Tibial shaft fractures are among the most common fractures in children and adolescents1. Most tibial shaft fractures in children are isolated, and approximately 30% are associated with a fracture of the fibula. Hansen et al. found that most tibial fractures in children are the result of a ground or low-level fall, a sports accident, or being struck by a bicycle2. Sarmiento et al.3 referred to these fractures as low-energy injuries. In general, most closed tibial shaft fractures in children are treated by immobilization in a long leg cast1,4-7. There is, however, no clear consensus regarding the position of immobilization of the knee or when weight-bearing should be allowed. While some authors have recommended initially the use of a long leg cast with the knee flexed 30° to 60° to preclude weight-bearing1,4-7, others have recommended the use of a long leg cast in ≤10° of knee flexion to encourage early weight-bearing8,9.
The purpose of this randomized controlled prospective study was to compare the effectiveness of treatment of tibial shaft fractures in children with the use of (1) a long leg cast with the knee in 60° of flexion and instructions not to bear weight during the initial phase of immobilization and (2) a long leg cast with the knee in 10° of knee flexion and instructions to bear weight as soon as comfort permitted. Our primary hypothesis was that treatment of low-energy tibial shaft fractures with use of these two methods would result in similar times to fracture union. Our secondary hypotheses were that allowing children to bear weight in a more extended cast would result in less disability and would not increase the risk of unacceptable shortening or angulation.
Study Design
We conducted an institutional review board-approved, single-center, randomized, open label trial, in which children who presented with a closed tibial shaft fracture were placed in one of two groups: patients in Group A received a long leg cast with the knee immobilized in 60° of flexion and were given instructions to avoid bearing weight on the affected extremity, and patients in Group B had a long leg cast applied with the knee immobilized in 10° of flexion and were given instructions to bear weight as tolerated on the affected extremity. The randomized trial was registered at ClinicalTrials.gov under the number NCT01238523. Written informed consent was obtained from the parents of each patient. For patients between the ages of seven and fourteen years, child assents were also obtained.
Patients were enrolled between July 2007 and December 2009. They were randomized with use of sealed, sequentially numbered envelopes, in which the sequence was concealed. Sample size estimation was based on time to fracture union, the primary outcome variable in the present study. Assuming that the variance of healing times between the groups is three weeks, for a significance level of 0.05 and power of 80%, the sample size was selected to detect differences of two weeks or greater in time to union. Since two weeks was considered a minimum difference in time to healing that would be of clinical importance, the sample size estimation was conducted to determine the minimum number of patients needed in each group to provide sufficient statistical certainty in detecting a difference of two weeks (or longer) in time to union between the groups. Accordingly, it was determined that a minimum of thirty-six patients in each group, for a total of seventy-two patients, would be included. For variables other than time to union, which were all considered secondary, the observed differences were noted along with the associated p values and any discussion of potential clinical importance. Additionally, post hoc power analysis demonstrated that the sample size was sufficiently large to detect 2 mm of shortening and 1.2° of angulation, given an alpha of 0.05 and 80% power.
Eligibility and Exclusion Criteria
Patients between the ages of four and fourteen years, who presented within seven days of sustaining a closed tibial shaft fracture, with or without an associated fracture of the fibula, regardless of the fracture pattern, were considered for inclusion in the present study. Patients whose fractures had >2 cm of shortening were excluded, as were open fractures, pathological fractures, or those associated with a neuromuscular disorder.
Data compiled on each patient included age, sex, weight, side of injury, affected bone(s) (tibia or both tibia and fibula), location and type of fracture (complete or incomplete), displacement of complete fractures (displaced or nondisplaced), and fracture pattern. All patients had low-energy injuries, defined as those due to falls and direct blows without associated injuries3.
Treatment
All patients were initially evaluated by the on-call orthopaedic resident at our urgent care facility. Anteroposterior and lateral radiographs of the affected tibia were made digitally by an experienced radiology technician, following standard techniques. The fracture was reduced closed with the patient under conscious sedation with the goal to obtain <5° of varus or valgus angulation, <10° of anterior or posterior angulation, <10° of rotational deformity, and <1 cm of shortening9.
For random assignment, a random sequence of eighty A and B values were generated with use of a computerized random-sequence generator by a statistician, who also numbered eighty envelopes on the outside from one to eighty. Slips of paper were inserted in each envelope stating Treatment A or Treatment B, following the sequence of the random A and B values. Envelopes were sealed and given to the clinic staff. The on-call orthopaedic resident then opened the next envelope in sequence of the outside number for each eligible patient. According to the randomization sequence, children in Group A received a long leg cast with the knee immobilized in 60° of flexion. Patients in Group B received a long leg cast with the knee immobilized in 10° of flexion. The amount of knee flexion in the cast was carefully checked with the use of a handheld goniometer. All reduction and casting procedures were performed by the on-call orthopaedic resident with the assistance of an experienced cast technician. Postreduction radiographs were evaluated for residual angulation and shortening.
Patients between the ages of four and eight years were given a walker; patients between nine and fourteen years were given crutches. All patients received specific instructions by the on-call orthopaedic resident with regard to the appropriate use of the assistive device. Patients in Group A were instructed to avoid weight-bearing on the affected extremity for a period of four weeks. Those in Group B were instructed to begin immediate weight-bearing on the affected extremity, guided by pain tolerance.
One week after their initial visit, the patients were evaluated by a pediatric orthopaedist. Casts were inspected for signs of weight-bearing. The patients were evaluated weekly for the first four weeks. Alignment of the fracture was checked at each visit with use of standard anteroposterior and lateral radiographs with neutral rotation of the affected leg. The cast was wedged if the radiographs demonstrated an alignment outside the parameters set at the time of closed reduction9.
Four weeks after the fracture, patients in both groups had the long leg cast replaced with a short leg cast, and all were encouraged to bear weight as tolerated on the affected side. Subsequently, the patients were evaluated at six, eight, ten, twelve, eighteen, and twenty-four weeks after the original injury. Any additional cast changes or complications were documented. The casts were removed once fracture union was obtained.
Outcome Measures
The primary outcome variable was the time to union of the fracture on radiographs. Radiographic union was defined as the presence of callus bridging across three of four cortices as seen on anteroposterior and lateral radiographs5. Radiographic union was determined by a senior pediatric orthopaedic surgeon, not associated with the evaluation process, who reviewed sequential radiographs for all patients in a blinded fashion. This determination was used to calculate the time to obtain radiographic union.
Secondary variables were angulation in the coronal and sagittal planes, and tibial shortening, as measured on the final radiographs. The sagittal and coronal alignment of each fracture was assessed on the radiographs made at radiographic union. Shortening of the affected tibia was measured on radiographs made at each follow-up visit. All radiographic measurements were performed with use of the Cobb tool available in our digital radiology package (version 3.4.0.38, OfficePACS; Stryker Imaging, Flower Mound, Texas).
Other secondary variables pertained to activity. Specifically, the Activities Scale for Kids-Performance (ASK-P) questionnaire was given to each patient at one, six, twelve, and twenty-four weeks after the initial injury. The ASK-P is a validated, highly reliable, self-reported measure that allows for the assessment of physical functioning in children10-13. This scale has been extensively used to evaluate different treatment modalities used for fractures in children14-18. In this study, the ASK-P questionnaire was used to determine the rate of progress during the recovery period.
Data Analysis
A Student t test or a Fisher exact test, as appropriate, was used to compare the two groups in terms of the demographic data, location and type of fracture, length of casting, prereduction and postreduction coronal and sagittal angulations, amount of shortening, any change in coronal or sagittal angulation, the ASK-P values, and the time to achieve healing. A Pearson product-moment coefficient of correlation was used to assess the association between patient factors and outcomes. Given the fact that six of the seven patients who were lost to follow-up (three in each group) had recorded ASK-P scores at the six-week interval, an intent-to-treat analysis was performed with the information on these six patients included.
Source of Funding
There was no external funding for this study.
From July 2007 to December 2009, 102 patients were screened for participation in the study. Of those, fourteen were not included. One patient presented more than seven days after the original injury, three decided to seek follow-up in a facility closer to home, one patient requested to be seen by a physician not participating in the study, and nine patients consented to participate in the study but never returned for follow-up. Eighty-eight patients (eighty-eight tibiae) were then included in the study. Forty-four children were assigned to Group A, and forty-four to Group B (Fig. 1). Seven patients were lost to follow-up before the twelfth week, despite several attempts to contact them. Their results were excluded from the final analysis. Therefore, the results of forty patients in Group A and forty-one patients in Group B were included in this analysis.
Patient and Fracture Characteristics
The characteristics of the patients and their fractures are summarized in the Appendix. There were no significant differences in the patient characteristics between the groups with regard to age, side, or sex. Patients in Group A weighed less than those in Group B (mean, 68.1 and 87.1 lb [30.7 and 39.2 kg], respectively; p = 0.02). There were no significant differences in the fracture characteristics between the groups.
The data for fracture alignment on the prereduction and postreduction radiographs are listed in Table I. The groups showed no clinically important differences in the amount of coronal plane angulation on the prereduction radiographs, although patients in Group B had a significantly greater amount of sagittal plane angulation (0.7° versus 1.8°; p = 0.02). Also, patients in Group A had less shortening than those in Group B (mean shortening of 0.5 mm versus 1.4 mm; p = 0.02). However, this difference was <1 mm. There were no clinically or significant differences in angulation or shortening on the postreduction radiographs.
Time to Union
There was no difference in the time to radiographic union between the groups (a mean of 10.8 weeks [95% confidence interval (95% CI), 9.0 to 12.5 weeks] in Group A versus 10.8 weeks [95% CI, 9.3 to 12.3 weeks] in Group B; p = 0.47). Radiographic union was observed in 2.5% of the patients in Group A versus no patients in Group B at four weeks (p = 0.49), in 40% and 34.1% of patients, respectively, at six weeks (p = 0.65), and in 90% and 92.7% of patients at twelve weeks (p = 0.71). The fractures in all patients in both groups were healed at twenty-four weeks.
Overall, there was a significant association between age and time to healing (r = 0.45, p < 0.0001). The mean age of the patients with fracture union at four, six, twelve, and twenty-four weeks was 5.4, 7.0, 8.5, and 12.1 years, respectively. However, there was no significant difference in age between the groups with regard to those with fracture union at six (p = 0.1), twelve (p = 0.26), and twenty-four weeks (p = 0.15).
Alignment
Wedging of the cast for loss of alignment was required in six patients (15%) in Group A and in four patients (9.7%) in Group B (p = 0.35). Of the ten wedges, five were performed during the first week of follow-up, four during the second week, and one during the third week. Radiographs made after the wedging of the cast demonstrated alignment had been restored to acceptable parameters.
At the time of fracture union, there was no clinically or significant difference in the average coronal or sagittal alignment between the groups (Table I). On average, the final mean coronal plane alignment in Group A was 1.3° compared with 1.2° in Group B (p = 0.9). Similarly, the mean final sagittal plane alignment in Group A was 1° of posterior angulation compared with 0.6° of posterior angulation in Group B (p = 0.2). No patient in either group had >6° of angulation in any plane. There was no association between the age of the patient and the final alignment in either the coronal (r = 0.14) or sagittal (r = 0.26) planes.
Shortening
At the time of fracture union, there was no difference in the average shortening between the two groups (mean, 0.1 mm in Group A versus 0.5 mm in Group B; p = 0.2) (Table I). Only one patient had >1.0 cm of shortening; this was a patient in Group B, in whom the initial shortening was 9.2 mm and was not improved with the fracture reduction.
Activities Scale for Kids-Performance (ASK-P)
The ASK-P data are provided in Table II. Although patients in both groups reported an overall improvement in physical functioning over time, patients in Group B showed greater functional improvement (p = 0.03) (Fig. 2) and better standing skills (p = 0.01) (Fig. 3) at six weeks after the fracture. Although patients in Group B showed a trend toward better personal care activities (p = 0.06), transfer ability (p = 0.08), and locomotion (p = 0.17) at the six-week interval, these differences were not significant. The intent-to-treat analysis performed at the six-week interval confirmed that patients in Group B, compared with patients in Group A, had greater functional improvement (72.9 versus 65.2; p = 0.03) and better standing skills (60.4 versus 49.7; p = 0.01), but also demonstrated better transfer skills (83.0 versus 76.2; p = 0.08).
Complications
No complications were recorded for either group. The mean total length of time in a cast was similar in both groups, with 11.1 weeks (range, 8.3 to 15.4 weeks) for patients in Group A and 11.3 weeks (range, 8.1 to 16.3 weeks) for patients in Group B (p = 0.3).
Sarmiento et al.3 promoted the concept that early weight-bearing enhanced the rate of healing and return to function in adult patients with a tibial shaft fracture. In a large series of 1000 closed tibial shaft fractures, those authors observed that, in general, closed fractures exhibited their maximal shortening at the time of injury and that final shortening did not increase beyond the initial one. As well, the authors suggested that controlled motion at the fracture site was conducive to osteogenesis, presumably by promoting increased vascularity. This treatment philosophy did not seem to influence the care of tibial shaft fractures in children.
Most authors tend to treat tibial shaft fractures in children with an initial period of non-weight-bearing on the injured extremity, perhaps because they believe that early walking will lead to shortening and angulation of the fracture1,5,6,19,20. However, to our knowledge, no study has suggested that promoting early weight-bearing following these pediatric injuries has any adverse effect on the outcome.
Unacceptable angulation and shortening resulting from tibial shaft fractures have been rarely reported. Yang and Letts7, analyzing the outcome of the treatment of ninety-five children with fracture of the tibia with an intact fibula, suggested that a much stronger, thicker periosteum in children acts to minimize shortening. Children have some capacity to compensate for any residual shortening because of the growth stimulation caused by fracture hyperemia21,22. Residual angulatory deformity will remodel after tibial shaft fractures in children7,19,21,22.
In 1978, McCollough et al.23 studied the use of functional fracture brace treatment for fifty-six children with tibial shaft fractures. Their initial treatment consisted of closed reduction and application of a long leg cast, later replaced with either a below-the-knee functional weight-bearing cast or a fracture brace. The average healing time was 13.2 weeks. The authors conceded that the rapid healing time and relative lack of joint stiffness and atrophy characteristically seen in children limited the applicability of this treatment method. They suggested that the only potential advantage of using a fracture brace in children was the greater freedom afforded to both the child and the parents during the course of treatment.
Recently, Jenkins et al.8 retrospectively reviewed the cases of eighty children with a complete tibial shaft fracture who were treated in a long leg cast and were asked to begin to bear weight on the affected limb at an average of 6.8 days after the injury. Despite the fact that the children were encouraged to bear weight early, the authors observed that the children in their series delayed doing so until an average of fifteen days after the injury. The authors reasoned that the children may have delayed bearing weight on the injured limb until they were physically comfortable. None of these patients healed with unacceptable angulation or shortening of the fracture. The authors concluded that early weight-bearing may promote healing and can be allowed without an increase in the rate of complications.
The results of this randomized trial support the idea that early weight-bearing can be encouraged without jeopardizing the outcome of treatment in pediatric patients. None of our patients in Group B had unacceptable angulation or shortening at the time that the tibial shaft fracture had united, and there was no difference in the need to wedge a cast because of loss of the original alignment between the groups. Ten (12.3%) of eighty-one patients in the present study required wedging of casts to restore alignment. This was comparable with the 19.7% rate reported by Yang and Letts7 in their series of ninety-five children with tibial shaft fractures who were managed with an initial period of non-weight-bearing.
One potential benefit of early weight-bearing may be to lessen disability during the cast and early rehabilitation phases of treatment. We used the ASK-P questionnaire to assess physical disability at different time intervals during treatment. Although patients in both groups reported an overall improvement in physical functioning over time, patients in the group who were permitted to bear weight earlier showed greater functional improvement and better standing skills at six weeks after the original cast was applied. Patients in the flexed cast group (Group A) demonstrated significantly better dressing skills at the twelve-week interval. While we are unsure about the reason for this difference, we are reassured by the fact that other factors at twelve weeks, more closely associated with functional disability, were not significantly different between the groups.
This study has several limitations. The results observed are applicable only to low-energy tibial fractures and, therefore, cannot be generalized to all tibial shaft fractures in children. Despite the verbal instructions that were given, it was difficult to precisely determine when each child began to bear weight. Although assessed in clinic, we could not quantitate cast wear in any meaningful way. Future investigations may wish to incorporate pressure sensor monitors into the casts to more accurately assess the time of onset of weight-bearing after immobilization of these injuries. The accuracy of current computer tools to assess length on digital radiographs is limited, especially when these measurements are small. Although we used up-to-date technology, the accuracy of the current measurement software is not well established. Additionally, the time interval between the evaluations of each patient may have been too great to detect a difference in the time of fracture union. Perhaps if the patients were followed at weekly intervals throughout the immobilization period, a difference in healing time would have been detected, as was observed in the study by Jenkins et al.8. Lastly, other indirect measures of improvement such as the number of days the assistive devices were used and the number of missed school days were not recorded in this study. Those indirect measures could have provided clinically and socially relevant information, and we recommend that future investigations include these measures.
In summary, the results of this randomized trial suggest that low-energy tibial shaft fractures in children may be treated by immobilizing the limb with the knee flexed 10° to 15° and encouraging early weight-bearing on the affected extremity without affecting the time to union or increasing the risk of angulation and shortening at the fracture site. While it appears that early weight-bearing has the potential to lessen the patient’s disability while in the cast, it does not seem to change the patient’s functional activity in the longer term. Accordingly, the treatment of a low-energy tibial shaft fracture will likely continue to be a matter of preference by the treating orthopaedist. However, on the basis of the results of this study, we modified our clinical practice, treating all low-energy tibial shaft fractures in children between the ages of four and fourteen years with an initial long leg cast, with the knee immobilized in 10° to 15° of knee flexion, and allowing them to bear weight as tolerated on the injured extremity. We believe that the early functional gains demonstrated by this group of children may permit some children to return to school sooner and allow their parents to return to work earlier, potentially lowering the indirect cost of treating these fractures.
Disclosure: None of the authors received payments or services, either directly or indirectly (i.e., via his or her institution), from a third party in support of any aspect of this work. One or more of the authors, or his or her institution, has had a financial relationship, in the thirty-six months prior to submission of this work, with an entity in the biomedical arena that could be perceived to influence or have the potential to influence what is written in this work. No author has had any other relationships, or has engaged in any other activities, that could be perceived to influence or have the potential to influence what is written in this work. The complete Disclosures of Potential Conflicts of Interest submitted by authors are always provided with the online version of the article.