One of the most important goals in the management of severe
open injury of the lower limb is to obtain adequate soft-tissue
coverage. Soft-tissue-coverage procedures are performed to provide
a closed wound, to promote revascularization of injured bone and
soft tissue, and to prevent late infection and nonunion that may
occur secondary to persistent bone ischemia8,12,13.
The options for coverage of traumatic soft-tissue defects about
the tibia that cannot be managed by direct closure include rotational
flaps and free flaps. Rotational flaps for coverage of such soft-tissue
defects include gastrocnemius and soleus muscle flaps, myocutaneous
flaps, fasciocutaneous flaps, cross-leg flaps, and variations thereof4,16,21,22. The common denominator
of these procedures is that presumed healthy tissue proximate to
the zone of injury is rotated on a vascular pedicle to provide coverage.
Conversely, free flaps typically are harvested from an area remote
to the zone of injury, and their vascular supply is provided by
means of surgically constructed microvascular anastomoses.
The type of flap used for coverage of a soft-tissue defect generally
is chosen on the basis of anatomical considerations, specifically
the location of the defect on the leg, the size of the defect, and
the availability of local tissues for coverage. Previous authors
have recommended gastrocnemius rotational flaps for defects over
the proximal one-third of the tibia, soleus rotational flaps for
defects over the middle third, and free flaps for defects over the distal
third7,20,22. Others have postulated
that the timing of coverage is more important than the type of flap
and that flap coverage should be performed less than seven days
after the injury to decrease the prevalence of complications such
as osteomyelitis and flap failure8,9.
These studies have been limited by small numbers of patients and
by failure to control for patient and injury-related variables,
such as comorbidities and the severity of the injury.
We reviewed data from a prospective, multicenter study of severe
injuries of the lower extremity to assess the short-term complications
that occur after flaps are used to treat traumatic soft-tissue defects
about the tibia. We sought to identify the patient, injury, or treatment
characteristics that were significantly associated with the likelihood of
a wound complication after a soft-tissue-coverage procedure. The
purpose of the present study was to analyze whether the type of
flap (rotational or free) applied to a traumatic soft-tissue defect about
the tibia resulted in significantly different rates of wound complications
after taking into account other factors such as the patient's overall condition,
the injury sustained, and the treatment rendered.
Data for this study were collected as part of a larger study,
the Lower Extremity Assessment Project (LEAP), which was designed
to compare the clinical and long-term functional outcomes of limb
reconstruction with those of amputation in patients who sustained
high-energy trauma of the lower extremity. The goal of LEAP was
to identify clinical and patient-related predictors of successful
treatment14.
Study Population
Patients between the ages of sixteen and sixty-nine years who
were admitted to one of eight level-I trauma centers between March
15, 1994, and June 30, 1997, for treatment of a high-energy trauma
of the lower extremity were eligible for the study. High-energy
injuries of the lower extremity included: (1) a traumatic amputation,
(2) a Gustilo-Anderson type-III tibial fracture11,
(3) a dysvascular limb (such as after a knee dislocation, a closed
tibial fracture, or a penetrating wound with vascular injury requiring
repair), (4) a major soft-tissue injury of the tibia (degloving
or a severe crush or avulsion injury), and (5) a severe ankle and
foot injury (such as a Gustilo-Anderson type-IIIB ankle or pilon
fracture as well as a severe open hindfoot or midfoot injury)14.
The criteria for exclusion from LEAP included: (1) an age of
less than sixteen years or more than sixty-nine years, (2) a documented
psychiatric disorder, (3) an inability to speak English or Spanish, (4)
an associated moderate-to-severe injury of the central nervous system,
(5) a third-degree burn measuring more than one handbreadth on the
injured leg, (6) a previous leg amputation or an inability to walk
before the injury, (7) primary treatment that was received before
admission to the trauma center, and (8) the inability of the patient
to return for follow-up evaluations because he or she lived too
far away from the treatment center.
Each of the 601 patients in the study had been or was being followed
prospectively at three, six, twelve, and twenty-four months after
the injury. For each follow-up assessment, the patient was asked
to return to the hospital and undergo a clinical evaluation by an
orthopaedist, a functional status assessment by a physical therapist,
and an interview by the site coordinator. Of the 601 patients enrolled
in LEAP, 196 patients (201 limbs) were treated with a rotational
or free flap for coverage of a soft-tissue defect about the tibia
during the initial hospitalization. Because wound complications
are most likely to occur within the first six months after injury,
this analysis is based on 190 patients (195 limbs) treated with
either a rotational or a free flap for whom we were able to determine whether
a wound complication had occurred as of six months after the injury.
The remaining six patients (six limbs) were lost to follow-up.
Data Collection
Data were collected from three primary sources. First, baseline
and follow-up clinical assessments were conducted by orthopaedic
surgeons to characterize the severity of the injury of the limb
or limbs, to document all treatment received, and to record any
wound complications that developed either before or after discharge.
Second, interviews were conducted with each patient to determine
the background sociodemographic characteristics and the health habits
maintained before the injury and to ascertain the use of health services
after the injury. Finally, information pertaining to the circumstances
of the injury and the severity of injuries sustained to other body
regions was obtained either from the trauma registry of each LEAP
site or by abstracting the patient's medical records. A more detailed
description of the data collected from these three sources follows.
Baseline Clinical Assessment
The severity of the injury of the limb or limbs was characterized
by the extent of the osseous injury, the magnitude of the soft-tissue
defect, and the functional status of the neurovascular and muscular
structures of the four tibial compartments. The severity of the
tibial fracture and the severity of the soft-tissue injury about
the tibia were graded according to the ASIF/OTA classification system17,18 by the treating orthopaedic surgeon.
In that system, the tibial fracture is characterized in terms of the
location and degree of comminution, and the extent of the soft-tissue
damage is graded on an ordinal scale ranging from 1 to 5 points,
with 1 point indicating minimal skin breakage from the inside out
and 5 points indicating extensive degloving. In addition, the orthopaedist
documented whether each major artery, vein, nerve, and muscle within the
four tibial compartments was normal, impaired, or nonfunctional
or absent.
The orthopaedic surgeon also was responsible for documenting
all treatment that the injured limb received during the patient's
hospital stay. Detailed information was recorded concerning each
surgical procedure performed (such as vascular repair, wound d衲idement,
fasciotomy, fracture stabilization, revision of fixation, soft-tissue
coverage, bone-grafting, and amputation). Of most relevance to the
current analysis were the operative procedures for soft-tissue coverage
of the tibia. The date and time of all attempts at soft-tissue coverage were
recorded, as was the type of soft-tissue coverage used, whether
the attempt at soft-tissue coverage was successful, and, if not,
the reason for the failure.
Finally, all limb complications that had occurred during the
initial hospitalization and by the time of each follow-up encounter
were documented on the baseline and follow-up orthopaedic evaluation forms
by the attending orthopaedic surgeon. A section on the forms designed
to document the complications that occurred in the injured limb listed
twenty complications that might be anticipated to occur after a
limb injury requiring salvage (such as loss of reduction, pin-track
infection, deep-vein thrombosis, and so on), and additional space
was allocated for the documentation of complications not included
on the list. For each complication that occurred, the attending
surgeon documented: (1) the date that the complication was recognized,
(2) whether treatment was received, (3) whether the treatment was
operative or nonoperative, and (4) whether the treatment was rendered
in an inpatient or an outpatient setting. The specific treatment
was documented for all complications that required inpatient treatment.
Pertinent to the analysis were the complications that may have
been related to inadequate flap coverage: (1) wound infection, (2)
wound necrosis, and (3) flap loss. It should be emphasized that
the exact definition of these three complications was determined
by the treating surgeon and was not determined by specific criteria.
Patient Interviews
The sociodemographic characteristics and health habits of the
patients were determined through an interview conducted by the study
coordinator before the initial hospital discharge. The patients were
asked to report their age and to describe their health habits before
the injury. For example, they were asked whether they smoked and,
if so, how many cigarettes per day15.
They also were read a list of fourteen major medical conditions
(for example, asthma, emphysema, arthritis, hypertension, and so
on) and were asked whether a doctor had ever told them that they
had any of those conditions14.
Medical Record
All injuries were classified according to the Abbreviated Injury
Scale1, which rates individual
injuries by body region on an ordinal scale ranging from 1 point
(minor) to 6 points (unsurvivable). Because the Abbreviated Injury
Scale does not assess the combined effects of multiple injuries,
the Injury Severity Score was also calculated2,3,
by summing the squares of the highest Abbreviated Injury Scale scores
in the three most severely injured body regions.
Analysis
To determine whether the prevalence of short-term morbidity after
application of a rotational flap was significantly different from
that after application of a free flap, we compared the rates of
wound complications associated with each flap type. A wound complication
was defined as the occurrence, within the first six months after
the injury, of a wound infection (after flap application), the need
for a flap revision, or the loss of a flap. Complications that occurred
before the application of the flap were not considered wound complications for
the purposes of this study. Furthermore, wound complications that
resulted in operative treatment were distinguished from those that
did not. Before conducting the analysis, we compared the patient, injury,
and treatment characteristics of the patients for whom we could
determine complications with the characteristics of those for whom
we could not (because they had been lost to follow-up).
The analysis consisted of three phases. First, bivariate analyses
were performed to ascertain whether the two treatment groups differed
significantly with respect to the patient, injury, and treatment
characteristics that were hypothesized to influence the occurrence
of a wound complication. To determine the significance of any observed
differences, the chi-square statistic was used for categorical variables
and the Student t test was used for continuous variables. For both
test statistics, a p value of 0.05 or less was considered significant.
Second, a stratified analysis of the type of flap and the occurrence
of wound complications, controlling for one other injury and patient
characteristic at a time, was conducted to assess whether or not there
were any potential interaction terms or confounding factors that
needed to be taken into account in the multivariate modeling. Confounding factors
and interaction terms were evaluated with use of the Mantel-Haenszel
procedure23. Interaction terms
were considered in the multivariate model if they were significant
at p £ 0.25 according to the Breslow-Day test for homogeneity of
the odds ratios23.
Finally, to concurrently control for the effect of multiple factors
on the prevalence of wound complications that required operative
treatment, a multivariate logistic regression model was developed. The
prevalence of one or more wound complications resulting in operative
treatment was modeled as a function of: (1) the type of flap (free
or rotational), (2) the type of fixation (none, cast, or plate; external
fixation; or intramedullary nailing), (3) the timing of soft-tissue
coverage (zero to three days, four to seven days, or more than seven
days after the injury), (4) the severity of the soft-tissue defect
(ASIF 1, 2, or 3; ASIF 4; or ASIF 5), (5) the location of the soft-tissue
defect as defined by the fracture location (ASIF 41, ASIF 42, ASIF
43 or 44, or no fracture or multiple fracture sites) and the severity
of the fracture (none or ASIF/OTA type A or B, or ASIF/OTA type
C), (6) the interaction between the type of flap and the severity
of the underlying osseous injury as defined by the ASIF/OTA fracture
type, (7) the functional status of structures within each tibial
compartment (normal, or impaired or absent), (8) age as a continuous variable,
(9) smoking status (nonsmoker or less than ten cigarettes per day,
or more than ten cigarettes per day), (10) the presence of comorbidities (none
or one, or two or more), and (11) the Injury Severity Score as a
continuous variable.
We also separately modeled the prevalence of all wound complications
(whether or not operative treatment was required) as a function
of all of the same variables cited above. Factors were considered
significant in the multivariate logistic regression model when the
95 percent confidence interval around the odds ratio did not include
one.
Of the 195 limbs in the present study, eighty-eight (45 percent)
were treated with a rotational flap and 107 (55 percent), with a
free flap (Table I).
There was no significant difference in the follow-up rates between
the limbs treated with a rotational flap (eighty-eight [97 percent]
of the ninety-one limbs were followed) and the limbs treated with
a free flap (107 [97 percent] of the 110 limbs were followed). Furthermore,
there were no significant differences with respect to other treatment, patient,
or injury characteristics between the patients who were followed
and those who were not.
A six-month orthopaedic evaluation was completed for 185 of the
195 limbs for which we were able to determine whether a limb complication
had occurred. Of the ten limbs for which a six-month orthopaedic
evaluation was not completed, four had a wound complication either
during the initial hospitalization or before the three-month orthopaedic evaluation
and therefore were deemed to have had a wound complication within
the six-month period. The remaining six cases had no documentation of
a wound complication at the initial hospitalization, at the three-month
evaluation, or at the twelve-month orthopaedic assessment, so they were
deemed to have had no wound complication within the six-month period.
The patients were predominantly male; 152 (78 percent) of 195
limbs were in male patients (Table II). They were young, with a mean
age of thirty-six years (range, sixteen to sixty-nine years), and
they were relatively healthy before the injury; 110 (56 percent)
of the 195 limbs were in patients who reported the absence of a
chronic medical condition. Approximately two-thirds of the subjects
had been injured as a result of a motor-vehicle or motorcycle accident.
The mean duration of the initial hospital stay was 22.3 days (range,
two to seventy-eight days). In general, the patients treated with
a rotational flap were similar to those treated with a free flap;
the only significant difference with regard to patient characteristics
was that the patients treated with a rotational flap were more likely
to have sustained injuries to other body regions, as reflected by
an Injury Severity Score that was significantly higher than that
of the patients treated with a free flap (p = 0.001).
Ninety-one percent (178) of the 195 limbs in this cohort sustained
a Gustilo-Anderson type-IIIB tibial fracture (Table III). When different
characteristics of the injured limbs are compared with respect to
the type of flap used, several observations are worth noting. First, the
forty limbs that sustained a distal tibial or ankle fracture were
more likely to be treated with a free flap (thirty-five limbs, 88
percent) than a rotational flap (five limbs, 13 percent) (p = 0.001).
Second, limbs treated with a free flap were more likely to have
sustained a severe soft-tissue injury, according to the ASIF classification
of tibial skin defects, than were those treated with a rotational
flap (34 percent compared with 15 percent; p = 0.02). This same
trend also was reflected in the functional status of three of the
four tibial compartments. The limbs treated with a free flap were
significantly more likely to have at least one absent muscle unit in
the anterior (p = 0.001), deep posterior (p = 0.01), or lateral
compartment (p = 0.01) than were limbs treated with a rotational
flap. Finally, the two treatment groups differed significantly with regard
to the type of fracture fixation (p = 0.02) (Table IV). Twenty-three
(22 percent) of the 107 limbs treated with a free flap had fixation
with a plate compared with only eight (9 percent) of the eighty-eight
limbs treated with a rotational flap. In contrast, forty-one (38
percent) of the limbs with a free flap had fixation with an intramedullary
nail compared with forty-six (52 percent) of the limbs with a rotational
flap. The timing of flap coverage did not differ significantly between
the two treatment groups (Table IV).
Overall, the prevalence of wound complications in the study cohort
was 27 percent (fifty-three of 195 limbs) (Table V). The most
common type of wound complication was a wound infection, which occurred
in forty limbs (21 percent). Of the 195 limbs in the study cohort,
fourteen (7 percent) were treated with at least one additional flap
(one limb was treated with three flaps). Ten of them were treated
with the same type of flap (that is, a rotational or free flap), and
four were treated with a different type of flap during the second
attempt at soft-tissue coverage. Of the fifty-three limbs with a
complication, forty-six (87 percent) required operative treatment. Without
adjusting for any characteristics related to the patients, the severity
of injury, or the treatment, one type of flap procedure (that is,
application of a rotational or a free flap) was not significantly
more likely to result in a wound complication than was the other
(Table V). Furthermore,
there were no significant differences within each treatment group
- that is, local muscle flaps were not more likely to be associated with
complications than were fasciocutaneous flaps, and one type of free
flap was not more likely to be associated with complications than
was another type.
When the rate of operative wound complications according to the
type of flap was examined solely on the basis of the type of underlying
fracture (the ASIF/OTA classification), an interaction was detected
between the type of fracture and the type of flap (p = 0.10). In
the group of patients who had no fracture or who had sustained an
ASIF/OTA type-A or type-B fracture, there was no significant difference
in the rates of operative complications between the limbs treated
with a free flap (19 percent) and the limbs treated with a rotational
flap (17 percent). In contrast, the rate of wound complications
requiring operative treatment among the patients who sustained a
more severe (ASIF/OTA type-C) fracture was higher for those who
received a rotational flap (44 percent) than for those who received
a free flap (23 percent) (Table VI). It should be noted that, although
multiple interaction terms were considered in the multivariate analyses,
only the interaction effect of flap type and ASIF/OTA fracture type
was found to be significant.
To examine whether a wound complication resulting in operative
treatment was more likely to occur in one treatment group or another,
depending on the site of injury, we used fracture location as a proxy
for the location of the tibial soft-tissue defect and compared the
rates of complications for each treatment group by fracture site
(Table VII).
No interaction effect was found between the fracture location and
the type of flap. Stated another way, when used to treat a soft-tissue
defect associated with a fracture, neither a rotational nor a free
flap was significantly more or less likely to result in a wound
complication in one area of the tibia than in another.
Because previous investigators have found the timing of the soft-tissue
coverage to be an important determinant of wound complications8,10, a comparison was done between
the wound complications that resulted in operative treatment and the
timing of the initial soft-tissue coverage (Table VIII). Limbs
treated with a flap later in the hospital stay were not significantly
more likely to have a wound complication that required operative
treatment than were limbs with earlier wound coverage.
When the occurrence of a wound complication requiring operative
treatment was modeled multivariately as a function of different
patient, injury, and treatment characteristics, the interaction
effect between the type of flap selected and the grade of the underlying
osseous injury was significant (Table IX). Among the limbs that had an ASIF/OTA
type-C fracture, those that were treated with a rotational flap
were 4.3 times more likely (95 percent confidence interval, 1.38
to 13.64) to have a wound complication that required operative treatment than
were limbs that were treated with a free flap. Analysis of the limbs
that sustained a less severe osseous injury (that is, those that
had no fracture or that had an ASIF/OTA type-A or type-B fracture) revealed
no significant difference in the rate of operative complications
by flap type. The multivariate analyses did not show any other injury,
patient, or treatment characteristic to be a significant predictor
of wound complications that required operative treatment.
We also examined the prevalence of wound complications (regardless
of whether operative treatment was required) as a function of the
same variables considered in the multivariate logistic regression
analysis and found the same interaction effect to be significant.
Finally, because traditional teaching recommends that only free
flaps be applied to distal tibial injuries, the likelihood of wound
complications occurring was examined only for the 128 limbs with
an injury in the proximal two-thirds of the tibia. The results were
the same: the rate of operative complications for limbs with the
most severe grade of underlying osseous injury was significantly
higher for those treated with a rotational flap than for those treated
with a free flap.
High-energy trauma of the lower extremity is a treatment challenge
for orthopaedic and plastic surgeons. A number of investigators
have mentioned the crucial role that soft-tissue reconstruction
plays in the healing of a severely injured lower extremity6,8,9,13. The purpose of the present
study was to identify the factors that may influence the development
of wound complications and therefore the relative success or failure
of soft-tissue reconstruction.
Twenty-seven percent of the limbs in our study sample had at
least one wound complication within the first six months after the
injury. The rates of wound complications did not differ significantly with
respect to the type of flap when no other patient or injury characteristics
were taken into account. However, the two treatment groups were not
equivalent in all respects. First, patients treated with a free
flap had sustained more severe soft-tissue injury as measured by
the ASIF classification and by the number of impaired or absent
functional units in the four tibial compartments. Second, patients
treated with a free flap were more likely to have an underlying
distal tibial or ankle fracture than were patients treated with
a rotational flap. Third, patients treated with a rotational flap
had sustained more associated injuries to other body regions, as
reflected by higher Injury Severity Scores, than had those treated
with a free flap. Finally, the type of fracture fixation differed
significantly between the two groups. Patients treated with a rotational
flap were more likely to have the underlying tibial fracture treated
with intramedullary nailing, whereas patients treated with a free flap
were more likely to have plate fixation.
After controlling for these differences in the severity of the
limb injury, the severity of the overall injury, the treatment rendered,
and the patient characteristics, a significant interaction effect
was demonstrated between the type of flap and the severity of the
underlying osseous injury. The rate of complications differed significantly
with respect to the type of flap, depending on the severity of the underlying
osseous injury. For the limbs with the most severe grade of underlying
osseous injury (an ASIF/OTA type-C fracture), treatment with a rotational
flap was 4.3 times more likely (95 percent confidence interval,
1.38 to 13.64) to lead to an operative wound complication than was
treatment with a free flap. In contrast, the rate of complications
for the limbs with a less severe osseous injury did not differ significantly
with respect to the type of flap.
It is not surprising that limbs with a more severe osseous injury
are more likely to have wound complications than are limbs with
a less severe osseous injury. For limbs with a diaphyseal fracture,
the ASIF/OTA classification increases from A to B to C as the comminution
and the fracture complexity (both markers of injury severity) increase.
Similarly, juxta-articular (ASIF/OTA type-C) fractures, which involve
complete separation of the articular fragments from the intact portion
of the diaphysis, are presumed to be associated with greater severity of
limb injury than are most ASIF/OTA type-A and type-B fractures.
However, it is less clear why rotational flaps are significantly
more likely to be associated with wound complications than are free flaps,
depending on the severity of the underlying osseous injury. There
are at least two possible reasons for this difference. First, when
a muscle flap procedure is performed early in the course of treatment,
the absolute boundaries of the entire zone of injury may be unclear,
and local muscle or tissue selected to provide coverage for a defect
may itself be partially injured. Surgical dissection and rotation
of such tissue necessarily is associated with additional trauma,
and the combination of these factors may result in sufficient tissue
compromise to increase the complication rate. Second, because more
muscle mass is typically available for coverage when a free flap
is used, tenuous or tense coverage is less likely.
The Injury Severity Scores were significantly higher in the rotational
flap group (mean, 14 points) than in the free-flap group (mean,
11 points) (p = 0.001). The reasons for this difference are uncertain,
but one reason may be a perception of some surgeons that free-flap
procedures are associated with more local and systemic complications
than are rotational flap procedures and that these complications
may be more poorly tolerated by more severely injured patients.
Our data tend to refute the contention that free flaps are associated with
more wound complications than are rotational flaps. However, the
association of any one operative or nonoperative treatment with
the prevalence of systemic complications in multiply injured patients
is much more difficult to determine and is beyond the scope of this
investigation. Nonetheless, it has been the experience at many centers
that free-flap procedures can be well tolerated even by severely
injured patients.
Contrary to other studies5,6,8,9,
we did not find the timing of soft-tissue coverage to be an important
determinant of wound complications after severe injury of the lower
limb. In those studies, subjects who were treated with the flap
earlier were found to be less likely to have complications. The
difference between our results and those of previous studies can
most likely be explained by three factors. First, the definitions
of early compared with late coverage varied by study. According
to three5,8,9 of the four studies,
most of our subjects would be defined as having had early coverage.
Seventy-one percent of the limbs in our study sample were treated
with the flap within one week (mean, seven days; range, zero to
thirty-one days) after the initial injury. Second, the definition
of complications varied by study. For example, Byrd et al.5 defined complications as osteomyelitis,
nonunion of the fracture, flap loss, or amputation, whereas we focused
more narrowly on wound complications. Finally, none of the previous
studies controlled for differences in the severity of the limb injury
or the overall injury when complication rates were compared with
respect to the timing of soft-tissue coverage. However, patients
who are treated with a flap later may be more likely to have sustained
a more severe limb injury or overall injury compared with those
who have flap coverage early. It therefore may be misleading to
compare complication rates by the timing of coverage only - that
is, without controlling for differences in other patient or injury
characteristics.
Nieminen et al.19 recently
reported a 2 percent failure rate and a 5 percent amputation rate
in a series of 100 consecutive patients with 104 free-flap transfers
as treatment for traumatic soft-tissue defects about the tibia.
However, thirty of those free-flap procedures were performed because
of complications after the treatment of a closed fracture, and another
twenty-seven were performed following a Gustilo-Anderson type-I,
II, or IIIA open fracture11. In
comparison, all patients with a fracture in the present series had
a Gustilo-Anderson type-IIIB or type-IIIC open injury11. In the series of Nieminen et al.,
patients underwent free-flap reconstruction after a mean delay of twenty-two
weeks after the injury. Only 40 percent of the subgroup treated
within the first six months after the injury underwent flap reconstruction within
the first three weeks after the injury. In contrast, most patients
in the current series underwent soft-tissue reconstruction within
seven days after the injury (Table VIII). These comparisons highlight the
differences between injury profiles and treatment algorithms in the
two series and help to provide an explanation for the differences
in complication rates. The patients in the current series sustained
high-grade open injuries and were treated with early soft-tissue
reconstruction. In the context of this injury profile, the complication
rates associated with both the free flaps and the rotational flaps
are higher than those that would be expected in association with
similar procedures performed after less severe injuries.
Traditional teaching has suggested that gastrocnemius rotational
flaps should be used for coverage of soft-tissue defects in the
proximal third of the leg; soleus flaps, for defects in the middle
third; and free flaps, for defects in the distal third7. Our data did not demonstrate a significant
difference in wound complication rates by fracture location, even
after we controlled for other patient characteristics. However,
two limitations of our study must be kept in mind when interpreting
this result. First, we assumed that the location of the wound was
similar to that of the underlying fracture. However, there may have
been cases where this was not so. The second important limitation
is the sample size. Although our series included 108 subjects who
sustained a diaphyseal fracture, only twenty had a proximal fracture
and only forty had a distal tibial or ankle fracture. The sample
size limits our ability to compare differences in complication rates
adequately among some of these small subgroups (such as those with
a proximal tibial fracture or those with a distal tibial or ankle fracture).
Several other broader limitations of the current study are important
when considering the results. First, the study design is not a randomized,
controlled trial. All of the treatment received by each limb, including
the type of flap, was based on the judgment of the attending orthopaedic
trauma and/or plastic surgeon and was not dictated by a specific
protocol. The surgeons chose the type of flap on the basis of a
host of considerations (such as the location of the wound, the condition
of the local tissues, the severity of the overall injury and the condition
of the patient, and the surgeon and patient preferences). Consequently,
the two treatment groups differed significantly with respect to various
characteristics (such as the severity of the limb injury, the severity
of the overall injury, the location of the underlying fracture,
and the type of fracture fixation). We tried to consider all of
the factors that we thought might influence the development of a
wound complication when we compared complication rates by the type
of flap used. However, there may be other important factors that
we neglected or that were not adequately controlled for with use
of statistical techniques. A prospective randomized, controlled
trial specifically designed to assess the role of the flap type
in the prevalence of short-term wound complications would better
address this issue.
Second, although we found that wound complications (such as wound
infection, wound necrosis, and flap loss) were more likely to be
associated with one type of flap than with another in limbs with
severe underlying fracture, these results do not mean that flap
selection alone led to the wound complications that occurred. Although
we considered only wound complications that occurred after application
of the flap, it is possible that occult infection was occasionally
present before application of the flap. If this were true more often
for limbs treated with a rotational flap than for those treated
with a free flap, the results could be biased. Such a bias would
represent an inherent flaw in the study design. It is conceivable
that surgeons were often more careful about thorough d衲idement
of an open fracture before application of a free flap than they
were before application of a rotational flap.
A third broad limitation of the current study is related to the
methodology of the data collection and grouping. Multiple orthopaedic
surgeons at multiple centers were asked to determine the functional status
of each muscle-tendon structure at the time of soft-tissue coverage
and to grade each injury according to the ASIF/OTA classification
systems for fractures and soft-tissue injuries. Although the interobserver
reliability of each of these classifications is unknown, we were
unable to find a more reliable method of determining the functional
status of each muscle-tendon structure. Furthermore, we believed
that, given the relative simplicity of the ASIF/OTA fracture classification
system at the level of distinguishing type A or B from type C, the
interobserver reliability would likely be high; this theory, however,
is unproved. We grouped osseous injuries according to whether they
were likely to have been associated with high energy (that is, ASIF/OTA
type-C fractures) or were likely to have been associated with lower
energy (that is, no fracture or an ASIF/OTA type-A or type-B fracture)
for the purposes of simplifying statistical analyses. In addition,
metaphyseal and diaphyseal injuries were considered together. Although
some complex type-B injuries, particularly in the periarticular
regions, may be associated with higher energy transfer than some
simple type-C injuries, we believed that, in general, these groupings
placed most of the lower-energy injuries in one group and most of
the higher-energy injuries in the other.
The final broad limitation of the current study is that we considered
only short-term wound complications. It is certainly possible that
some subjects may have had a wound complication after six months;
however, this is unlikely. In 95 percent of the patients who had
a wound complication in our sample, the complication developed by
three months after the injury. Only 5 percent (three patients) had
a wound complication between three and six months after the injury.
Another important factor in understanding the limitations of considering
only short-term complications is that, although the presence or
absence of a wound complication may be related to the overall success
or failure of the soft-tissue reconstruction, a wound complication
in and of itself does not equate to overall failure of such procedures.
Despite these limitations, the results of the current study warrant
further investigation. To our knowledge, this is the first prospective
comparison of short-term wound complications by the type of flap
used after high-energy trauma of the lower extremity. We were able
to determine whether a wound complication occurred by six months
after the injury in 97 percent of a large study sample. These data
are based on the experience of eight orthopaedic trauma centers
and not on that of any one institution. We believe that this study
is the first in which the effects of injury, treatment, and patient
characteristics were considered when an attempt was made to identify
factors that may compromise successful soft-tissue reconstruction.
Our data suggest that, when an underlying osseous injury is severe,
treament of an acute traumatic soft-tissue defect of the lower limb
with a free flap is significantly less likely to have a short-term wound
complication than is treatment with a rotational flap.