Treatments for paediatric femoral fractures: a randomised trial

Articles

Treatments for paediatric femoral fractures: a randomised trial James G Wright,Elaine E L Wang, Janice L Owen, Derek Stephens, H Kerr Graham, Michael Hanlon, Gary R Nattrass, Rick A K Reynolds, Peter Coyte

Lancet 2005; 365: 1153–58 See Comment page 1116

Summary Background Treatments for femoral fractures in children vary widely and have been investigated only in case series. We did a multicentre randomised trial to compare malunion rates after external fixation and after early application of a hip spica cast for paediatric femoral shaft fractures. Methods All children aged 4–10 years with femoral fractures, admitted to four paediatric hospitals, were randomly assigned early application of hip spica or external fixation. The primary outcome was malunion at 2 years after the fracture. Secondary outcomes were scores on the RAND physical function child health questionnaire and the posthospitalisation behavioural questionnaire, and parents’ and children’s ratings of overall satisfaction with treatment. Analysis was by intention to treat based on children who reached the 2-year evaluation. Findings Of 60 children assigned to the hip-spica group, 56 reached the 2-year assessment; of them, six (11%) required other forms of treatment because of unacceptable loss of reduction. Of 48 children assigned external fixation, 45 reached the 2-year assessment; two (4%) had refractures and five (11%) required operative adjustment of the fixator. The rate of malunion was significantly higher in the hip-spica group than in the external-fixator group (25/56 [45%] vs 7/45 [16%]; 95% CI for difference 12–46%; p=0·002). The two groups had similar mean scores for the RAND physical function health questionnaire (0·34 vs 0·45; 95% CI for difference, –0·57 to 0·34; p=0·61), for the post-hospitalisation questionnaire (106·8 vs 106·3; –4·9 to 5·9; p=0·86), and for parents’ satisfaction (4·3 vs 4·2; –0·3 to 0·6; p=0·5) and children’s ratings of happiness with treatment (6·9 vs 7·7; –2·2 to 0·5; p=0·21). Interpretation Early application of hip spica has a small role in the treatment of paediatric femoral fractures. Future trials need to compare external fixation with flexible intramedullary nails.

Introduction Femoral fractures in children are common, with annual rates of roughly 20 per 100 0001,2 and treatment costs of up to US$13 490 per patient.3–5 The absence of evidence from high-quality randomised trials has led to wide treatment variability for children’s femoral fractures. A survey of paediatric orthopaedic surgeons showed that treatment ranges from long hospital stays to operative fixation.6 Published work on treatment of femoral fractures consists almost entirely of uncontrolled case series.7 An evidence-based working group of the Pediatric Orthopaedic Society of North America in 2002 was unable to make any specific recommendations about the preferred treatment of paediatric femoral fractures.6 Because children’s bones remodel with time after a fracture, femoral fractures are generally perceived to heal satisfactorily irrespective of the form of treatment.8 Treatment choices for femoral fractures in children are therefore usually based on surgeons’ preferences and family considerations. Furthermore, shifts in practice have occurred, such as increased use of flexible intramedullary nails, not based on evidence but because of general perceptions. For example, external fixation is not commonly used because of the belief that this treatment is unacceptable to families.9 Early application of hip spica (a cast from chest to toes) allows the child to leave hospital within a few days.10,11 This treatment, however, is cumbersome.7,12 External www.thelancet.com Vol 365 March 26, 2005

fixation (pins through the bone proximal and distal to the fracture with an external connecting bar) also allows early discharge from hospital,13 is less cumbersome than hip spica, and can be effective at controlling the fracture position, theoretically leading to reduced rates of malunion.14–16 However, there is widespread belief that external fixation is associated with high rates of refracture17–19 and is less acceptable to families than hip spica.20 We did a randomised trial to investigate malunion rates, physical function, behavioural disturbances, and satisfaction with early application of hip spica compared with external fixation for paediatric femoral fractures.

Divisions of Orthopaedic Surgery (Prof J G Wright MD, J L Owen BScPT), and Population Health Sciences (Prof J G Wright, E E L Wang MD, D Stephens MSc), Hospital for Sick Children, Toronto, ON, Canada; Departments of Surgery (Prof J G Wright), Public Health Sciences (Prof J G Wright), and Health Policy, Management, and Evaluation (Prof J G Wright, Prof P Coyte PhD), University of Toronto, Toronto, ON, Canada; Clinical Development, Connaught Campus, Aventis Pasteur, Toronto, ON, Canada (E E L Wang); Department of Orthopaedic Surgery, University of Melbourne, Royal Children’s Hospital, Parkville, Victoria, Australia (Prof H K Graham MD); Department of Paediatric Orthopaedic Surgery, Starship Children’s Hospital, Auckland, New Zealand (M Hanlon MD); Department of Paediatric Orthopaedic Surgery, Royal Children’s Hospital, Melbourne, Australia (G R Nattrass MD); and Department of Orthopaedics, USC School of Medicine, Children’s Hospital Los Angeles, Los Angeles, CA, USA (R A K Reynolds MD) Correspondence to: Prof James G Wright, Department of Surgery, The Hospital for Sick Children, 555 University Avenue, Toronto, ON M5G 1X8, Canada [email protected]

Methods Participants Children aged 4–10 years admitted to the Hospital for Sick Children, Toronto, Canada; Royal Children’s Hospital, Melbourne, Australia; Los Angeles Children’s Hospital, Los Angeles, USA; and Starship Children’s Hospital, Auckland, New Zealand with femoral fractures were eligible for study. The recruitment was expected to take 3 years but took 6 years. The centres entered the study at different times but the study began in October, 1994, and the last patient was enrolled in October, 2000. The last patients’ follow-up at 2 years (the primary endpoint) occurred in October 2002. Reasons for exclusion were hip fracture, distal femoral physeal fracture, significant head injury (score on the Glasgow coma scale less than 1121), 1153

Articles

pathological fracture, or an open fracture. The trial received ethics approval from the Institutional Review Board (Declaration of Helsinki, Edinburgh 2000). Study coordinators not involved in the provision of care to the children obtained written informed consent from the parents before treatment was assigned.

234 patients assessed for eligibility 126 excluded

108 randomised

Procedures To ensure masking at randomisation, children were allocated in blocks of variable size by three strata: hospital, surgeon, and age (4–6 years and 7–10 years). Opaque envelopes, which were sequentially numbered, were kept in the operating room. Once consent for inclusion in the study was obtained, the envelopes were opened by the operating surgeon to allow for appropriate preparation for the operating-room staff. A biostatistician who had no further involvement in the trial created the computer-generated randomisation schedule using SAS (version 8.2). Protocols were standardised with written descriptions of treatment guidelines. Children in the hip-spica group were given general anaesthesia and placed in cast incorporating the fractured limb, not including the foot, with the hip and knee flexed about 70º. We used previously reported treatment guidelines for early application of hip spica in children 10 years and younger;10,11 adequate closed reduction was defined as 1–2 cm of shortening, no posterior angulation, less than 20º of anterior angulation, no varus angulation, and less than 15º of valgus angulation. Children were allowed to walk with crutches, if able, and were discharged from hospital and reviewed weekly as outpatients. A change in position to more than 20º anterior or posterior angulation or more than 15º varus or valgus angulation during these 3 weeks was treated by a wedging of the cast in the clinic with sedation or readmission to hospital for repeat closed reduction under general anaesthesia. Children in the external-fixator group were given general anaesthesia for a closed reduction of the fracture and application of a dynamised Orthofix external fixator (Orthofix, McKinney, TX, USA).13 Satisfactory reduction was defined as up to 1 cm of overlap, less than 15º of varus or valgus angulation, and less than 20º anterior or posterior angulation. Children were allowed to walk with crutches, if able, and were discharged from hospital in 1–2 days and reviewed weekly. A change in position (as defined above) before fracture healing was treated with a repeat closed reduction of the fracture and adjustment of the fixator under general anaesthesia. The fixator was removed in the clinic after the child had received oral analgesics. The primary outcome measure was fracture malunion at 2 years after the fracture, defined as any of: limb-length discrepancy (as assessed by CT) of more than 2 cm, more than 15º of anterior or posterior angulation, or more than 10º of varus or valgus angulation (as assessed from radiographs). Because children’s fractures remodel with 1154

60 allocated early hip spica 4 lost to follow-up

48 allocated external fixation 3 lost to follow-up

56 reached 2-year evaluation

45 reached 2-year evaluation

Figure 1: Trial profile

time,22 the acceptable initial closed reduction was greater by 5º than the final definition of malunion. Clinical assessments at 3 months, 9 months, 15 months, and 24 months after the fracture were done by physicians unaware of the child’s treatment regimen, previous assessments, and study protocol. Children wore tights to mask treatment allocation. For assessment of the primary outcome of fracture malunion a single reader, unaware of the treatment children received, the date of the radiograph, and the study protocol, reviewed all radiographs using a standard protocol at a central location. The intraclass correlation coefficient of the intra-observer test-retest reliability of the assessment of malunion from radiographs was 0·97 (95% CI 0·94–0·99). In the original protocol we had also planned to include more than 15º of rotational deformity (determined from clinical examination of hip rotation) as part of the composite definition of malunion. Despite standardised protocols, rotational assessment was unreliable; the intra-observer test-retest reliability range of the intraclass correlation coefficient was roughly 0·05–0·40 for different features of hip rotation. Thus analyses were done both with and without rotational deformity in the definition of malunion. Secondary outcomes included the self-administered reliable and valid RAND child health status scale.23,24 The RAND scale has four subscales: mental health; general health perception; developmental milestones; and physical function. The 13-question physical function subscale of RAND, completed by parents and their child together, is scored from 0 to 13 with low scores indicating better function. The post-hospitalisation behavioural questionnaire25,26 is completed by the parent. Parents rated children’s dysfunctional behaviours using 27 questions responding in five categories (scored from 1 to 5, respectively) of “much less”, “less”, “no change”, “more”, and “much more” than before admission at 6 weeks and 12 weeks after the fracture. The total scores ranged from 0 www.thelancet.com Vol 365 March 26, 2005

Articles

to 135, with scores greater than 81 indicating more, and scores less than 81 indicating fewer dysfunctional behaviours. To assess acceptability of treatment, including scarring from pin-tracts of the external fixator, children and their parents separately rated treatment at 3 months and 24 months on ordinal rating scales with 11 categories from “very unhappy” to “very happy” for children and five categories of “very dissatisfied” to “very satisfied” for parents. Duration of treatment was defined as the number of days the child wore the hip spica or external fixator.

Statistical analyses Acceptable union was the primary outcome of treatment.27–29 A difference of 20% representing the expected reduction in malunion rate from published work7 was deemed to be clinically significant and necessary to have clinicians widely adopt the more invasive external fixator. With the assumption of satisfactory union rates of 75% in the hip-spica group and 95% in the external-fixation group and a type I error of 0·05 and a type II error of 0·2, 50 patients were needed in each group. The sample size was increased by 10% to allow for dropouts. No interim analysis was done. The analyses were undertaken with and without adjustment for randomisation strata and other confounding characteristics.30 Final interpretations for unadjusted analysis were based on the difference between two independent malunion rates by the 2 test. For the adjusted analysis, a logistic regression was used. Patients who underwent the 2-year assessment were analysed according to the group to which they had been randomly allocated.31 The two treatment groups were also compared for physical function, behaviour, and patients’ satisfaction (independent two-tailed t tests). For the secondary outcomes, not all patients completed the questionnaires at the 2-year assessment. The secondary outcomes were analysed for patients with completed questionnaires and also with imputed values. All tests were done with the use of a two-sided  level of 0·05.

Role of the funding source Canadian Institutes of Health Research, a national funding agency, provided peer review on the grant application. Other than this involvement, the funding sources had no role in the study design, data collection, data analyses, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Results Of 234 children whose families were approached to participate, 108 (46%) were enrolled. Of those excluded, informed consent could not be obtained for 44 (35%), parents preferred one or the other treatment in 41 (33%), doctors chose another form of treatment for 21 (17%), parents refused to consider participation in 13 (10%), www.thelancet.com Vol 365 March 26, 2005

and no reason was stated for 7 (6%) children. Of 108 enrolled children, 101 (94%) completed the 2-year assessment (figure 1). The baseline characteristics of the participants are shown in table 1. Of 56 children randomly assigned hip spica who completed the 2-year assessment, six (11%) received other forms of treatment due to unacceptable loss of reduction; three patients received external fixation, and three received flexible intramedullary nails. Two other patients were admitted for repeat closed reduction and reapplication of the hip spica. Of 45 children randomly assigned external fixation who completed the 2-year assessment, none received another form of treatment but two (4%) had refractures after the external fixator was removed—each child was treated with a plate (which was not removed). Five children (11%) were readmitted to hospital for repeat closed reduction and adjustment of the external fixator. The mean duration of treatment—ie, the length of time the child was immobilised with a hip spica or time until removal of the external fixator, was longer in the external-fixator group than in the hip-spica group (77 vs 58 days; p=0·01). As shown in table 2, the rate of malunion was significantly higher in the hip-spica group than in the external-fixator group (odds ratio=4·38 [95% CI 1·67–11·47]; relative risk=2·86 [95% CI 1·36–6·02]). The difference in malunion rates was closely similar by centre, age-group, fracture type, and initial shortening (figure 2). The number needed to treat was 3·4 (95% CI 2·2–8·3), thus a malunion was avoided for every four patients treated with an external fixator rather than a hip spica.32 With a definition of malunion that included the clinical assessment of malrotation, the malunion rate was 57% in the hip-spica group compared with 40% in the external-fixation group (p=0·09). Multiple logistic regression was used to assess whether sex, age, mechanism of injury, centre, and initial shortening confounded the relation between malunion and the two treatment groups. The unadjusted odds ratio for Hip-spica group (n=60) Mean age, years 6·3 Age-group 4–6 years 22 (37%) Age-group 7–10 years 38 (63%) Boys/girls 44 (73%)/16 (27%) Types of fracture Spiral 27 (45%) Oblique 12 (20%) Transverse 21 (35%) Diaphyseal fracture 60 (100%) Mechanism of injury Fall 33 (55%) Pedestrian—motor-vehicle collision 11 (18%) Other 11 (18%) Motor-vehicle collision 5 (8%)

External-fixation group (n=48) 6·5 19 (39%) 29 (61%) 32 (67%)/ 16 (33%) 14 (30%) 17 (35%) 17 (35%) 48 (100%) 24 (50%) 14 (29%) 8 (17%) 2 (4%)

Data are number and percentage, unless otherwise stated.

Table 1: Baseline characteristics of patients

1155

Articles

Number of patients

Difference (95% CI)

p (2)

29% (12 to 46) 6% (–5 to 17) 23% (7 to 38) 3% (–4 to 10) 7% (–10 to 25)

0·002 0·33 0·007 0·42 0·43

Hip-spica group External-fixation (n=56) group (n=45) Overall malunion 25 (45%) Leg length discrepancy 2cm 7 (13%) Anterior/posterior angulation 15° 19 (34%) Varus/valgus angulation 10° 3 (5%) Rotational malunion 19 (34%)

7 (16%) 3 (7%) 5 (11%) 1 (2%) 12 (27%)

*Analyses were by intention to treat of children who completed the 2-year assessment.

Table 2: Primary outcomes*

malunion for external fixation was 0·23 (95% CI 0·09–0·60) and after adjustment the odds ratio was 0·12 (0·04–0·42). None of the variables listed above met the 5% level of significance. Scores for the RAND physical function, the overall RAND, the post-hospitalisation behavioural questionnaire, parents’ satisfaction, and children’s satisfaction were closely similar in the hip-spica and external-fixator groups (table 3). Comparisons between the two groups that used imputed values for RAND, post-hospitalisation behavioural questionnaire, and parents’ and children’s satisfaction provided similar results (data not shown). The Spearman rank correlation between parents’ and children’s ratings of satisfaction with treatment was r=0·45 (p=0·001). None of the comparisons for the other RAND subscale scores showed a significant difference. Those with malunion had similar RAND physical function scores to those without malunion (0·3 vs 0·4, p=0·5; difference=–0·1, 95% CI –0·6 to 0·3). Adverse Age category

Initial shortening

0·7

Malunion rates

0·6 0·5

Spica

Spica

Spica

Spica

0·4 Overall rate of malunion

Overall rate of malunion

0·3 0·2

Ex-fix

Ex-fix

Ex-fix

Ex-fix

0·1 0 4–6 years

Age category

7–10 years 0–2·5 cm

Initial shortening

2·5 cm

Fracture

Centre 0·7 Spica

Malunion rates

0·6 0·5

Spica Spica

Spica

0·4 Overall rate of malunion 0·3

Ex-fix

0·2

Spica

Spica Overall rate of malunion Spica

Ex-fix

Ex-fix

Ex-fix

0·1

Ex-fix

Ex-fix

Ex-fix

0 Toronto

Melbourne

Los Angeles

Auckland

Spiral

Centre

Figure 2: Malunion by treatment group and other characteristics

1156

Oblique Fracture

Transverse

events, other than loss of reduction in the hip-spica group, needing other forms of treatment were restricted to the external-fixator group. Of the 45 patients in this group, 20 (45%) were treated for pin-tract infections with oral antibiotics, but none required admission to hospital and no chronic infections were reported.

Discussion We have shown that early application of hip spica compared with external fixation results in significantly higher rates of malunion 2 years after the fracture. If families of children with femoral fractures choose hipspica treatment, they should be advised that outcomes in terms of overall physical function, behaviour, and satisfaction are similar to those with external fixation. Although the long-term implications are unknown, the rate of malunion (by a generally accepted definition) would be more than three times greater than with external fixation. We chose 2 years as the time of final assessment because 85% of remodelling should have occurred by this time.22 Thus most children with malunion at 2 years after the fracture will have permanent deformity. Although there are no universally accepted criteria for malunion, those used in our study are similar to those used by other investigators.11,20,33 The long-term consequences of malunion are uncertain but children treated with hip spica have persistent weakness and gait abnormalities even 2 years after the fracture.34,35 By contrast, Hedin and Larsson36 found no difference between muscle strength in patients treated with external fixation and healthy controls. Furthermore, several research groups have suggested that permanent leg deformity could result in premature arthritis.37–39 We used commonly accepted treatment protocols in this study, which suggest up to 2 cm shortening in the hip-spica treatment group.10,11 Although control of fracture shortening is more difficult with hip spica than with external fixation, the acceptance of up to 2 cm of shortening in the hip-spica group (by contrast with 1 cm in the external-fixation group) could have resulted in greater shortening 2 years after the fracture. Treatment could potentially also have affected the rates of remodelling and overgrowth. The malunion rates recorded here are higher than previously reported7 and greater than the 25% expected malunion rate for patients who received hip spica, probably because of the complete follow-up, intention-to-treat analyses, explicit criteria, and masked assessment of outcomes.40 Our study was a multicentre international trial. The advantage of having several sites is that clinically important trends might be missed at individual sites because of small numbers. Three issues could affect the generalisability of this trial. First, surgeons might not be proficient in one or both treatments. The difference in malunion rates between the two treatment groups was consistent across all four study sites, indicating the www.thelancet.com Vol 365 March 26, 2005

Articles

difference is due to the procedure and not technical proficiency. Staff from all four centres were experienced in both techniques and, therefore, the results are probably typical of other paediatric centres. Second, the age-range of patients enrolled in this study was wide; the difference in malunion rates might have been smaller in younger children. However, malunion rates for the two treatments were similar in the two age-groups investigated. Finally, are the results applicable for different fracture types? A previous study described the telescope test,41 whereby shortening under general anaesthesia was predictive of fracture malunion due to shortening at final assessment. However, malunion rates were similar in different fracture types and for different amounts of initial shortening. Thus the results of this study are robust and relevant to paediatric femoral fractures. External fixation is generally perceived to be associated with high rates of refracture.20 The low refracture rate in our study is probably due to the standardised treatment protocol. Furthermore, reports of high refracture rates after removal of the external fixators can be attributed to the use of fixators in children older than the participants in our study. Additionally, external fixators are commonly used for children with open fractures, which are known to be slower to heal, probably leading to increased refracture rates.17,18 However, refractures are an important complication because they necessitate repetition of the whole treatment process. Another perceived disadvantage of external fixation is their unacceptability to families.9 External fixation required readjustment in 11% of children and long treatment duration and pin-tract infections were common. The similarity of ratings of children and parents for the two treatments shows that contrary to clinicians’ beliefs, however, external fixation is acceptable to families. The trade-off for families in choosing treatment will be a 4% risk of refracture, an extra 2 days in hospital, and an extra 19 days treatment with external fixation compared with early application of hip spica for which malunion rates are three times greater. We expected that the functional outcomes in the two treatment groups at 2 years after the fracture would be similar. Furthermore, the low scores in both groups suggest little short-term physical disability from femoral fracture. By contrast, in both groups the rate of behavioural disturbance was high.42 Acute illness and hospital admission lead to behavioural disturbances in children.25,26,43–45 The average scores of the posthospitalisation behavioural questionnaire in this study of more than 100 contrast with average scores of 77 in children 6 weeks after admission to an intensive-care unit45 and of 84 in children a week after cardiac catheterisation.43 The high degree of behavioural disturbances in both groups in our study needs further investigation. A possible explanation is that these behavioural disturbances are not a result of the fractures, www.thelancet.com Vol 365 March 26, 2005

Number of patients

Difference (95% CI) p (t test)

Hip-spica group External-fixation (n=56) group (n=45) Overall RAND score 68 RAND physical subscale 0·3 Post-hospitalisation behavioural questionnaire 106·8 Initial stay in hospital, days 3·4 Total stay in hospital, days 4·1 Duration of treatment, days 58 Parental satisfaction 4·3 Child satisfaction 6·9

69 0·4 106·3 5·3 5·9 77 4·2 7·7

1·0 (–4·2 to 2·1) –0·1 (–0·6 to 0·3) 0·5 (–4·9 to 5·9) –1·9 (–2·8 to -1·0) –1·8 (–3·9 to -0·3) –19 (–28 to -11) 0·1 (–0·3 to 0·6) –0·8 (–2·2 to 0·5)

0·5 0·61 0·86 0·01 0·02 0·01 0·49 0·21

*Analyses were by intention to treat of children who completed the 2-year assessment.

Table 3: Secondary outcomes*

but show the behaviour or personality indicative of children who sustain fractured femurs. A limitation of this study was that other treatments for paediatric femoral fractures were not assessed. To address the study objective of choosing the best treatment, and recognising the complexities of a randomised clinical trial, we chose the two interventions in this trial because they are commonly used and favourable outcomes have been reported.10,11,13,15,33 A popular treatment for paediatric femoral fractures not investigated here was flexible intramedullary nails.46 Although flexible nails result in early discharge and can control fracture position,47 potential disadvantages include the need for a second general anaesthetic if the nails are removed, the possibility they could migrate,48 and reported rates of complications, such as malunion, no union, and pain and skin erosion at insertion site, in up to 49% of patients.49 Although a randomised trial has compared external fixation with flexible nails,9 the trial had only 19 patients in total, no statistical analysis was provided, and no comprehensive assessment of fracture malunion was undertaken. For the future, further research will need to compare the results of external fixation with flexible intramedullary nails and address the importance of femoral malunion in children, rates of remodelling and overgrowth, and the high degree of behavioural disturbances with treatments. Contributors All of the authors contributed to the conception of the research. J G Wright, E E Wang, J L Owen, and D Stephens did the analyses. All authors reviewed and approved the final version of the report. Conflict of interest statement We declare that we have no conflict of interest. Acknowledgments Financial support for this study was provided by grants from the Medical Research Council (MRC) of Canada from 1994–2002 (grants MT12788, MA-12788, MA-12788 renewal, and 95048) and the Canadian Orthopaedic Research Education Association. Peter Coyte is the Canadian Health Services Research Foundation/Canadian Institutes of Health Research Chair in Health Care Settings and Department of Health Policy, Management and Evaluation, University of Toronto. James Wright was the recipient of a Canadian Institutes of Health Research (CIHR) (formerly MRC) Investigator Award and is the Robert B Salter Chair of Surgical Research.

1157

Articles

References 1 Nafei A, Teichert G, Mikkelsen SS, Hvid I. Femoral shaft fractures in children: an epidemiological study in a Danish urban population, 1977–86. J Pediatr Orthoped 1992; 12: 499–502. 2 Hinton RY, Lincoln A, Crockett MM, Sponseller P, Smith G. Fractures of the femoral shaft in children: incidence, mechanisms, and sociodemographic risk factors. J Bone Joint Surg 1999; 81: 500–07. 3 Buechsenschuetz KE, Mehlman CT, Shaw KJ, Crawford AH, Immerman EB. Femoral shaft fractures in children: traction and casting versus elastic stable intramedullary nailing. J Trauma 2002; 53: 914–21. 4 Coyte P, Bronskill SE, Hirji ZZ, Daigle-Takacs G, Trerise BS, Wright JG. An economic evaluation of two treatments for paediatric femoral shaft fractures. Clin Orthop Rel Res 1997; 336: 205–15. 5 Newton PO, Mubarak SJ. Financial aspects of femoral shaft fracture treatment in children and adolescents. J Pediatr Orthoped 1994; 14: 508–12. 6 Sanders JO, Browne RH, Mooney JF, et al. Treatment of femoral fractures in children by pediatric orthopaedists: results of a 1998 survey. J Pediatr Orthoped 2001; 21: 436–41. 7 Wright JG. The treatment of femoral shaft fractures in children: a systematic overview and critical appraisal of the literature. Can J Surg 2000; 43: 180–89. 8 Macnicol MF. Fracture of the femur in children. J Bone Joint Surg 1997; 79B: 891–92. 9 Bar-On E, Sagiv S, Porat S. External fixation or flexible intramedullary nailing for femoral shaft fractures in children: a prospective randomised study. J Bone Joint Surg [Br] 1997; 79B: 975–78. 10 Irani RN, Nicholson JT, Chung SMK. Long-term results in the treatment of femoral shaft fractures in young children by immediate spica immobilization. J Bone Joint Surg 1976; 58A: 945–51. 11 Sugi M, Cole WG. Early plaster immobilization for fractures of the femoral shaft in childhood. J Bone Joint Surg [Br] 1987; 69B: 743–45. 12 Hughes BF, Sponseller PD, Thompson JD. Pediatric femur fractures: effects of spica cast treatment on family and community. J Pediatr Orthoped 1995; 15: 457–60. 13 De Bastiani G, Aldegheri R, Brivio LR. The treatment of fractures with a dynamic axial fixator. J Bone Joint Surg 1984; 66B: 538–45. 14 Aronson J, Turskey EA. External fixation of femur fractures in children. J Pediatr Orthoped 1992; 12: 157–63. 15 Blasier RD, Aronson J, Turskey EA. External fixation of pediatric femur fractures. J Pediatr Orthoped 1997; 17: 342–46. 16 Evanoff M, Strong ML, MacIntosh R. External fixation maintained until fracture consolidation in the skeletally immature. J Pediatr Orthoped 1993; 13: 98–101. 17 Kirschenbaum D, Albert MC, Robertson W, Davidson RS. Complex femur fractures in children: treatment with external fixation. J Pediatr Orthoped 1990; 10: 588–91. 18 Probe R, Lindsey RW, Hadley NA, Barnes DA. Refracture of adolescent femoral shaft fractures: a complication of external fixation. J Pediatr Orthoped 1993; 13: 102–05. 19 Skaggs DL, Leet AI, Money MD, Shaw BA, Hale JM, Tolo VT. Secondary fractures associated with external fixation in pediatric femur fractures. J Pediatr Orthoped 1999; 19: 582–86. 20 Flynn JM, Ganley TJ, Skaggs D, Kay RM, Leitch KK, Sponseller PD. The operative management of pediatric fractures of the lower extremity. J Bone Joint Surg [Am] 2002; 84: 2288–300. 21 Tepas JJ, Mollitt DL, Talbert JL, Bryant M. The paediatric trauma score as a predictor of injury severity in the injured child. J Pediatr Surg 1987; 22: 14–18. 22 Reynolds DA. Growth changes in fractured long bones: a study of 126 children. J Bone Joint Surg [Br] 1981; 63B: 83–88. 23 Eisen M, Donald CA, Ware JEJ, Brook RH. Conceptualization and measurement of health for children in the health insurance study; RAND R-2313-HEW. Santa Monica, CA, USA: RAND Health, 1980. 24 Eisen M, Ware JE, Donald CA, Brook RH. Measuring components of children’s health status. Med Care 1979; 17: 902–21. 25 Vernon DTA, Schulman JL, Foley JM. Changes in children’s behavior after hospitalization: some dimensions of response and their correlates. Am J Dis Child 1966; 111: 581–93.

1158

26

27 28 29 30

31

32 33

34

35

36

37

38 39 40

41

42

43

44

45

46

47

48

49

Kain ZN, Wang S-M, Mayes LC, Krivutza DM, Teague BA. Sensory stimuli and anxiety in children undergoing surgery: a randomized, controlled trial. Anesth Analg 2001; 92: 897–903. Skeletal trauma in children, 2nd edn. Philadelphia, PA: WB Saunders Company, 1998. Fractures in children, 3rd edn. Philadelphia, PA: JB Lippincott Company, 1991. Skeletal injury in the child, 2nd edn. Philadelphia, PA: WB Saunders Company, 1990. Localio AR, Berlin JA, Have TRT, Kimmel SE. Adjustments for center in multicenter studies: an overview. Ann Intern Med 2001; 135: 112–23. Hollis S, Campbell F. What is meant by intention to treat analysis? Survey of published randomised controlled trials. BMJ 1999; 319: 670–74. Cook RJ, Sackett DL. The number needed to treat: a clinically useful measure of treatment effect. BMJ 1995; 310: 452–54. Griffin PP, Anderson M, Green WT. Fractures of the shaft of the femur in treatment: treatment and results. Orthop Clin North Am 1972; 3: 213–23. Wong J, Boyd R, Keenan NW, et al. Gait patterns after fracture of the femoral shaft in children, managed by external fixation or early hip spica cast. J Pediatr Orthop 2004; 24: 463–71. Hennrikus WL, Kasser JR, Rand F, Millis MB, Richards KM. The function of the quadriceps muscle after a fracture of the femur in patients who are less than seventeen years old. J Bone Joint Surg 1993; 75A: 508–13. Hedin H, Larsson S. Muscle strength in children treated for displaced femoral fractures by external fixation: 31 patients compared with 31 matched controls. Acta Orthop Scand 2003; 74: 305–11. Eckhoff DG, Kramer RC, Alongi CA, VanGerven DP. Femoral anteversion and arthritis of the knee. J Pediatr Orthoped 1994; 14: 608–10. Verbeek HOF, Bender J, Sawdis K. Rotation of the femur after fractures of the femoral shaft in children. Injury 1976; 8: 43–48. Weber BG. Fractures of the femoral shaft in children. Injury 1969; 1: 65–68. Ioannidis JPA, Haidich A-B, Pappa M, et al. Comparison of evidence of treatment effects in randomized and nonrandomized studies. JAMA 2001; 286: 821–30. Thompson JD, Buehler KC, Sponseller PD, et al. Shortening in femoral shaft fractures in children treatment with spica cast. Clin Orthop 1997; 338: 74–78. Loder RT, Warschausky S, Schwartz EM, Hensinger RN, Greenfield ML. The psychosocial characteristics of children with fractures. J Pediatr Orthoped 1995; 15: 41–46. Campbell L, Clark M, Kirkpatrick SE. Stress management training for parents and their children undergoing cardiac catheterization. Am J Orthopsychiatry 1986; 56: 234–43. Kain ZN, Mayes LC, Wang S-M, Hofstadter MB. Postoperative behavioral outcomes in children: effects of sedative premedication. Anesthesiology 1999; 90: 758–65. Rennick JE, Johnston CC, Dougherty G, Platt R, Ritchie JA. Children’s psychological responses after critical illness and exposure to invasive technology. J Dev Behav Pediatr 2002; 23: 133–44. Ligier JN, Metaizeau JP, Prevot J, Lascombes P. Elastic stable intramedullary nailing of femoral shaft fractures in children. J Bone Joint Surg 1988; 70B: 74–77. Flynn JM, Hresko T, Reynolds RA, Blasier RD, Davidson R, Kasser J. Titanium elastic nails for pediatric femur fractures: a multicenter study of early results with analysis of complications. J Pediatr Orthoped 2001; 21: 4–8. Rohde RS, Mendelson SA, Grudziak JS. Acute synovitis of the knee resulting from intra-articular knee penetration as a complication of flexible intramedullary nailing of pediatric femur fractures: report of two cases. J Pediatr Orthoped 2003; 23: 635–38. Luhmann SJ, Schootman M, Schoenecker PL, Dobbs MB, Gordon JE. Complications of titanium elastic nails for pediatric femoral shaft fractures. J Pediatr Orthoped 2003; 23: 443–47.

www.thelancet.com Vol 365 March 26, 2005