Journal of Orthopaedic Trauma Publish Ahead of Print

DOI: 10.1097/BOT.0000000000000353 Optimal Proximal Locking Screw Length in Retrograde Nails

1
2
3 Is There an Optimal Proximal Locking Screw Length in Retrograde Intramedullary
4 Femoral Nailing?
5 Can we Stop Measuring for these Screws?
6
7
8 Cory A. Collinge, MD1 John D. Koerner, MD2, Richard S. Yoon, MD3, Michael J.
9 Beltran, MD4, and Frank A. Liporace, MD3
10

D
11 1. Division of Orthopaedic Trauma
12 Harris Methodist Fort Worth Hospital
13 Fort Worth, TX, 76104

TE
14
15 2. Division of Orthopaedic Trauma
16 Department of Orthopaedic Surgery, UMDNJ – New Jersey Medical School
17 Newark, NJ 07101
18
19 3. Division of Orthopaedic Trauma
20 Department of Orthopaedic Surgery, NYU Hospital for Joint Diseases
EP
21 New York, NY 10003
22
23 4. Department of Orthopedic Surgery
24 San Antonio Military Medical Center
25 San Antonio, TX
26
27
28
C

29 Conflicts of Interest: Dr. Collinge and Dr. Liporace have received royalties for lower
30 extremity intramedullary nails from Biomet, Parsippany, NJ. No other authors have direct or
31 indirect benefits to report in the preparation or completion of this manuscript.
C

32
33 Ethical Review Board: Institutional review board approval was required and obtained prior to
34 the completion of this manuscript.
35
A

36
37 Corresponding Author:
38 Cory A Collinge, MD

39 Director of Orthopaedic Trauma, Harris Methodist Fort Worth Hospital and Staff Physician,
40 John Peter Smith Orthopedic Surgery Residency Program
41 800 5th Street, Suite 500, Fort Worth, TX 76104
42 Cell: 817-253-9392
43 Office: 817-878-5300
44 Fax: 817-878-5305
45 Email: ccollinge@msn.com

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails

Summary
The insertion of locking screws through the proximal thigh while locking retrograde femoral nails is arguably
more difficult and traumatic to local tissues than locking at other intramedullary nail sites. The purpose of this
study was to evaluate whether a “standard” screw length for proximal interlocking of retrograde nails is
possible, therefore assessing whether the act of measuring for these screws can be omitted. This paper
retrospective evaluates screw position and estimated proximal locking screw length in patients undergoing
retrograde nailing using a large radiographically measured computed tomography (CT) cohort, with validation
via a smaller clinical cohort. According to these data, it appears reasonable to skip depth gauge measurement
during AP interlocking of retrograde femoral nails and insert a standard length screw based on location
relative to the lesser trochanter. This should decrease the amount of local trauma to the patient at the locking

D
screw site, while increasing operating room efficiency by avoiding what can often become a difficult step
during the procedure.

TE
Keywords: femoral nail, retrograde, locking screw, interlocking screw, screw length, nailing, trauma
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails
INTRODUCTION
Retrograde intramedullary femoral nailing is routinely used to treat midshaft and distal-third femoral
shaft fractures. Relative indications for retrograde nailing include distal fractures, ipsilateral femoral neck or
acetabular fracture, obesity, pregnancy, ipsilateral tibial fracture, and others.1,2 Modern nailing systems rely
on proximal and distal interlocking for axial and rotational stability. The vast majority of retrograde nails
implanted are full-length nails inserted to a level at or just proximal to the lesser trochanter (LT)- this increases
the working length of the nail, prevents the formation of a stress riser in the subtrochanteric femur, and limits
injury to branches of the femoral artery and nerve adjacent to the proximal femur.3

Compared to other areas of the body, proximal interlocking of retrograde femoral nails is technically
more challenging. These screws are inserted using free-hand, “perfect circle” technique, without a radiolucent

D
targeting jig available. Although some nailing systems offer navigated targeting thereby eliminating the need
for fluoroscopy altogether, (e.g. electromagnetic technology, TRIGEN SURESHOT®, Smith Nephew, Memphis,
TN, USA), the option to lock with a lateral-medial screw, the vast majority of nails must be locked with one or
two proximal screws placed from anterior to posterior. Because the soft tissue envelope about the proximal

TE
femur is composed of taut fascia enveloping the thick quadriceps musculature, drilling for and placing these
screws can be challenging and time consuming. For this reason, some systems have gone to a more complex
locking mechanism where the screw is firmly held by thescrew driver. Further complicating this is screw
measurement. A number of depth gauges designs are available from manufacturers, but our impression is that
few were specifically designed for the typical anatomy in this area and may be better suited for use in other
locations where locking screws are used. As a result, often larger wounds and potentially more soft tissue
trauma during locking screw application are the result.
EP
The purpose of this article was to analyze the proximal femur in patients with femoral shaft fractures in
order to establish an average screw length at and adjacent to the lesser trochanter (LT). We theorized that a
standard screw length can reliably be chosen based on whether the interlocking screw is placed above,
adjacent to, or below the LT, obviating the need for routine depth gauge measurement prior to placement of
these screws.
C

ANALYSIS
Institutional review board approval was obtained from all institutions involved in this study.
C

Radiographic Computed Tomography (CT) Cohort

Data from consecutive patients treated with retrograde intramedullary femoral nails was collected at a
level 1 and level 2 trauma center. We included only those patients with a unilateral injury and a preoperative
CT scans of both proximal femorae, which was typically obtained as part of the initial trauma workup. AP
A

diameter of the proximal femur was measured at three locations on the uninjured extremity: at the LT
(centered from the superior to inferior most aspect), 1cm proximal to the superior-most extent of the LT, and
1cm inferior to the inferior extent of the LT. One of two fellowship-trained orthopedic traumatologists
independently performed all measurements at their respective institution: measures were taken from the
outermost aspect of the anterior cortex to the outer aspect of the posterior cortex along a line passing
centrally through the canal at each level. CT scanners at our institutions were calibrated daily to ensure
accuracy. The medical record of each patient was reviewed to gather age, gender, height, weight, ethnicity,
and body mass index (BMI) (Table 1).

CT scans were obtained for 348 consecutive patients (76 female and 272 male), and a total of 320
consecutive patients (61 female and 259 male) met inclusion criteria (Table 1). The mean estimated AP screw
2

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails
length determined from CT increased from distal to proximal (Tables 2 and 4). Men had a significantly larger
AP depth than women 1 cm inferior to the LT (p<0.001), otherwise there were no significant differences
among any of the other study variables.

Thirty-five patients receiving 47 screws had a pre-operative CT that allowed for measurement of the AP
thickness of the pilot hole at each level. Demographic data is presented in Table 3 and the measured lengths
of pilot holes for proximal locking are reported in Table 4. The clinical patient cohort was insufficient in
number for subgroup analysis.

Clinical Cohort
Clinical validation was assessed in a consecutive series of patients treated for femur fracture by one of

D
the authors (blinded) between Jan 2009 and Jan 2013. A retrospective analysis of measured proximal locking
screw lengths using a standard depth gauge (Biomet, Parsippany, NJ) was performed by reviewing the
operative record. Measured pilot hole length and actual inserted screw length were recorded in the operative
note as part of that surgeon’s routine. Screw position relative to the LT was determined via magnified

TE
radiograph. A horizontal line was drawn through the center of the proximal locking screw, and its vertical
position relative to the LT was measured. The electronic medical record of each patient was also reviewed to
gather age, gender, height, weight, ethnicity, and body mass index (BMI) (Table 2).

Comparing Cohorts for Agreement of Clinical Measurement (Validation)

The radiographic cohort was then assessed for agreement with the clinical cohort to establish whether
EP
it was reasonable to use CT measurements as a surrogate for a larger clinical group. Agreement between the
two methods for measuring optimal screw length was assessed by measuring difference of values from the
mean of the two measures, absolute mean differences between measurements,4,5 as well as with intraclass
correlation coefficient (ICC) 5 .

Analysis of “Optimal” Measured Screw Length with Commercially Available Screws

C

Using the clinical data, we calculated what the relative length from “perfect’ would have been using a
screw that was the “next size up” for the average pilot hole length at each of the three levels. Traditionally,
locking screws have been manufactured in 5mm increments (most manufacturers), but screws are now also
available in 2mm or 2.5mm increments by most implant manufacturers: we compared our patients’ data to all
C

three increments. The ”best” screw length was defined as the screw available equal to our measured pilot
hole length or next longer for screws available in 2.0mm and 2.5mm increments, or screw available 1mm less
and up to 4mm longer than measured pilot hole length for screws available in 5mm increments.
A

Comparing Cohorts for Agreement of Clinical Measurement

Analysis assessing agreement of the CT measurements and the clinical measurements were plotted
(Table 1) and showed a mean difference of 0.23 percentage points with the limits of agreement at -4.6 and
4.7, small enough to be confident that CT measurements are a valid substitute for measuring clinically using a
depth gauge. The ICC between the two measures of screw length was 0.81 (95% confidence interval, 0.63 to
0.91), indicating very good agreement.4,5

Analysis of “Optimal” Measured Screw Length with Commercially Available Screws

Thirty-eight of 47 (81%) patients would have received a screw within 2.0mm of the measured length
with the “next size up” rule. Three patients would have received a screw 4mm shorter than measured, and
one patient would have received a screw 6mm longer than measured. If a screw that came in 2.5mm
3

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails
increments had been used with a similar approach, 32 of 47 (69%) patients would have received screws within
2.0mm of measured. Two patients would have received a screw that was 5mm shorter than measured and 1
would have received a screw 6.5mm longer than measured. Finally, 25 of 47 (54%) of patients locked treated
with a nail that’s locking screws came in 5.0mm increments would have received a screw within 2mm of
measured. Six patients would have received a screw ≥5mm longer (two 5mm, one 6mm, one 7mm) or shorter
(two 5mm) than measured (Table 5).

Statistical Analysis
Continuous variables were summarized and reported as means with standard deviations; comparative
analysis between gender and other subgroups was analyzed via Student’s t-test. It is important to note that

D
appropriately, equal variances were not assumed when comparing gender cohorts due to unequal cohort
sizes. Frequencies and percentages were analyzed and reported for categorical data. ICC reported as standard
with higher agreement noted towards 1. Significance was set to p<0.05. All analysis was performed via SPSS
18.0 (IBM Corp. Armonk, NY).

TE
Source of Funding
The authors of this paper have received no outside funding in the preparation of this manuscript.

DISCUSSION
According to our analysis, it appears reasonable to forgo depth gauge measurement and proceed with
placement of standard screw sizes during retrograde femoral nailing. It seems reasonable to insert two
EP
proximal locking screws 35-40mm in length, perhaps on the longer side of that range for the screws just above
the lesser trochanter and shorter for those just below the lesser trochanter. Using standard screw sizes may
have an impact in decreasing operating room time and potentially surgical trauma to the proximal femur.
According to our data, using this strategy would only rarely result in a substantial mismatch between the
screw placed and the actual pilot hole that is drilled.
C

Whether the placement of a screw that is too short or too long is clinically relevant remains unknown
and is not readily addressed in the literature. The purpose of interlocking screws is to maintain axial and
rotational stability. Anecdotally, short proximal locking screws (not completely bicortical) can result in screw
back-out, thus, it is important to attempt achieve bicortical contact. An excessively long locking screw in this
C

area is unlikely to be clinically important. The most likely clinical malady would be some amount of irritation of
the quadratus femoris muscle whose insertion is broadly along the posterior aspect of the proximal femur
proximal to the level of the lesser trochanter: we are, however, unaware of any documented cases of
symptomatic proximal interlocking screws as a result of long screw placement.1,6-11 Nerve or artery injury after
A

placement of an inappropriately long interlocking screw during intramedullary nailing has been reported in
other anatomical regions,12 however to our knowledge this has not been associated with proximal locking of
retrograde femoral nails. Theoretical injury to the sciatic nerve or branches of the femoral nerve may be at risk
during AP proximal locking, but is not reported in the literature and the risk should be minimized with
predetermined screw lengths.3,13

Our study is not without its limitations. With only two trauma centers involved, despite a large study
cohort, this study alone cannot clearly capture the diversity observed around the United States and globally.
However, the purpose of the study was not to suggest concrete screw lengths for each male and female, but
to offer a general starting point to increase operative efficiency in an effort to decrease patient time under
anesthesia and perhaps the amount of operative trauma to the anterior thigh. Finally, we chose three regions
4

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails
centered at the LT from which to quantify and compared locking screw lengths. Obviously, this analysis might
be considered more “exact” by dividing the proximal femur into more than three sections (i.e. allowing for
analysis shorter segments of the bone). The major strength of this study is that it is the first attempt that we
are aware of that has tried to assess the optimal locking screw length of locking screws in this area to see if
standard lengths could be used. However, it is also important to note that our recommendations only apply
for nails placed at the measurements parameters analyzed; for nails that fall further away from the LT, these
standard screw lengths should not be implemented.

In summary, standard screw lengths for two proximal screws (35-40mm) appear to be a reasonable
option for AP locking of retrograde femoral nails, obviating the need for depth gauge measurement and the
problems inherent with that particular technique - namely prolonged operative time, mistaken measurements

D
when percutaneously done and the depth gauge not fully seated, and local soft tissue injury. Further
investigation might focus on standardization of interlocking screws in other locations of the body during
intramedullary nailing, which has the potential to decrease implant inventory, as well as operative time.

TE
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails
References
i
1. Ostrum RF, DiCicco J, Lakatos R, et al. Retrograde intramedullary nailing of femoral diaphyseal
fractures. J Orthop Trauma, 1998;12: 464-8.
2. Ricci WM, Gallagher B, Haidukewych GJ. Intramedullary nailing of femoral shaft fractures: current
concepts. J Am Acad Orthop Surg, 2009;17: 296-305.
3. Riina J, Tornetta P 3rd, Ritter C, et al. Neurologic and vascular structures at risk during anterior-
posterior locking of retrograde femoral nails. J Orthop Trauma, 1998; 12:379-81.
4. Altman DG. Assessing new methods of clinical measurement. Br J Gen Pract, 2009;59: 399-400.
5. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical
measurement. Lancet. 1986;1:307-10.
6. Acharya KN, Rao MR. Retrograde nailing for distal third femoral shaft fractures: a prospective study. J

D
Orthop Surg (Hong Kong). 2006;14: 253-8.
7. Han HS, Oh KW, Kang SB. Retrograde intramedullary nailing for periprosthetic supracondylar fractures
of the femur after total knee arthroplasty. Clin Orthop Surg. 2009;1:201-6.
8. Moed BR, Watson JT, Cramer KE, et al. Unreamed retrograde intramedullary nailing of fractures of the

TE
femoral shaft. J Orthop Trauma. 1998;12:334-42.
9. Ricci WM, Bellabarba C, Evanoff B, et al. Retrograde versus antegrade nailing of femoral shaft
fractures. J Orthop Trauma. 2001;15:161-9.
10. Sanders R, Koval KJ, DiPasquale T, et al. Retrograde reamed femoral nailing. J Orthop Trauma. 1993;7:
293-302.
11. Tornetta P 3rd, Tiburzi D. Antegrade or retrograde reamed femoral nailing. A prospective, randomised
trial. J Bone Joint Surg Br. 2000;82:652-4.
EP
12. Coupe KJ, Beaver RL. Arterial injury during retrograde femoral nailing: a case report of injury to a
branch of the profunda femoris artery. J Orthop Trauma. 2001;15:140-3.
13. Beltran MJ, Collinge CA, Patzkowski JC, et al. The safe zone for external fixator pins in the femur. J
Orthop Trauma. 2012;26:643-7.
14. Brown GA, Firoozbakhsh K, Summa CD. Potential of increased risk of neurovascular injury using
proximal interlocking screws of retrograde femoral nails in patients with acetabular fractures. J Orthop
C

Trauma. 2001;15:433-7.
15. Handolin L, Pajarinen J, Tulikoura I. Injury to the deep femoral artery during proximal locking of a distal
femoral nail--a report of 2 cases. Acta Orthop Scand. 2003;74:111-3.
16. Shuler FD, Busam M, Beimesch CF, et al. Retrograde femoral nailing: computed tomography
C

angiogram demonstrates no relative safe zone for prevention of small diameter arterial vascular injury
during proximal anteroposterior interlocking. J Trauma. 2010;69:E42-5.
17. Collinge C, Hartigan B, Lautenschlager EP. Effects of surgical errors on small fragment screw fixation. J
Orthop Trauma, 2006; 20:410-3.
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Table 1. Demographic data for entire computed tomography cohort (n=320 sides).
Demographic Mean (SD)
Parameter
Mean age, yrs 32.7 (14.5)
Gender, % male 80.9
Ethnicity
White (%) 44.4
Black (%) 39.1
Hispanic (%) 15.0
Other (%) 1.6

D
Height, in 68.5 (3.5)
Weight, lbs 186.8 (40.9)
BMI, kg/m2 28.0 (6.1)

TE
BMI = Body mass index
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Table 2. Mean measurement data for total computed tomography study cohort and by

gender.

Location Total Cohort Males Females P value

(n=320) (n=259) (n=61)

Mean AP length 1cm superior to 35.7 (5.8) 35.8 (5.7) 35.2 (6.2) 0.50
LT, mm (SD)

D
Mean AP length at LT, mm (SD) 32.4 (3.8) 32.6 (3.6) 31.4 (4.4) 0.07

Mean AP length 1cm inferior to 30.2 (3.0) 30.5 (2.9) 28.7 (3.0) <0.001*
LT, mm (SD)

TE
*Significance P<0.05, comparing gender cohorts; AP = Anteroposterior; LT = Lesser

trochanter

SD = Standard deviation
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Table 3. Demographic data for clinical validation cohort (n=47 screws).
Demographic Parameter Mean (SD)
Mean age, yrs 46.7 (24.0)
Gender, % male 65.0
Ethnicity
White (%) 63.3
Black (%) 8.3
Hispanic (%) 25.0
Other (%) 3.3
Height, in 65.5 (6.5)

D
Weight, lbs 176.8 (77.2)
BMI, kg/m2 27.0 (11.1)
BMI = Body mass index

TE
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Table 4. Comparison of clinical validation cohort’s computed tomography (CT) measurement estimating

pilot hole length, measured pilot hole length§ and inserted screw length.

Location Mean AP length Mean length of Inserted screw Mean difference in

estimating pilot hole drilled pilot hole§ length pilot hole§ vs. inserted
measured on CT screw lengths

1cm superior to 35.7 (5.8) 36.3 (1.4) 39.6 (1.9) 3.0

LT, mm (SD)

D
At LT, mm (SD) 32.4 (3.8) 31.5 (1.5) 35.3 (3.2) 3.5

1cm inferior to LT, 30.2 (3.0) 30.6 (2.4) 33.0 (2.4) 2.2

TE
mm (SD)
§
Measured with a standard depth gauge; LT = Lesser trochanter; SD = Standard deviation
EP
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
 Optimal Proximal Locking Screw Length in Retrograde Nails

1 Table 5: Comparison of measured pilot hole length and available screws for “best” screw insertion at 5mm, 2.5mm, and 2mm lengths.
2
Screw location Measured “Best” screw Difference between “Best” screw Difference between “Best” screw Difference between
length length (5mm measured and length (2.5mm measured and length (2mm measured and “best”
increments)† “best” screw increments)† “best” screw increments)† screw lengths for
lengths for 5mm lengths for 2.5mm 2.5mm increments
increments increments

D
≥1cm superior to 35.3 (2.3) 35 -0.2 (1.6) 35 -0.3 (2.3) 36 0.64 (2.3)
LT,mm (SD)

TE
At LT, mm (SD) 31.5 (1.5) 35 3.3 (1.7) 32.5 0.7 (1.7) 32 0.27 (1.7)

≥1cm inferior to 30.6 (2.4) 30 1.1 (2.4) 32.5 1.8 (2.3) 32 1.1 (2.4)
LT, mm (SD)

EP
3 †”Best” screw length defined as screw available equal to measured pilot hole length or next longer for screws available in 2.0mm and 2.5mm
4 increments, or screw available 1mm less up to 4mm longer than measured pilot hole length for screws available in 5mm increments.
5 SD = Standard deviation
6 ***Of note, ICC between measured and clinical comparison group: 0.81
7
8
C
C
A

Copyright © 2015 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.