1761

C OPYRIGHT Ó 2019 BY T HE J OURNAL OF B ONE AND J OINT S URGERY, I NCORPORATED

The Pediatric “Floating Knee” Injury

A State-of-the-Art Multicenter Study
CORTICES*

Background: Isolated femoral and tibial fractures are 2 of the top 5 causes of pediatric orthopaedic hospital admission,
yet their simultaneous ipsilateral presentation, the “floating knee” injury, remains rare. Historically, treatment consisted
of traction and cast immobilization, which resulted in prolonged periods of immobilization, lengthy hospitalizations, and
high rates of malunion. As such, previous authors have recommended fixation of at least 1 bone in the setting of a floating
knee injury. This strategy, however, has never been evaluated and the outcomes of modern treatment are unknown.
Methods: We performed a multicenter retrospective review of the records of pediatric patients with ipsilateral femoral
and tibial fractures that had been treated at 11 tertiary care level-I pediatric trauma centers from 2004 to 2014. Outcomes
and treatment strategies were assessed with standardized means.
Results: Over the study period, 130 floating knees in 129 patients met the inclusion criteria for evaluation. The average
patient age was 10.2 years, and 63.1% were male. One-third of the patients presented with open injuries, and 83.8% of
injuries were related to vehicular trauma. Simple diaphyseal fractures (OTA/AO 32-A and B femoral fractures and OTA/AO
42-A and B tibial fractures) were most common. Intramedullary fixation (rigid or flexible) was the most common treatment
strategy for femoral fractures (69.2%). Tibial fractures were treated most commonly with casting (27.7%), followed by
flexible intramedullary nailing (24.6%). The mean duration of hospitalization was 9.7 days. Outcomes were excellent in
66.6% of cases and good in 26.4% of cases.
Conclusions: Previous literature on pediatric floating knee injuries consisted of small case series that were published
prior to the introduction of flexible intramedullary nailing to North America. This multicenter study of a large cohort
demonstrates a change in practice pattern from a largely nonoperative treatment strategy to operative fixation of at least
the femoral fracture. In the present study, this approach led to good or excellent results in 93.1% of cases and was
associated with a short duration of hospitalization.
Level of Evidence: Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

I
solated femoral and tibial fractures are 2 of the top 5 causes Pediatric fractures have distinct patterns and character-
of pediatric orthopaedic hospital admission, yet their istics in comparison with adult fractures. Associated injuries,
simultaneous ipsilateral presentation, the “floating knee” treatment priorities, and prognostic implications are different
injury, remains rare1-5. Historically, these injuries were treated in the pediatric population. While there have been several
with casting, traction, and limited internal fixation2. In recent studies on the treatment of the adult floating knee, there have
years, the advent and popularization of flexible intramedullary been only a few small case series on pediatric floating knee
nailing and minimally invasive stabilization has rendered these injuries1-4,6-8. The majority of the injuries in those studies were
injuries more likely to be treated operatively than in the past. treated nonoperatively and had poor radiographic and clinical
*Members of the Children’s Orthopedic Trauma and Infection Consortium for Evidence-based Studies (CORTICES) include Jason Anari, MD, The
Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; Benjamin Shore, MD, MPH, FRCSC, Boston Children’s Hospital, Boston, Massachusetts;
Jaime Rice Denning, MD, Cincinnati Children’s Hospital, Cincinnati, Ohio; Ying Li, MD, C.S. Mott Children’s Hospital, Michigan Medicine, Ann Arbor,
Michigan; David Spence, MD, Le Bonheur Children’s Hospital, Memphis, Tennessee; Anthony Riccio, MD, Texas Scottish Rite Hospital for Children,
Dallas, Texas; Jason Stoneback, MD, Children’s Hospital Colorado, Aurora, Colorado; Joseph Janicki, MD, Lurie Children’s Hospital, Chicago, Illinois;
Jaclyn Hill, MD, Scott Rosenfeld, MD, Texas Children’s Hospital, Houston, Texas; Megan Johnson, MD, Jon Schoenecker, MD, PhD, Monroe Carell
Children’s Hospital at Vanderbilt University, Nashville, Tennessee; Brian Brighton, MD, MPH, Carolinas Health Care System, Charlotte, North Carolina;
and Keith Baldwin, MD, MSPT, MPH, The Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania.

Disclosure: The author indicated that no external funding was received for any aspect of this work. On the Disclosure of Potential Conflicts of Interest
forms, which are provided with the online version of the article, the author checked “yes” to indicate that the author had a relevant financial relationship in
the biomedical arena outside the submitted work and “yes” to indicate that the author had other relationships or activities that could be perceived to
influence, or have the potential to influence, what was written in this work (http://links.lww.com/JBJS/F433).

J Bone Joint Surg Am. 2019;101:1761-7 d http://dx.doi.org/10.2106/JBJS.18.01446

 1762
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d
outcomes, leading many authors to recommend operative
fixation1,3. We are not aware of any recent studies that have
investigated the outcomes of surgical treatment of the pediatric
floating knee.
The primary goal of the present study was to examine a
large cohort of pediatric patients with a floating knee in order
to describe the mechanism of injury, fracture characteristics,
and fracture pattern. Secondary goals were to contrast treat-
ment methods, radiographic and clinical outcomes, and
complications with those in previously published reports on
pediatric and adult floating knee injuries.
Materials and Methods
A fter institutional review board approval and a Data Use
Agreement were obtained from the 11 institutions of the
Children’s Orthopedic Trauma and Infection Consortium for
Evidence-based Studies (CORTICES) (https://cortices.org/)
involved in this study (all level-I tertiary care pediatric trauma
centers), a data collection sheet and data dictionary were
distributed to the primary investigator at each site (see
Appendix). Patients 0 to 18 years of age with ipsilateral
femoral and tibial fractures that had been sustained during the
same event between January 1, 2004, and December 31, 2014,
were identified at each institution. Exclusion criteria were
buckle fractures of both the femur and tibia, sustained pen-
etrating trauma, and incomplete clinical and radiographic
data.
Patient charts and radiographs were retrospectively re-
viewed. Fracture union was defined as bridging callus on 3 of 4
cortices on the anteroposterior and lateral radiographs and was
assessed separately for tibial and femoral fractures9. We used
the Karlström criteria to determine the clinical outcome of
treatment (Fig. 1)10. The Letts classification system3 was used to
classify the floating knee injury, and the OTA/AO classification
system was used to classify the pattern of each fracture11. Simple
Fig. 1
10
Karlström criteria . For the purpose of rating, the criterion that was lowest was considered the rating of overall outcome. (Reproduced from: Karlström G,
Olerud S. Ipsilateral fracture of the femur and tibia. J Bone Joint Surg Am. 1977 Mar;59[2]:240-3.)
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
diaphyseal femoral fractures were classified as OTA/AO 32-A1-
3 and 32-B1-2. Simple diaphyseal tibial fractures were classified
as OTA/AO 42-A2 and 42-B1-3. The Gustilo and Anderson
classification system was used to describe open fractures12.
Raw proportions and measures of central tendency were
calculated. Historical data were obtained from a previous
systematic review of pediatric floating knee injuries, and for
comparison, data on injuries in adults were tabulated from a
similar multicenter study from the adult literature. For
dichotomous variables, we used the Fisher exact test or the
chi-square test. For continuous variables with non-normal
distributions, we used the Mann-Whitney U test or the
Kruskal-Wallis test. For analyses with normal distributions,
we used the Student t test. To determine factors involved in
the outcome, we used binary logistic regression with the
backward likelihood ratio method with criteria of 0.1 signif-
icance on univariate analysis to enter the model. We used
SPSS software (version 25; IBM) to calculate all statistics.
Results
Fracture and Surgical Characteristics
e identified 130 floating knees in 129 pediatric patients.
W The average age (and standard deviation) was 10.2 ± 4.7
years. The sample was 36.9% female and 63.1% male. The
racial breakdown was 68.5% white, 13.1% black, 8.5% Latino,
and 9.9% other races or unknown. A majority of the patients
were injured as the result of impact with a motor vehicle (113
of 129 patients; 87.6%). The most common mechanisms of
injury were being a passenger in a motor-vehicle collision (58
of 129 patients; 45.0%), being struck by a vehicle (43 of 129
patients; 33.3%), and being involved in an all-terrain vehicle
(ATV) accident (12 of 129 patients; 9.3%). The average dura-
tion of follow-up was 57.4 ± 55.2 weeks.
One-third of patients (33.3%; 43 of 129) had at least
1 open fracture. Open fractures were significantly more
 1763
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d

common in the tibia than the femur (26.1% versus 10.8%; p <
0.01). Femoral fractures were more likely than tibial fractures
to be classified as type 3 (57.1% versus 41.2%), but this dif-
ference was not significant (p = 0.216). Articular involvement
occurred in only 6 femora (4.6%) and 13 tibiae (10%). The
majority of femoral fractures (104 fractures; 80%) were
diaphyseal (OTA/AO 32); among these 104 fractures, simple
patterns were most common (with 67 fractures being classified
as type A and 31 being classified as type B) and comminuted
patterns were uncommon (with 6 fractures being classified as
type C). Most of the tibial fractures (65.4%; 85 of 130 fractures)
were diaphyseal (OTA/AO 42); among these fractures, simple
patterns predominated (with 58 fractures being classified as
type A and 21 being classified as type B fractures) and complex
patterns were rare (with 6 fractures being classified as type C).
A description of the floating knee injuries with use of the Letts
classification system is shown in Figure 2, and the anatomical
locations of injuries are shown in Figure 3.
Surgical intervention was the most common treatment
strategy for the femoral fractures (91.5%; 119 of 130). The
most common fixation method was intramedullary nailing;
flexible intramedullary nails were used in 37.7% of femora,
followed by rigid nails in 31.5%. Only 8.5% of patients were
treated nonoperatively with cast immobilization. The average
age of the patients in the casting subgroup was 3.4 years, and
only 1 patient in that group was older than 6 years. In contrast,
tibial fractures were significantly more likely than femoral
fractures to be treated with a cast (p < 0.001). More than a
quarter of tibiae (27.7%) underwent cast immobilization. The
next most common treatment strategy for tibial fractures was
flexible intramedullary nailing (24.6%). A mix of other fixation
methods were utilized for the remaining tibial fractures. The
mean duration of hospitalization was 9.7 ± 11.7 days.
Clinical Outcomes
According to the Karlström criteria, 66.6% of patients had an
excellent outcome at the time of the latest follow-up, 26.4% had
a good outcome, 3.5% had an acceptable outcome, and 3.5%
had a poor outcome. The majority of patients with available
data (69.4%; 75 of 108) regained full range of motion or were
limited by <10° (19.4%; 21 of 108) at the time of final follow-
up; the remaining patients (11.1%; 12 of 108) were limited by
>10°. A majority of patients (86.6%; 84 of 97) reported normal
walking ability, and the remaining patients reported being
distance-limited; only 1 patient was unable to walk at the time
of the latest follow-up. Most patients (85.4%; 35 of 41) who
had participated in sports prior to the injury reported that they
were able to return to the same sport.

Fig. 2
Diagram illustrating the Letts classifications of the floating knee injuries in the present study. The percent of each type of fracture in the sample is shown.
(Note: One patient was missing data in terms of the Letts classification. The mechanism of injury was motor vehicle collision, but no information was
available about whether the injuries were open or closed. Thus, the numbers of fractures shown in parentheses add up to only 129 fractures, but the
percentages are based on all 130 fractures.)
 1764
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d
Fig. 3
Diagram illustrating the numbers of fractures in each anatomical region.
The femoral fractures united by an average of 15.1 ± 11.7
weeks, whereas the tibial fractures united by an average of 16.7
± 19.3 weeks (p = 0.425). All femoral fractures healed, whereas
2 tibial fractures resulted in nonunion. There were 12 cases of
malunion (9.2%), most commonly observed in association
with Letts type-D injuries (58.3% of malunions, p = 0.01). In
addition, there were 8 cases of limb-length discrepancy of >1
cm (6.2%), 3 cases of growth arrest (2.3%), 4 cases of delayed
union or nonunion (3.1%), 13 cases of infection or wound
complication requiring either medical or surgical intervention
(10%), 3 cases of compartment syndrome (2.3%), and 2 cases
of nerve palsy (1.5%). Of the 13 patients who had an infection
or wound complication requiring treatment, 8 (61.5%) had an
open fracture. No specific fracture-management strategy was
associated with a higher risk of developing one of the com-
plications. Symptomatic hardware was reported in 6 patients
who underwent operative treatment. Four of the patients with
symptomatic hardware had a rigid nail.
Predictors of Outcomes
Patients who experienced excellent outcomes were slightly
younger (average age, 9.0 versus 9.8 years) and lighter (average
body mass index, 14.3 versus 15.6 kg/m2), but these differences
were not significant (p = 0.482 and 0.864, respectively). Among
patients with simple diaphyseal femoral fractures, 73.8% (48 of
65) had excellent outcomes and, of these, 62.5% (30 of 48) were
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
managed with intramedullary fixation. In contrast, among
patients with more complex femoral fractures, only 16.7% (1 of
6) (p = 0.03) experienced excellent outcomes with intramed-
ullary nailing. Notably, there was no significant difference in
the rate of good or excellent outcomes associated with intra-
medullary nailing of simple diaphyseal femoral fractures as
compared with other methods (p = 0.999), although intra-
medullary nailing was the most frequently employed treatment
method (81.3%; 78 of 96 simple fractures). Although outcomes
were less predictable for more complex fracture patterns,
alternative fixation strategies such as open reduction and plate
fixation, closed reduction and percutaneous pinning, and
external fixation trended toward better outcomes when com-
pared with intramedullary nailing (rate of excellent results,
56.3% [9 of 16] with other methods, compared with 16.7% [1
of 6] with intramedullary nails), although the result was not
significant (p = 0.311). Outcomes of casting for femoral frac-
tures were generally excellent (71.4%; 5 of 7), although cast-
related problems were common, with 27.7% of patients either
requiring a cast change or developing skin irritation under the
cast. Casting resulted in malunion in 1 case (14.3%; 1 of 7).
No significant differences were identified in outcomes
between various fixation strategies for tibial fractures (p =
0.277). Simple patterns did not demonstrate a significant dif-
ference from more complex patterns in terms of the rate of
excellent outcomes (68.5% for simple fractures versus 63.6%
 1765
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d
for complex fractures; p = 0.639). Casting was employed for the
treatment of 27.6% of tibial fractures, and only 1 fracture
required a cast change.
On logistic regression analysis, after controlling for
femoral and tibial fracture patterns, articular involvement, type
of fracture (closed or open), age, and body mass index, only the
femoral fracture pattern was an independent predictor of an
excellent outcome (p = 0.008).
Discussion
ediatric floating knee is a rare injury. The literature on
P pediatric floating knee injuries is sparse and mostly dated.
The classic work by Letts et al. was published >3 decades ago
and was based on 15 cases3. The goal of the present multicenter
study was to provide a more current perspective on mecha-
nisms of injury, treatment methods, and clinical and radio-
graphic outcomes and to compare our findings with those in
previous reports on pediatric and adult patients. Table I details
demographic characteristics of the patients in 3 cohorts: the
current study, a systematic review of past studies of floating
knee injuries in children, and a multicenter study of floating
knee injuries in adults. To our knowledge, our study presents
data on the largest cohort of pediatric patients with a floating
knee that is available in the literature.
Pediatric floating knees are the result of a high-energy
mechanism of injury. Motor-vehicle accidents, including
automobile-pedestrian and ATV accidents, have been shown
to account for the majority of pediatric and adult cases7 (Table II).
In our series, nearly 9 of 10 children sustained this injury as
the result of a motor-vehicle accident (either as a pedestrian
or as a passenger). One notable difference compared with
prior studies is that 12.4% of the cases in our study involved
children who had been injured on a motorcycle or an ATV.
Bohn previously reported on children who had sustained this
injury after a motorcycle injury13. Floating knee injuries that
resulted from an ATV accident in the present study may reflect
a similar mechanism of injury. Similar to other authors1-4, we
found that diaphyseal fractures were the most common
fracture pattern and that intramedullary fixation was the most
common treatment strategy. Open injuries were found in
approximately one-third of patients in all 3 cohorts. Because
of the high-energy mechanism of injury, clinicians should
perform a careful examination and scrutinize radiographs for
signs of air so that open fractures are not missed initially. Over
TABLE I Data on Current Study, Historical Systematic Review, and Previous Adult Multicenter Study
Characteristic Pediatric (Modern)*
Dates 2004-2014
Source Multicenter series
Age† (yr) 10.2
No. of patients 129 (130 fractures)
*Present study. †The values are given as the mean.
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
60% of the patients in the present study who had an infection
had an open injury.
Letts et al. noted high rates of malunion (33%) and
nonunion (13%) in association with traction when applied to
pediatric floating knee injuries3. Those authors made the rec-
ommendation to fix at least 1 of the 2 fractures rigidly. This was
a bold recommendation at the time, as the standard of care for
individual tibial and femoral fractures was nonoperative
treatment14. The report by Ligier et al. on elastic stable intra-
medullary nailing of pediatric femoral shaft fractures15 was not
published until 2 years after the report by Letts et al. Yue et al.
also noted a high rate of malunion (25%) in association with
nonoperatively treated pediatric floating knees and a higher
proportion of good and excellent outcomes in association with
operative treatment1. Those authors recommended fixation of
one and preferably both fractures. Although traction histori-
cally has been a mainstay of pediatric femoral fracture treat-
ment, we did not have a single case that was treated definitively
with traction (with or without delayed spica casting) in our
series.
The majority (91.5%) of femoral fractures in our cohort
were treated operatively, and, given the excellent outcomes
associated with primarily operative treatment and the 30-day-
shorter average duration of hospitalization compared with the
findings in a recent systematic review, we recommend operative
treatment of the femoral fracture in cases of floating knee
injury in children >4 years of age2. In terms of the tibial frac-
ture, nearly one-third of our patients were managed with a cast,
and we did not note an increased malunion rate. Therefore, we
believe that select tibial fractures in floating knee injuries may
be treated successfully with a cast, although monitoring for the
development of compartment syndrome in the first 48 hours is
advised16.
Adult studies in the late 1970s and early 1980s demon-
strated the benefits of operative treatment of ipsilateral femoral
and tibial fractures10,17,18. Today, adult floating knee injuries are
treated operatively, and, in spite of a higher rate of articular
fractures and increased severity of injury, over half of patients
have good or excellent results7. The more aggressive surgical
approach, particularly with respect to the femur, resulted in a
higher rate of good or excellent results in the current study as
compared with pediatric historical controls (92.6% vs. 82.6%)
(Table III). Because of the nature of the previous studies, it is
somewhat difficult to state with any confidence if this 10%
2 7
Pediatric (Historical) Adult
1975-2003 2000-2006
Systematic review Multicenter series
9.3 31
97 172
 1766
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d

TABLE II Fracture and Fixation Characteristics of Current Study, Historical Systematic Review, and Previous Adult Multicenter Study
2 7
Characteristic Pediatric (Current) Pediatric (Historical) Adult
1,3,4,13
Most common mechanism of injury Motor vehicle-related (87.6%) Motor vehicle-related (85.4%) Motor vehicle-related
(90%)
Most common fracture pattern, femur Simple diaphyseal (73.8%) Diaphyseal (81.3%) Diaphyseal (83.1%)
Most common fracture pattern, tibia Simple diaphyseal (56.9%) Diaphyseal (77.5%) Diaphyseal (79.7%)
Most common femoral fixation Intramedullary (elastic stable Nonoperative (traction or cast) Intramedullary (71.5%)
intramedullary nail [ESIN] (66.3%)
or reamed) (69.2%)
Most common tibial fixation Intramedullary (ESIN) Nonoperative (cast) 67.4% Intramedullary 54.1%
(24.6%) or casting (27.7%)
Percent open 33.1% 42.3% 38.0% femur,
57% tibia

difference is clinically meaningful. However, our results cor-
roborate the recommendations made by Letts et al. and Yue
et al. decades ago, which, until now, were not supported by any
data1,3. Interestingly, we found that fractures other than simple
diaphyseal femoral fractures did not fare as well with intra-
medullary fixation. Because of sample heterogeneity, it is dif-
ficult to ascertain the reason for this finding. Other authors
have noted that intra-articular fractures in adults have worse
clinical outcomes than diaphyseal fractures7,19. Intra-articular
involvement combined with length instability of many complex
fracture patterns render these injuries less amenable to flexible
nailing and could account for some of the observed difference
in outcomes.
Feron et al. noted that distal-third fractures, comminuted
fractures, open fractures, infection, or vascular injury por-
tended a worse outcome7. The fact that all of these injuries are
less common in children may account for some of the observed
outcome differences between children and adults. Additionally,
in contrast to children, adults experience higher rates of non-
union and revision surgery, local wound complications, and
systemic complications19,20. No serious systemic complications
were recorded in the present study. This finding is consistent
with those in the studies by Al-Mahdi et al. and Hedequist et al.,
which showed that systemic complications, such as systemic
inflammatory response syndrome, are rare in children even in
the face of serious injury21,22.
The present study had some limitations. Because of the
retrospective nature of the study, sources of potential detection
and inclusion bias and confounding, including the potential for
misclassification, are possible. We attempted to minimize
heterogeneity in reporting through the use of a data dictionary
and rigorous definitions for all of our major outcomes. Addi-
tionally, follow-up and measurements were not standardized
across centers. Another limitation is that the average duration
of follow-up in the present study was just over 1 year. Angular
deformities or limb-length differences that would be identified
and managed over a longer period of time would not be de-
tected. As such, our malunion rate is likely an underestimation.
We still noted a 9.2% rate of malunion, which was more
common in Letts type-D fractures. These weaknesses not-
withstanding, we believe that this is the largest study to date of
this rare injury in children.
In conclusion, this multicenter study of a modern cohort
of pediatric patients with floating knees demonstrates an

TABLE III Patient Outcomes in Current Study, Historical Systematic Review, and Previous Adult Study
2 7
Characteristic Pediatric (Current) Pediatric (Historical) Adult

Length of stay* (d) 9.7 ± 11 39.8† 30

Time to femoral union* (wk) 15.1
Time to tibial union* (wk) 16.7
Delayed union or nonunion 3.1%
Common complications Malunion (9.2%), wound
complications (10%)
Outcome 93% good or excellent
Not reported Not reported
Not reported Not reported
7.2% 21%
Limb-length discrepancy Systemic complications (14%),
(33%), malunion (21%), nonunion (21%), revision
wound complications (9%) surgery (23%),
local complications (33%)
83% good or excellent‡ 53% good or excellent

*The values are given as the mean, with or without the standard deviation. †Based on frequency weighting. ‡Based on the data from only 2 of the
1,4 2
studies that were included the systematic analysis .
 1767
TH E JO U R NA L O F B O N E & JO I N T SU RG E RY J B J S . O RG
d
V O LU M E 101 -A N U M B E R 19 O C T O B E R 2, 2 019
d d
evolution in our understanding of the mechanism of injury and
treatment. We identified an increased incidence of ATV and
motor-vehicle accidents, highlighting the need for safety
awareness. Although it is difficult to determine whether oper-
ative fixation leads to clinically important improvement in
outcomes over techniques of the past, this study demonstrates
that current fixation strategies result in excellent outcomes, a
low rate of complications, and a shorter length of hospital stay.
Given the recommendations of previous authors and the find-
ings the present study, we recommend fixation of the femoral
fracture in patients >4 years of age and judicious use of internal
fixation of the tibial fracture. Surgeons should have a high index
of suspicion regarding open injuries as approximately one-third
of floating knee injuries involve an open fracture. Patients should
be followed closely and an orthoroentgenogram should be
considered at the time of follow-up to identify limb-length
References
1. Yue JJ, Churchill RS, Cooperman DR, Yasko AW, Wilber JH, Thompson GH. The
floating knee in the pediatric patient. Nonoperative versus operative stabilization.
Clin Orthop Relat Res. 2000 Jul;(376):124-36.
2. Anari JB, Neuwirth AL, Horn BD, Baldwin KD. Ipsilateral femur and tibia fractures
in pediatric patients: a systematic review. World J Orthop. 2017 Aug 18;8(8):
638-43.
3. Letts M, Vincent N, Gouw G. The “floating knee” in children. J Bone Joint Surg Br.
1986 May;68(3):442-6.
4. Arslan H, Kapukaya A, Kesemenli C, Subaşi M, Kayikçi C. Floating knee in chil-
dren. J Pediatr Orthop. 2003 Jul-Aug;23(4):458-63.
5. Nakaniida A, Sakuraba K, Hurwitz EL. Pediatric orthopaedic injuries requiring hos-
pitalization: epidemiology and economics. J Orthop Trauma. 2014 Mar;28(3):167-72.
6. Rollo G, Falzarano G, Ronga M, Bisaccia M, Grubor P, Erasmo R, Rocca G, Tomé-
Bermejo F, Gómez-Garrido D, Pichierri P, Rinonapoli G, Meccariello L. Challenges in
the management of floating knee injuries: results of treatment and outcomes of 224
consecutive cases in 10 years. Injury. 2019 Feb;50(2):453-61. Epub 2018 Dec 11.
7. Feron JM, Bonnevialle P, Pietu G, Jacquot F. Traumatic floating knee: a review of a
multi-centric series of 172 cases in adult. Open Orthop J. 2015 Jul 31;M11(Suppl 1):
356-60.
8. Nouraei MH, Hosseini A, Zarezadeh A, Zahiri M. Floating knee injuries: results of
treatment and outcomes. J Res Med Sci. 2013 Dec;18(12):1087-91.
9. Whelan DB, Bhandari M, Stephen D, Kreder H, McKee MD, Zdero R, Schemitsch
EH. Development of the radiographic union score for tibial fractures for the
assessment of tibial fracture healing after intramedullary fixation. J Trauma. 2010
Mar;68(3):629-32.
10. Karlström G, Olerud S. Ipsilateral fracture of the femur and tibia. J Bone Joint
Surg Am. 1977 Mar;59(2):240-3.
11. Meinberg EG, Agel J, Roberts CS, Karam MD, Kellam JF. Fracture and dislocation
classification compendium-2018. J Orthop Trauma. 2018 Jan;32(Suppl 1):S1-10.
T H E P E D I AT R I C “ F L O AT I N G K N E E ” I N J U RY
inequality or coronal plane deformity, particularly in cases of
Letts type-D injuries.
Appendix
Supporting material provided by the authors is posted
with the online version of this article as a data supplement
at jbjs.org (http://links.lww.com/JBJS/F434). n
NOTE: The authors acknowledge the assistance of Divya Talwar, PhD, with data coordination and
data use agreements, Stuart Almond with figure design, Aubrie Ashie, DO, with data accumulation
and cleaning, Marilyn Elliott, BA, with data accumulation and cleaning, and Shivani Gohel, BS, with
manuscript preparation.
Email address for K. Baldwin: baldwink@email.chop.edu
ORCID iD for K. Baldwin: 0000-0002-2333-9061
12. Gustilo RB, Anderson JT. Prevention of infection in the treatment of one thou-
sand and twenty-five open fractures of long bones: retrospective and prospective
analyses. J Bone Joint Surg Am. 1976 Jun;58(4):453-8.
13. Bohn WW, Durbin RA. Ipsilateral fractures of the femur and tibia in children and
adolescents. J Bone Joint Surg Am. 1991 Mar;73(3):429-39.
14. Tachdijian MO. Pediatric orthopedics. 2nd ed, vol 2. Saunders; 1990.
15. Ligier JN, Metaizeau JP, Prévot J, Lascombes P. Elastic stable intramedullary
nailing of femoral shaft fractures in children. J Bone Joint Surg Br. 1988 Jan;70(1):
74-7.
16. Flynn JM, Bashyal RK, Yeger-McKeever M, Garner MR, Launay F, Sponseller PD.
Acute traumatic compartment syndrome of the leg in children: diagnosis and out-
come. J Bone Joint Surg Am. 2011 May 18;93(10):937-41.
17. Fraser RD, Hunter GA, Waddell JP. Ipsilateral fracture of the femur and tibia.
J Bone Joint Surg Br. 1978 Nov;60-B(4):510-5.
18. Veith RG, Winquist RA, Hansen ST Jr. Ipsilateral fractures of the femur and tibia.
A report of fifty-seven consecutive cases. J Bone Joint Surg Am. 1984 Sep;66(7):
991-1002.
19. Rethnam U, Yesupalan RS, Nair R. The floating knee: epidemiology, prognostic
indicators & outcome following surgical management. J Trauma Manag Outcomes.
2007 Nov 26;1(1):2.
20. Anastopoulos G, Assimakopoulos A, Exarchou E, Pantazopoulos T. Ipsilateral
fractures of the femur and tibia. Injury. 1992;23(7):439-41.
21. Al-Mahdi W, Ibrahim MM, Spiegel DA, Arkader A, Nance M, Baldwin K. Is sys-
temic inflammatory response syndrome relevant to pulmonary complications and
mortality in multiply injured children? J Pediatr Orthop. 2017 Oct 9. [Epub ahead of
print].
22. Hedequist D, Starr AJ, Wilson P, Walker J. Early versus delayed stabilization of
pediatric femur fractures: analysis of 387 patients. J Orthop Trauma. 1999 Sep-Oct;
13(7):490-3.