012. systematic review. J Orthop Traumatol 2016;17:361–7. https://doi.org/10.1007/
s10195-016-0405-z.

11