- Email: [email protected]
Health outcomes of delayed union and nonunion of femoral and tibial shaft fractures
Injury, Int. J. Care Injured 45 (2014) 1653–1658
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Health outcomes of delayed union and nonunion of femoral and tibial shaft fractures Wei-Han Tay a, Richard de Steiger b, Martin Richardson b, Russell Gruen c, Zsolt J. Balogh d,* a
Department of Orthopaedic Surgery, John Hunter Hospital, New South Wales, Australia Department of Surgery, Epworth HealthCare, University of Melbourne, Victoria, Australia c Department of Surgery, National Trauma Research Institute, The Alfred, Monash University, Victoria, Australia d Department of Traumatology, John Hunter Hospital, University of Newcastle, New South Wales, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 26 June 2014
Introduction: Knowledge about the functional consequences of lower limb long bone fractures is helpful to inform patients, clinicians and employers about their recovery process and prognosis. This study aims to describe the epidemiology and health outcomes of femoral and tibial shaft fractures treated at two level I trauma centres, by comparing the differences between patients with delayed union or nonunion and patients with union. Patients and methods: An analysis of registry data over two years, supplemented with medical record review, was conducted. Fracture healing was retrospectively assessed by clinical and radiological evidence of union, and the need for surgical intervention. SF-12 scores, and work and pain status were prospectively recorded at six and twelve months post injury. Results: 285 fractures progressed to union and 138 fractures developed delayed union or nonunion. There was a significant difference between the two cohorts with regards to the mechanism of injury, association with multi-trauma, open fractures, grade of Gustilo classification, patient fund source, smoking status and presence of comorbidities. The SF-12 physical component score was less than 50 at both six and twelve months with improvement in the union group, but not in the delayed union or nonunion group. 72% of patients with union had returned to work at one year, but 54% continued to have pain. The difference compared to patients with delayed union or nonunion was significant. Discussion: Even patients whose fractures unite in the expectant time-frame will have residual physical disability. Patients with delayed union or nonunion have still poorer outcomes, including ongoing problems with returning to work and pain. It is important to educate patients about their injury so that they have realistic expectations. This is particularly relevant given that the patients most likely to sustain femoral or tibial shaft fractures are working-age healthy adults, and up to a third of fractures may develop delayed union or nonunion. Conclusion: Despite modern treatment, the patient-reported outcomes of lower limb long bone shaft fractures do not return to normal at one year. Patients with delayed union or nonunion can expect poorer outcomes. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Trauma Outcomes Femur Tibia Fracture Nonunion Delayed union
Introduction Femoral and tibial shaft fractures are major limb injuries that can lead to significant physical impairment [1]. Because they usually result from high-energy trauma, delayed union and nonunion are common occurrences of these fractures [2].
* Corresponding author at: Royal Newcastle Centre, John Hunter Hospital, Lookout Road, New Lambton, New South Wales 2305, Australia. Tel.: +61 2 49214259. E-mail address: [email protected] (Z.J. Balogh). http://dx.doi.org/10.1016/j.injury.2014.06.025 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Fracture nonunion is a chronic condition associated with pain, and functional and psychosocial disability [3] that has been shown to have a greater negative impact on the quality of life than patients with end-stage congestive heart failure, and patients receiving renal dialysis or chemotherapy for cancer [4,5]. In addition to the considerable economic costs to society [6], delayed fracture healing and multiple surgeries cause patients to have ongoing pain, swelling, stiffness and inability to bear weight in their limb [7]. The assessment of fracture healing using traditional parameters, such as time to union or range of motion for adjacent joints,
1654
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
does not necessarily correlate with the patient’s experience of their injury or recovery [8]. While patient-reported outcomes have now been studied for several orthopaedic conditions [9–11], review of the literature confirms a paucity of comparable data on fracture nonunion. In particular, a direct comparison of the health status of patients with delayed union or nonunion and with normal fracture healing has yet to be documented. The exact determination of the burden of delayed union and nonunion of lower extremity long bone fractures is important to inform patients and their families, clinicians, employers and insurers about their recovery process and functional prognosis. This could also guide the design of more focused research in this area and provide relevant information to potential funding bodies. The aim of this study is to describe the epidemiology and health outcomes of femoral and tibial shaft fractures treated at two level I trauma centres, by comparing the differences between patients with delayed union or nonunion and patients with union. Patients and methods A retrospective analysis of prospective registry data, supplemented with patient medical record review, was conducted. Patients were selected from the Victorian Orthopaedic Trauma Outcomes Registry (VOTOR) database with approval of the respective hospital ethics committees. VOTOR includes all patients with orthopaedic injuries admitted to the two level I adult trauma centres in Victoria, Australia [12]. Participants were recruited into VOTOR using an opt-off method of consent, whereby all eligible patients are automatically registered upon admission and given the option to withdraw from the registry at any time [12]. Patients were enrolled in the study from August 2003 to August 2004 and again from February 2005 to July 2006 [12]. The five month gap between the two sample collection periods was due to temporary funding issues with the registry. Femoral and tibial shaft fractures were identified by reviewing the VOTOR database for AO fracture types 32 and 42, respectively. Exclusion criteria included patients who sustained a pathological fracture or whose injury was managed by a non-orthopaedic team [12], fractures with joint involvement, and fractures in a subsequently amputated limb or in a patient deceased less than six weeks post injury. Fig. 1 illustrates the patient inclusion process. Fracture healing was assessed by clinical and radiological evidence of union. The indication and timing for surgical intervention was determined by reviewing hospital medical records. Evidence of clinical union included the absence of tenderness at the fracture site and ability to bear full weight on that limb without pain [13]. A fracture was deemed to have evidence of radiological union by the presence of adequate callus bridging the fracture site and disappearance of the fracture line [13,14]. The earliest timing for surgery for delayed healing was set at six weeks post fracture based on common local practice and included nail dynamisation. A fracture was considered to have progressed to nonunion if it required surgical intervention at greater than six months after injury, given that the majority of united fractures should have achieved the aforementioned clinical and radiological milestones. The outcome of a fracture was regarded as unknown if follow-up was for less than twelve weeks or if their clinical and radiological progress were unclear in the patient’s records. Health, work and pain status were prospectively recorded at six and twelve months post injury as part of the VOTOR follow-up protocol [12]. The health status of patients was assessed using the SF-12 [12], which has been tested for reliability [15–18] and validated for use within the Australian population [19] and in trauma patients [20]. Responses from the twelve questions in the self-reported survey are used to calculate a Physical Component
Total patients enlisted during VOTOR data collection periods: 4,448 patients VOTOR database searched for AO fracture types 32 and 42 Excluded: 3,853 patients No femoral or tibial shaft fracture
665 fractures in 595 patients Femur: 300 fractures in 287 patients Tibia: 365 fractures in 351 patients Review of hospital medical records Excluded: 106 patients Non-diaphyseal: 9 fractures Intra-articular: 60 fractures Limb amputated: 8 fractures Patient deceased: 7 fractures Incorrectly recorded: 6 fractures Medical record missing: 33 fractures
542 fractures in 489 patients Femur: 260 fractures in 248 patients Tibia: 282 fractures in 272 patients Fig. 1. Flowchart of patient inclusion process.
Summary (PCS) and a Mental Component Summary (MCS) score which correspond to the physical and mental health status of respondents, respectively [18,21]. Each component is scored between 0 and 100, with higher scores indicating better health [18,21] and a total greater than 50 representing no disability [22]. Patients were also asked whether they had returned to any form of work and whether they still had any pain from their injury [12]. Participants were interviewed by a trained research nurse [12]. Four contact attempts were made before a patient was considered lost to follow-up [12]. Patient demographics and injury details including age, gender, fund source, mechanism of injury, open or closed fracture, and the Gustilo classification for open fractures were retrieved from the VOTOR database. Hospital records were reviewed to determine other injuries and the cigarette smoking status of patients. To categorise other injuries, patients were identified as having sustained an isolated injury or multi-trauma (other than ipsilateral extra-articular fractures of the fibula associated with fractures of the tibia, minor abrasions or lacerations, or loss of consciousness for less than 30 minutes without neurological sequelae) with and without head injury. Head injury was defined as having evidence of traumatic cerebral oedema, diffuse or focal brain injury, or intracranial haemorrhage. In addition, the Charlson Comorbidity Index [23,24] (CCI) was calculated for each patient using the ICD-10-AM diagnosis codes assigned for that admission. This was utilised as an aggregate measure of a patient’s general health. The data was analysed using the statistical software package, Stata [25]. The p values for the categorical variables were calculated using Pearson’s chi-squared and Fisher’s exact twotailed tests. The SF-12 PCS and MCS scores were analysed by linear regression modelling to calculate a correlation coefficient. The return to work and pain status results were analysed by logistic regression modelling to calculate a risk ratio. Multi-variate analysis was employed to account for other variables.
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
1655
Table 2 Patient demographics.
Total fractures included for outcome review: 542 fractures
Union (%), n = 285
Review of hospital medical records
Age 251 (88) <65 years Gender Male 202 (71) Fund source Compensable 205 (72) Cigarette smokinga Smoker 95 (41) Charlson Comorbidity Index 1, 2, 3 or 6 59 (21)
Excluded: 119 fractures (22%) Malunion: 4 fractures Unknown: 115 fractures
Total fractures included for outcome analysis: 423 fractures Union: 285 fractures (53%) Delayed union or nonunion: 138 fractures (25%)
a
DU/NU (%), n = 138
p value
127 (92)
0.216
107 (78)
0.148
115 (83)
0.010
66 (53)
0.006
45 (33)
0.011
Smoking status documented: Union n = 231, DU/NU n = 125.
Fig. 2. Flowchart of fracture outcome review.
The majority of fractures occurred in healthy patients. 21% of patients with union had one or more comorbidities compared to 33% of patients with delayed union or nonunion (p value 0.011).
Results 423 fractures (78%) had a known healing endpoint that was eligible for assessment of their patient-reported outcomes. From this sample, 285 fractures (53%) progressed to union and 138 fractures (25%) developed delayed union or nonunion. 119 fractures (22%) developed malunion or had an unknown outcome and were excluded from further analysis. The outcomes of the fractures selected for evaluation are summarised in Fig. 2. The patient follow-up rates of each health outcome measured for the two study groups are shown in Table 1. The lowest response rate for the SF-12 was 72% for patients with delayed union or nonunion at six months following fracture. At least 65% of patients reported whether they had returned to work in both groups at six months post injury. The lowest follow-up rate for pain status was 70% for patients with delayed union or nonunion at the six months interview. Patient demographics The demographic distribution of patients in the union group compared to patients in the delayed union or nonunion group is shown in Table 2. The majority of fractures were sustained by male patients less than 65 years old. There was no significant difference in the age (p value 0.216) or gender (p value 0.148) allocation between patients in either study cohort. The larger proportion of injuries was compensable. This consisted of 72% of fractures in the union group compared to 83% of fractures that required surgical intervention for delayed union or nonunion (p value 0.010). The smoking status of patients was documented for 356 fractures. 41% of patients in the union group were smokers compared to 53% of patients who progressed to delayed union or nonunion (p value 0.006).
Injury profile The injury profile for the two fracture healing outcome groups is shown in Table 3. Most fractures were the result of a road traffic accident. 73% of fractures with union were caused by a road vehicle injury compared to 86% of fractures with delayed union or nonunion (p value 0.007). The larger proportion of injuries resulted from multi-trauma. 46% and 15% of fractures that united were part of multi-trauma with and without head injury, respectively. In comparison, 59% and 7% of fractures that required surgical intervention for delayed union or nonunion belonged to patients who sustained multitrauma with and without head injury, respectively (p value 0.015). 30% of fractures in the union cohort were open injuries compared to 43% of fractures with delayed union or nonunion (p value 0.006). The Gustilo classification was recorded for 110 out of 145 open fractures. 27% of fractures with normal union were grade III compared to 63% of fractures with delayed union or nonunion (p value 0.000). SF-12 physical and mental health The SF-12 PCS and MCS median scores at six and twelve months post injury for patients with union and with delayed union or nonunion are shown in Table 4, together with their respective correlation coefficient values unadjusted and adjusted for other variables. The PCS median scores were less than 50 in both outcome groups at the two follow-up periods. There was a noticeable improvement in PCS scores from six months to twelve months following injury in the union cohort. In contrast, the MCS median scores were greater than 50 at both follow-up interviews for the
Table 1 Follow-up rates (%) of health outcomes. Health outcome 6 months SF-12 PCS SF-12 MCS RTW Pain 12 months SF-12 PCS SF-12 MCS RTW Pain
Union, n = 285
DU/NU, n = 138
76 76 65 75
72 72 65 70
74 74 71 74
77 77 76 77
DU/NU = delayed union or nonunion; PCS = Physical Component Summary; MCS = Mental Component Summary; RTW = return to work.
Table 3 Injury profile.
Mechanism Road traffic accident Injury Multi-trauma without head injury Multi-trauma with head injury Wound Open Gustilo classificationa III a
Union (%), n = 285
DU/NU (%), n = 138
p value
208 (73)
119 (86)
0.007
132 (46) 43 (15)
82 (59) 10 (7)
0.015
85 (30)
60 (43)
0.006
17 (27)
30 (63)
0.000
Gustilo classification documented: Union n = 62, DU/NU n = 48.
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
1656
Table 4 Linear regression of effect of DU/NU on SF-12 median scores.
b
6 PCS 6 MCS 12c PCS 12 MCS a b c
Union (IQR), n = 285
DU/NU (IQR), n = 138
37 56 44 56
31 52 32 51
(16) (15) (22) (13)
(10) (24) (18) (19)
Unadjusted coefficient (95% CI) 6.67 4.60 6.96 3.00
( ( ( (
9.40, 8.26, 9.65, 5.59,
3.93) 0.95) 4.26) 0.41)
Adjusteda coefficient (95% CI) 7.09 5.38 7.33 3.25
( ( ( (
9.79, 8.93, 9.89, 5.78,
4.39) 1.83) 4.76) 0.72)
For age, gender and multiple injuries. Six months post injury. Twelve months post injury.
Table 5 Logistic regression of effect of DU/NU on return to work and pain.
6 RTW 6 Pain 12 RTW 12 Pain a
Union (%), n = 285
DU/NU (%), n = 138
Unadjusted risk ratio (95% CI)
Adjusteda risk ratio (95% CI)
97/186 (52) 146/215 (68) 145/202 (72) 114/212 (54)
40/90 (44) 71/96 (74) 62/105 (59) 76/106 (72)
0.85 1.09 0.82 1.33
0.82 1.11 0.76 1.37
(0.62, (0.92, (0.65, (1.13,
1.09) 1.22) 0.98) 1.50)
(0.59, (0.94, (0.57, (1.16,
1.08) 1.24) 0.94) 1.54)
For age, gender and multiple injuries.
two study arms and did not significantly change from six to twelve months post fracture. Patients in the union group scored higher in all categories compared to patients with delayed union or nonunion. These differences were significant (95% confidence) unadjusted and adjusted for age, gender and multiple injuries. Return to work and pain The return to work and pain status at six and twelve months post injury of patients with fracture union compared to patients with impaired bone healing is shown in Table 5, together with their respective risk ratios unadjusted and adjusted for other variables. This analysis did not include 34 patients in the union group and 11 patients in the delayed union or nonunion group, who reported that they were not working prior to their fracture. At six months post injury, 52% of patients in the union group had returned to work, which increased to 72% by twelve months. In comparison, 44% of subjects with delayed union or nonunion had resumed employment at six months. This improved to 59% at the final interview. The difference between the two groups at twelve months was significant (95% confidence) with a risk ratio of 0.82 and 0.76 for patients with delayed union or nonunion to return to work, unadjusted and adjusted for age, gender and multiples injuries, respectively. Patients with union and with delayed union or nonunion continued to complain of pain in 68% and 74% of cases, respectively, at six months following fracture. At twelve months, 54% of patients with union had ongoing pain, while this was reported in 72% of patients with delayed union or nonunion. This difference was significant (95% confidence) at twelve months with a risk ratio of 1.33 and 1.37 for patients with delayed union or nonunion to complain of pain, unadjusted and adjusted for age, gender and multiple injuries, respectively. Discussion Despite modern advances in intramedullary fixation and established treatment practices, the health outcomes of patients with lower limb long bone shaft fractures do not return to normal. Even patients who achieve union of their fracture can expect to have residual physical disability during the first year post injury. In addition, just over a half of individuals will have returned to some form of work at six months, and a similar proportion of patients will continue to have pain at twelve months. Not surprisingly,
patients with delayed union or nonunion of their femoral and tibial shaft fractures fare even worse. They can expect greater physical impairment than if their fracture had healed without complication during their first year post injury. These patients are similarly more likely to remain unemployed and have pain at twelve months. This is of particular concern given that nearly a third of lower limb long bone diaphyseal injuries treated at a tertiary trauma centre will develop delayed union or nonunion. This incidence is comparable to other Australian study cohorts [26,27]. Reasons for the poorer health status of patients with delayed or nonunion are probably multi-factorial. They include the direct debilitating effects from a painful un-united fracture, and the further recovery required from multiple surgeries and extended rehabilitation. Frustration and personal loss may also arise from the disruption to the patient’s life while undergoing prolonged treatment and recuperation, which can adversely affect their relationships and career. This can be demoralising and difficult for the patient and their family to comprehend, especially when there may be a lay perception of their orthopaedic injury as simply a ‘‘broken bone’’ that is acute, ‘‘fixable’’ and without sequelae. Interestingly, the SF-12 did not demonstrate any disability resulting from the mental health of patients with lower extremity long bone diaphyseal fractures. These findings are consistent with a previous outcomes study of isolated tibial shaft fractures [28], and infer that the preliminary effects of these injuries during the recovery phase are primarily physical. The patient and injury profile of our sample population were consistent with the literature. Patients who were smokers or had medical comorbidities demonstrated a greater proportion of fracture healing problems [1,2]. High-energy trauma such as from road traffic accidents, open injuries and Gustilo grade III fractures likewise recorded higher rates of delayed union and nonunion [1,2]. In keeping with the abundant callus healing response commonly seen post brain injury [29], patients with associated head trauma showed the lowest rate of delayed union and nonunion. Perhaps as a predictor for surgical intervention, patients whose treatments were compensable were also more likely to be reoperated on for their delayed union or nonunion. Potential explanations for this trend include easier access to surgery, since patients may be referred to private hospitals for their operation, and greater financial incentive for the surgeon to operate. There are few trauma registries in the world that monitor the complications of fractures or patient-reported outcomes. Our study’s strength was the use of prospectively-collected data, which minimises bias from retrospective review and allows for follow-up
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
at uniform time-points. In addition, our sample data was obtained from the two major trauma centres in the state, which can be considered representative of the wider population and the accepted standard of care in Australia. Our study’s main weakness was the retrospective identification of the healing outcomes of fractures by reviewing hospital medical records. Because of the lack of consensus in the assessment of fracture healing among surgeons [30] and to minimise the need for subjective interpretation of patients’ notes, surgical intervention was used as the major criterion for establishing a fracture with delayed union or nonunion. However, the decision-making for surgery may have biased our analysis. For example, although the SPRINT trial showed no effect of dynamisation on union for tibial fractures [31], this was published after our study recruitment and surgeons may have regularly dynamised the nail after six weeks post injury. The authors were blinded to all patient-reported outcomes during the fracture union selection process. In our study, patients with delayed union and nonunion, femoral and tibial fractures, and isolated and multiple injuries were considered together. This limits this paper’s comparability with other data and can lead to bias from confounding factors. Patients with delayed union and nonunion were analysed collectively, since both conditions derive from the same disease process and only differ in temporal relation. Given their similar aetiology for fracture and function as the two weight-bearing long bones in the lower limbs, the femur and tibia were also categorised together. To reduce confounding, comparison with a larger control group using multi-variate analysis to adjust for other variables, including age, gender and multiple injuries, was employed. Non-responder bias due to loss to follow-up is a further inevitable limitation of this study. The response rate of participants were maximised by a thorough follow-up protocol, which included multiple contact attempts via both telephone and post [12]. The follow-up rates for this project compared favourably to the overall rates for VOTOR and other prospective orthopaedic trauma registries [20,32,33]. It is essential that patients with lower extremity long bone shaft fractures are informed about the consequences of their injury, including delayed fracture healing and prolonged recovery. This will help patients to appreciate the severity and potential chronicity of their condition, and in turn motivate their compliance with therapy. The findings from this study may be used to educate patients’ expectations regarding their physical recovery and its practical implications like returning to work. This is particularly relevant given that the patients most likely to sustain femoral or tibial shaft fractures are working-age and previously healthy adults, and would also provide invaluable information to employers, insurers and other stakeholders. At the same time, surgeons and physicians can utilise this prognostic data on patients’ functional outcomes as a basis for future research in treatment methods and rehabilitation programmes. Surgeons should also be mindful to allow increased time for tibial shaft fractures to heal before intervening [31], as repeated surgeries may cause harm and paradoxically slow the patient’s recovery. In contrast, more aggressive surgical intervention may be necessary for patients at increased risk of nonunion. For example, open fractures sustained from high-energy trauma with segmental bone loss may require early and staged bone grafting. Rehabilitation strategies should include vocation retraining and optimising pain management to assist patients to return to their premorbid function.
1657
fractures unite in the expectant time-frame will have residual physical disability. Three out of four patients will have returned to work at twelve months post injury, but more than half of individuals will continue to have pain. Patients with delayed union or nonunion can expect poorer outcomes. Conflict of interest statement The authors of this paper certify that they have no affiliations with or involvement in any organisation or entity with any financial or personal interest in the subject matter or materials discussed in this manuscript. Acknowledgements 1. Victorian Orthopaedic Trauma Outcomes Registry (VOTOR)* Steering Committee and follow-up staff, 2. Department of Orthopaedic Surgery, Royal Melbourne Hospital (Victoria, Australia), and 3. Department of Orthopaedic Surgery, The Alfred (Victoria, Australia). References
Conclusion
[1] Rudloff MI. Fractures of the lower extremity. In: Canale ST, Beaty JH, editors. 12th ed., Canale & Beaty: Campbell’s operative orthopaedics, vol. III, 12th ed. Philadelphia: Mosby; 2013. Ch. 54. [2] Cleveland KB. Delayed union and nonunion of fractures. In: Canale ST, Beaty JH, editors. 12th ed., Canale & Beaty: Campbell’s operative orthopaedics, vol. III, 12th ed. Philadelphia: Mosby; 2013. Ch. 59. [3] Lerner RK, Esterhai Jr JL, Polomano RC, Cheatle MD, Heppenstall RB. Quality of life assessment of patients with posttraumatic fracture nonunion, chronic refractory osteomyelitis, and lower-extremity amputation. Clin Orthopaed Relat Res 1993;295:28–36. [4] Ware Jr JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30(6):473–83. [5] Ware Jr JE, Snow KK, Kosinski M, Gandek B. The SF-36 health survey: manual and interpretation guide. Boston: The Health Institute; 1993. [6] Kanakaris NK, Giannoudis PV. The health economics of the treatment of long-bone non-unions. Injury 2007;38(S2):77–84. [7] Babhulkar S, Pande K, Babhulkar S. Nonunion of the diaphysis of long bones. Clin Orthopaed Relat Res 2005;431:50–6. [8] Zlowodzki M, Obremskey WT, Thomison JB, Kregor PJ. Functional outcome after treatment of lower-extremity nonunions. J Trauma 2005;58(2):312–7. [9] Fanuele JC, Birkmeyer NJ, Abdu WA, Tosteson TD, Weinstein JN. The impact of spinal problems on the health status of patients. Have we underestimated the effect? Spine 2000;25(12):1509–14. [10] Kiebzak GM, Vain PA, Gregory AM, Mokris JG, Mauerhan DR. SF-36 general health status survey to determine patient satisfaction at short-term followup after total hip and knee arthroplasty. J South Orthop Assoc 1997;6(3):169–72. [11] Oliver CW, Twaddle B, Agel J, Routt Jr ML. Outcome after pelvic ring fractures: evaluation using the Medical Outcomes Short Form SF-36. Injury 1996;27(9): 635–641. [12] Victorian Trauma Foundation. Victorian Orthopaedic Trauma Outcomes Registry: evaluation and monitoring of orthopaedic trauma management across Victorian hospitals. Melbourne: Monash University; 2006. [13] Morshed S, Corrales L, Genant H, Miclau III T. Outcome assessment in clinical trials of fracture-healing. J Bone Joint Surg 2008;90(S1):62–7 (American). [14] Whelan DB, Bhandari M, McKee MD, Guyatt GH, Kreder HJ, Stephen D, et al. Interobserver and intraobserver variation in the assessment of the healing of tibial fractures after intramedullary fixation. J Bone Joint Surg 2002;84B(1): 15–18 (British). [15] Lenert LA. The reliability and internal consistency of an internet-capable computer program for measuring utilities. Qual Life Res 2000;9(7):811–7. [16] Resnick B, Parker R. Simplified scoring and psychometrics of the revised 12-Item Short-Form Health Survey. Outcomes Manag Nurs Pract 2001;5(4): 161–166. [17] Salyers MP, Bosworth HB, Swanson JW, Lamb-Pagone J, Osher FC. Reliability and validity of the SF-12 health survey among people with severe mental illness. Med Care 2000;38(11):1141–50. [18] Ware Jr JE, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34(3):220–33. [19] Sanderson K, Andrews G. The SF-12 in the Australian population: crossvalidation of item selection. Aust N Z J Public Health 2002;26(4):343–5.
The patient-reported outcomes of lower limb long bone shaft fractures do not return to normal at one year. Even patients whose
* Funded by the Transport Accident Commission (TAC) through the Institute for Safety, Compensation and Recovery Research (ISCRR).
1658
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
[20] Kiely JM, Brasel KJ, Guse CE, Weigelt JA. Correlation of SF-12 and SF-36 in a trauma population. J Surg Res 2006;132(2):214–8. [21] Ware Jr JE, Kosinski M, Keller SD. SF-12: How to score the SF-12 Physical and Mental Health Summary scales. 3rd ed. Lincoln: QualityMetric; 1998 . [22] Avery J, Dal Grande E, Taylor A. Quality of life in South Australia as measured by the SF12 health status questionnaire: population norms for 2003, trends from 1997–2003. Adelaide: South Australian Department of Human Services; 2004. [23] Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47(11):1245–51. [24] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40(5):373–83. [25] Gould B. Stata Release 13. College Station, TX, USA: StataCorp LP; 2013. [26] Harris I, Hatfield A, Donald G, Walton J. Outcome after intramedullary nailing of femoral shaft fractures. Aust N Z J Surg 2003;73:387–9. [27] Harris I, Lyons M. Reoperation rate in diaphyseal tibia fractures. Aust N Z J Surg 2005;75:1041–4.
[28] Dogra AS, Ruiz AL, Marsh DR. Late outcome of isolated tibial fractures treated by intramedullary nailing: the correlation between disease-specific and generic outcome measures. J Orthopaed Trauma 2002;16(4):245–9. [29] Spencer RF. The effect of head injury on fracture healing. A quantitative assessment. J Bone Joint Surg 1987;69B(4):525–8 (British). [30] Bhandari M, Guyatt GH, Swiontkowski MF, Tornetta III P, Sprague S, Schemitsch EH. A lack of consensus in the assessment of fracture healing among orthopaedic surgeons. J Orthopaed Trauma 2002;16(8):562–6. [31] Bhandari M, Guyatt GH, Tornetta III P, Schemitsch EH, Swiontkowski MF, Sanders D, et al. Randomized trial of reamed and unreamed intramedullary nailing of tibial shaft fractures. J Bone Joint Surg 2008;90:2567–78 (American). [32] Skoog A, So¨derqvist A, To¨rnkvist H, Ponzer S. One-year outcome after tibial shaft fractures: results of a prospective fracture registry. J Orthop Trauma 2001;15(3):210–5. [33] Swiontkowski MF, Agel J, McAndrew MP, Burgess AR, MacKenzie EJ. Outcome validation of the AO/OTA fracture classification system. J Orthopaed Trauma 2000;14(8):534–41.
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Health outcomes of delayed union and nonunion of femoral and tibial shaft fractures Wei-Han Tay a, Richard de Steiger b, Martin Richardson b, Russell Gruen c, Zsolt J. Balogh d,* a
Department of Orthopaedic Surgery, John Hunter Hospital, New South Wales, Australia Department of Surgery, Epworth HealthCare, University of Melbourne, Victoria, Australia c Department of Surgery, National Trauma Research Institute, The Alfred, Monash University, Victoria, Australia d Department of Traumatology, John Hunter Hospital, University of Newcastle, New South Wales, Australia b
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 26 June 2014
Introduction: Knowledge about the functional consequences of lower limb long bone fractures is helpful to inform patients, clinicians and employers about their recovery process and prognosis. This study aims to describe the epidemiology and health outcomes of femoral and tibial shaft fractures treated at two level I trauma centres, by comparing the differences between patients with delayed union or nonunion and patients with union. Patients and methods: An analysis of registry data over two years, supplemented with medical record review, was conducted. Fracture healing was retrospectively assessed by clinical and radiological evidence of union, and the need for surgical intervention. SF-12 scores, and work and pain status were prospectively recorded at six and twelve months post injury. Results: 285 fractures progressed to union and 138 fractures developed delayed union or nonunion. There was a significant difference between the two cohorts with regards to the mechanism of injury, association with multi-trauma, open fractures, grade of Gustilo classification, patient fund source, smoking status and presence of comorbidities. The SF-12 physical component score was less than 50 at both six and twelve months with improvement in the union group, but not in the delayed union or nonunion group. 72% of patients with union had returned to work at one year, but 54% continued to have pain. The difference compared to patients with delayed union or nonunion was significant. Discussion: Even patients whose fractures unite in the expectant time-frame will have residual physical disability. Patients with delayed union or nonunion have still poorer outcomes, including ongoing problems with returning to work and pain. It is important to educate patients about their injury so that they have realistic expectations. This is particularly relevant given that the patients most likely to sustain femoral or tibial shaft fractures are working-age healthy adults, and up to a third of fractures may develop delayed union or nonunion. Conclusion: Despite modern treatment, the patient-reported outcomes of lower limb long bone shaft fractures do not return to normal at one year. Patients with delayed union or nonunion can expect poorer outcomes. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Trauma Outcomes Femur Tibia Fracture Nonunion Delayed union
Introduction Femoral and tibial shaft fractures are major limb injuries that can lead to significant physical impairment [1]. Because they usually result from high-energy trauma, delayed union and nonunion are common occurrences of these fractures [2].
* Corresponding author at: Royal Newcastle Centre, John Hunter Hospital, Lookout Road, New Lambton, New South Wales 2305, Australia. Tel.: +61 2 49214259. E-mail address: [email protected] (Z.J. Balogh). http://dx.doi.org/10.1016/j.injury.2014.06.025 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Fracture nonunion is a chronic condition associated with pain, and functional and psychosocial disability [3] that has been shown to have a greater negative impact on the quality of life than patients with end-stage congestive heart failure, and patients receiving renal dialysis or chemotherapy for cancer [4,5]. In addition to the considerable economic costs to society [6], delayed fracture healing and multiple surgeries cause patients to have ongoing pain, swelling, stiffness and inability to bear weight in their limb [7]. The assessment of fracture healing using traditional parameters, such as time to union or range of motion for adjacent joints,
1654
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
does not necessarily correlate with the patient’s experience of their injury or recovery [8]. While patient-reported outcomes have now been studied for several orthopaedic conditions [9–11], review of the literature confirms a paucity of comparable data on fracture nonunion. In particular, a direct comparison of the health status of patients with delayed union or nonunion and with normal fracture healing has yet to be documented. The exact determination of the burden of delayed union and nonunion of lower extremity long bone fractures is important to inform patients and their families, clinicians, employers and insurers about their recovery process and functional prognosis. This could also guide the design of more focused research in this area and provide relevant information to potential funding bodies. The aim of this study is to describe the epidemiology and health outcomes of femoral and tibial shaft fractures treated at two level I trauma centres, by comparing the differences between patients with delayed union or nonunion and patients with union. Patients and methods A retrospective analysis of prospective registry data, supplemented with patient medical record review, was conducted. Patients were selected from the Victorian Orthopaedic Trauma Outcomes Registry (VOTOR) database with approval of the respective hospital ethics committees. VOTOR includes all patients with orthopaedic injuries admitted to the two level I adult trauma centres in Victoria, Australia [12]. Participants were recruited into VOTOR using an opt-off method of consent, whereby all eligible patients are automatically registered upon admission and given the option to withdraw from the registry at any time [12]. Patients were enrolled in the study from August 2003 to August 2004 and again from February 2005 to July 2006 [12]. The five month gap between the two sample collection periods was due to temporary funding issues with the registry. Femoral and tibial shaft fractures were identified by reviewing the VOTOR database for AO fracture types 32 and 42, respectively. Exclusion criteria included patients who sustained a pathological fracture or whose injury was managed by a non-orthopaedic team [12], fractures with joint involvement, and fractures in a subsequently amputated limb or in a patient deceased less than six weeks post injury. Fig. 1 illustrates the patient inclusion process. Fracture healing was assessed by clinical and radiological evidence of union. The indication and timing for surgical intervention was determined by reviewing hospital medical records. Evidence of clinical union included the absence of tenderness at the fracture site and ability to bear full weight on that limb without pain [13]. A fracture was deemed to have evidence of radiological union by the presence of adequate callus bridging the fracture site and disappearance of the fracture line [13,14]. The earliest timing for surgery for delayed healing was set at six weeks post fracture based on common local practice and included nail dynamisation. A fracture was considered to have progressed to nonunion if it required surgical intervention at greater than six months after injury, given that the majority of united fractures should have achieved the aforementioned clinical and radiological milestones. The outcome of a fracture was regarded as unknown if follow-up was for less than twelve weeks or if their clinical and radiological progress were unclear in the patient’s records. Health, work and pain status were prospectively recorded at six and twelve months post injury as part of the VOTOR follow-up protocol [12]. The health status of patients was assessed using the SF-12 [12], which has been tested for reliability [15–18] and validated for use within the Australian population [19] and in trauma patients [20]. Responses from the twelve questions in the self-reported survey are used to calculate a Physical Component
Total patients enlisted during VOTOR data collection periods: 4,448 patients VOTOR database searched for AO fracture types 32 and 42 Excluded: 3,853 patients No femoral or tibial shaft fracture
665 fractures in 595 patients Femur: 300 fractures in 287 patients Tibia: 365 fractures in 351 patients Review of hospital medical records Excluded: 106 patients Non-diaphyseal: 9 fractures Intra-articular: 60 fractures Limb amputated: 8 fractures Patient deceased: 7 fractures Incorrectly recorded: 6 fractures Medical record missing: 33 fractures
542 fractures in 489 patients Femur: 260 fractures in 248 patients Tibia: 282 fractures in 272 patients Fig. 1. Flowchart of patient inclusion process.
Summary (PCS) and a Mental Component Summary (MCS) score which correspond to the physical and mental health status of respondents, respectively [18,21]. Each component is scored between 0 and 100, with higher scores indicating better health [18,21] and a total greater than 50 representing no disability [22]. Patients were also asked whether they had returned to any form of work and whether they still had any pain from their injury [12]. Participants were interviewed by a trained research nurse [12]. Four contact attempts were made before a patient was considered lost to follow-up [12]. Patient demographics and injury details including age, gender, fund source, mechanism of injury, open or closed fracture, and the Gustilo classification for open fractures were retrieved from the VOTOR database. Hospital records were reviewed to determine other injuries and the cigarette smoking status of patients. To categorise other injuries, patients were identified as having sustained an isolated injury or multi-trauma (other than ipsilateral extra-articular fractures of the fibula associated with fractures of the tibia, minor abrasions or lacerations, or loss of consciousness for less than 30 minutes without neurological sequelae) with and without head injury. Head injury was defined as having evidence of traumatic cerebral oedema, diffuse or focal brain injury, or intracranial haemorrhage. In addition, the Charlson Comorbidity Index [23,24] (CCI) was calculated for each patient using the ICD-10-AM diagnosis codes assigned for that admission. This was utilised as an aggregate measure of a patient’s general health. The data was analysed using the statistical software package, Stata [25]. The p values for the categorical variables were calculated using Pearson’s chi-squared and Fisher’s exact twotailed tests. The SF-12 PCS and MCS scores were analysed by linear regression modelling to calculate a correlation coefficient. The return to work and pain status results were analysed by logistic regression modelling to calculate a risk ratio. Multi-variate analysis was employed to account for other variables.
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
1655
Table 2 Patient demographics.
Total fractures included for outcome review: 542 fractures
Union (%), n = 285
Review of hospital medical records
Age 251 (88) <65 years Gender Male 202 (71) Fund source Compensable 205 (72) Cigarette smokinga Smoker 95 (41) Charlson Comorbidity Index 1, 2, 3 or 6 59 (21)
Excluded: 119 fractures (22%) Malunion: 4 fractures Unknown: 115 fractures
Total fractures included for outcome analysis: 423 fractures Union: 285 fractures (53%) Delayed union or nonunion: 138 fractures (25%)
a
DU/NU (%), n = 138
p value
127 (92)
0.216
107 (78)
0.148
115 (83)
0.010
66 (53)
0.006
45 (33)
0.011
Smoking status documented: Union n = 231, DU/NU n = 125.
Fig. 2. Flowchart of fracture outcome review.
The majority of fractures occurred in healthy patients. 21% of patients with union had one or more comorbidities compared to 33% of patients with delayed union or nonunion (p value 0.011).
Results 423 fractures (78%) had a known healing endpoint that was eligible for assessment of their patient-reported outcomes. From this sample, 285 fractures (53%) progressed to union and 138 fractures (25%) developed delayed union or nonunion. 119 fractures (22%) developed malunion or had an unknown outcome and were excluded from further analysis. The outcomes of the fractures selected for evaluation are summarised in Fig. 2. The patient follow-up rates of each health outcome measured for the two study groups are shown in Table 1. The lowest response rate for the SF-12 was 72% for patients with delayed union or nonunion at six months following fracture. At least 65% of patients reported whether they had returned to work in both groups at six months post injury. The lowest follow-up rate for pain status was 70% for patients with delayed union or nonunion at the six months interview. Patient demographics The demographic distribution of patients in the union group compared to patients in the delayed union or nonunion group is shown in Table 2. The majority of fractures were sustained by male patients less than 65 years old. There was no significant difference in the age (p value 0.216) or gender (p value 0.148) allocation between patients in either study cohort. The larger proportion of injuries was compensable. This consisted of 72% of fractures in the union group compared to 83% of fractures that required surgical intervention for delayed union or nonunion (p value 0.010). The smoking status of patients was documented for 356 fractures. 41% of patients in the union group were smokers compared to 53% of patients who progressed to delayed union or nonunion (p value 0.006).
Injury profile The injury profile for the two fracture healing outcome groups is shown in Table 3. Most fractures were the result of a road traffic accident. 73% of fractures with union were caused by a road vehicle injury compared to 86% of fractures with delayed union or nonunion (p value 0.007). The larger proportion of injuries resulted from multi-trauma. 46% and 15% of fractures that united were part of multi-trauma with and without head injury, respectively. In comparison, 59% and 7% of fractures that required surgical intervention for delayed union or nonunion belonged to patients who sustained multitrauma with and without head injury, respectively (p value 0.015). 30% of fractures in the union cohort were open injuries compared to 43% of fractures with delayed union or nonunion (p value 0.006). The Gustilo classification was recorded for 110 out of 145 open fractures. 27% of fractures with normal union were grade III compared to 63% of fractures with delayed union or nonunion (p value 0.000). SF-12 physical and mental health The SF-12 PCS and MCS median scores at six and twelve months post injury for patients with union and with delayed union or nonunion are shown in Table 4, together with their respective correlation coefficient values unadjusted and adjusted for other variables. The PCS median scores were less than 50 in both outcome groups at the two follow-up periods. There was a noticeable improvement in PCS scores from six months to twelve months following injury in the union cohort. In contrast, the MCS median scores were greater than 50 at both follow-up interviews for the
Table 1 Follow-up rates (%) of health outcomes. Health outcome 6 months SF-12 PCS SF-12 MCS RTW Pain 12 months SF-12 PCS SF-12 MCS RTW Pain
Union, n = 285
DU/NU, n = 138
76 76 65 75
72 72 65 70
74 74 71 74
77 77 76 77
DU/NU = delayed union or nonunion; PCS = Physical Component Summary; MCS = Mental Component Summary; RTW = return to work.
Table 3 Injury profile.
Mechanism Road traffic accident Injury Multi-trauma without head injury Multi-trauma with head injury Wound Open Gustilo classificationa III a
Union (%), n = 285
DU/NU (%), n = 138
p value
208 (73)
119 (86)
0.007
132 (46) 43 (15)
82 (59) 10 (7)
0.015
85 (30)
60 (43)
0.006
17 (27)
30 (63)
0.000
Gustilo classification documented: Union n = 62, DU/NU n = 48.
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
1656
Table 4 Linear regression of effect of DU/NU on SF-12 median scores.
b
6 PCS 6 MCS 12c PCS 12 MCS a b c
Union (IQR), n = 285
DU/NU (IQR), n = 138
37 56 44 56
31 52 32 51
(16) (15) (22) (13)
(10) (24) (18) (19)
Unadjusted coefficient (95% CI) 6.67 4.60 6.96 3.00
( ( ( (
9.40, 8.26, 9.65, 5.59,
3.93) 0.95) 4.26) 0.41)
Adjusteda coefficient (95% CI) 7.09 5.38 7.33 3.25
( ( ( (
9.79, 8.93, 9.89, 5.78,
4.39) 1.83) 4.76) 0.72)
For age, gender and multiple injuries. Six months post injury. Twelve months post injury.
Table 5 Logistic regression of effect of DU/NU on return to work and pain.
6 RTW 6 Pain 12 RTW 12 Pain a
Union (%), n = 285
DU/NU (%), n = 138
Unadjusted risk ratio (95% CI)
Adjusteda risk ratio (95% CI)
97/186 (52) 146/215 (68) 145/202 (72) 114/212 (54)
40/90 (44) 71/96 (74) 62/105 (59) 76/106 (72)
0.85 1.09 0.82 1.33
0.82 1.11 0.76 1.37
(0.62, (0.92, (0.65, (1.13,
1.09) 1.22) 0.98) 1.50)
(0.59, (0.94, (0.57, (1.16,
1.08) 1.24) 0.94) 1.54)
For age, gender and multiple injuries.
two study arms and did not significantly change from six to twelve months post fracture. Patients in the union group scored higher in all categories compared to patients with delayed union or nonunion. These differences were significant (95% confidence) unadjusted and adjusted for age, gender and multiple injuries. Return to work and pain The return to work and pain status at six and twelve months post injury of patients with fracture union compared to patients with impaired bone healing is shown in Table 5, together with their respective risk ratios unadjusted and adjusted for other variables. This analysis did not include 34 patients in the union group and 11 patients in the delayed union or nonunion group, who reported that they were not working prior to their fracture. At six months post injury, 52% of patients in the union group had returned to work, which increased to 72% by twelve months. In comparison, 44% of subjects with delayed union or nonunion had resumed employment at six months. This improved to 59% at the final interview. The difference between the two groups at twelve months was significant (95% confidence) with a risk ratio of 0.82 and 0.76 for patients with delayed union or nonunion to return to work, unadjusted and adjusted for age, gender and multiples injuries, respectively. Patients with union and with delayed union or nonunion continued to complain of pain in 68% and 74% of cases, respectively, at six months following fracture. At twelve months, 54% of patients with union had ongoing pain, while this was reported in 72% of patients with delayed union or nonunion. This difference was significant (95% confidence) at twelve months with a risk ratio of 1.33 and 1.37 for patients with delayed union or nonunion to complain of pain, unadjusted and adjusted for age, gender and multiple injuries, respectively. Discussion Despite modern advances in intramedullary fixation and established treatment practices, the health outcomes of patients with lower limb long bone shaft fractures do not return to normal. Even patients who achieve union of their fracture can expect to have residual physical disability during the first year post injury. In addition, just over a half of individuals will have returned to some form of work at six months, and a similar proportion of patients will continue to have pain at twelve months. Not surprisingly,
patients with delayed union or nonunion of their femoral and tibial shaft fractures fare even worse. They can expect greater physical impairment than if their fracture had healed without complication during their first year post injury. These patients are similarly more likely to remain unemployed and have pain at twelve months. This is of particular concern given that nearly a third of lower limb long bone diaphyseal injuries treated at a tertiary trauma centre will develop delayed union or nonunion. This incidence is comparable to other Australian study cohorts [26,27]. Reasons for the poorer health status of patients with delayed or nonunion are probably multi-factorial. They include the direct debilitating effects from a painful un-united fracture, and the further recovery required from multiple surgeries and extended rehabilitation. Frustration and personal loss may also arise from the disruption to the patient’s life while undergoing prolonged treatment and recuperation, which can adversely affect their relationships and career. This can be demoralising and difficult for the patient and their family to comprehend, especially when there may be a lay perception of their orthopaedic injury as simply a ‘‘broken bone’’ that is acute, ‘‘fixable’’ and without sequelae. Interestingly, the SF-12 did not demonstrate any disability resulting from the mental health of patients with lower extremity long bone diaphyseal fractures. These findings are consistent with a previous outcomes study of isolated tibial shaft fractures [28], and infer that the preliminary effects of these injuries during the recovery phase are primarily physical. The patient and injury profile of our sample population were consistent with the literature. Patients who were smokers or had medical comorbidities demonstrated a greater proportion of fracture healing problems [1,2]. High-energy trauma such as from road traffic accidents, open injuries and Gustilo grade III fractures likewise recorded higher rates of delayed union and nonunion [1,2]. In keeping with the abundant callus healing response commonly seen post brain injury [29], patients with associated head trauma showed the lowest rate of delayed union and nonunion. Perhaps as a predictor for surgical intervention, patients whose treatments were compensable were also more likely to be reoperated on for their delayed union or nonunion. Potential explanations for this trend include easier access to surgery, since patients may be referred to private hospitals for their operation, and greater financial incentive for the surgeon to operate. There are few trauma registries in the world that monitor the complications of fractures or patient-reported outcomes. Our study’s strength was the use of prospectively-collected data, which minimises bias from retrospective review and allows for follow-up
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
at uniform time-points. In addition, our sample data was obtained from the two major trauma centres in the state, which can be considered representative of the wider population and the accepted standard of care in Australia. Our study’s main weakness was the retrospective identification of the healing outcomes of fractures by reviewing hospital medical records. Because of the lack of consensus in the assessment of fracture healing among surgeons [30] and to minimise the need for subjective interpretation of patients’ notes, surgical intervention was used as the major criterion for establishing a fracture with delayed union or nonunion. However, the decision-making for surgery may have biased our analysis. For example, although the SPRINT trial showed no effect of dynamisation on union for tibial fractures [31], this was published after our study recruitment and surgeons may have regularly dynamised the nail after six weeks post injury. The authors were blinded to all patient-reported outcomes during the fracture union selection process. In our study, patients with delayed union and nonunion, femoral and tibial fractures, and isolated and multiple injuries were considered together. This limits this paper’s comparability with other data and can lead to bias from confounding factors. Patients with delayed union and nonunion were analysed collectively, since both conditions derive from the same disease process and only differ in temporal relation. Given their similar aetiology for fracture and function as the two weight-bearing long bones in the lower limbs, the femur and tibia were also categorised together. To reduce confounding, comparison with a larger control group using multi-variate analysis to adjust for other variables, including age, gender and multiple injuries, was employed. Non-responder bias due to loss to follow-up is a further inevitable limitation of this study. The response rate of participants were maximised by a thorough follow-up protocol, which included multiple contact attempts via both telephone and post [12]. The follow-up rates for this project compared favourably to the overall rates for VOTOR and other prospective orthopaedic trauma registries [20,32,33]. It is essential that patients with lower extremity long bone shaft fractures are informed about the consequences of their injury, including delayed fracture healing and prolonged recovery. This will help patients to appreciate the severity and potential chronicity of their condition, and in turn motivate their compliance with therapy. The findings from this study may be used to educate patients’ expectations regarding their physical recovery and its practical implications like returning to work. This is particularly relevant given that the patients most likely to sustain femoral or tibial shaft fractures are working-age and previously healthy adults, and would also provide invaluable information to employers, insurers and other stakeholders. At the same time, surgeons and physicians can utilise this prognostic data on patients’ functional outcomes as a basis for future research in treatment methods and rehabilitation programmes. Surgeons should also be mindful to allow increased time for tibial shaft fractures to heal before intervening [31], as repeated surgeries may cause harm and paradoxically slow the patient’s recovery. In contrast, more aggressive surgical intervention may be necessary for patients at increased risk of nonunion. For example, open fractures sustained from high-energy trauma with segmental bone loss may require early and staged bone grafting. Rehabilitation strategies should include vocation retraining and optimising pain management to assist patients to return to their premorbid function.
1657
fractures unite in the expectant time-frame will have residual physical disability. Three out of four patients will have returned to work at twelve months post injury, but more than half of individuals will continue to have pain. Patients with delayed union or nonunion can expect poorer outcomes. Conflict of interest statement The authors of this paper certify that they have no affiliations with or involvement in any organisation or entity with any financial or personal interest in the subject matter or materials discussed in this manuscript. Acknowledgements 1. Victorian Orthopaedic Trauma Outcomes Registry (VOTOR)* Steering Committee and follow-up staff, 2. Department of Orthopaedic Surgery, Royal Melbourne Hospital (Victoria, Australia), and 3. Department of Orthopaedic Surgery, The Alfred (Victoria, Australia). References
Conclusion
[1] Rudloff MI. Fractures of the lower extremity. In: Canale ST, Beaty JH, editors. 12th ed., Canale & Beaty: Campbell’s operative orthopaedics, vol. III, 12th ed. Philadelphia: Mosby; 2013. Ch. 54. [2] Cleveland KB. Delayed union and nonunion of fractures. In: Canale ST, Beaty JH, editors. 12th ed., Canale & Beaty: Campbell’s operative orthopaedics, vol. III, 12th ed. Philadelphia: Mosby; 2013. Ch. 59. [3] Lerner RK, Esterhai Jr JL, Polomano RC, Cheatle MD, Heppenstall RB. Quality of life assessment of patients with posttraumatic fracture nonunion, chronic refractory osteomyelitis, and lower-extremity amputation. Clin Orthopaed Relat Res 1993;295:28–36. [4] Ware Jr JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30(6):473–83. [5] Ware Jr JE, Snow KK, Kosinski M, Gandek B. The SF-36 health survey: manual and interpretation guide. Boston: The Health Institute; 1993. [6] Kanakaris NK, Giannoudis PV. The health economics of the treatment of long-bone non-unions. Injury 2007;38(S2):77–84. [7] Babhulkar S, Pande K, Babhulkar S. Nonunion of the diaphysis of long bones. Clin Orthopaed Relat Res 2005;431:50–6. [8] Zlowodzki M, Obremskey WT, Thomison JB, Kregor PJ. Functional outcome after treatment of lower-extremity nonunions. J Trauma 2005;58(2):312–7. [9] Fanuele JC, Birkmeyer NJ, Abdu WA, Tosteson TD, Weinstein JN. The impact of spinal problems on the health status of patients. Have we underestimated the effect? Spine 2000;25(12):1509–14. [10] Kiebzak GM, Vain PA, Gregory AM, Mokris JG, Mauerhan DR. SF-36 general health status survey to determine patient satisfaction at short-term followup after total hip and knee arthroplasty. J South Orthop Assoc 1997;6(3):169–72. [11] Oliver CW, Twaddle B, Agel J, Routt Jr ML. Outcome after pelvic ring fractures: evaluation using the Medical Outcomes Short Form SF-36. Injury 1996;27(9): 635–641. [12] Victorian Trauma Foundation. Victorian Orthopaedic Trauma Outcomes Registry: evaluation and monitoring of orthopaedic trauma management across Victorian hospitals. Melbourne: Monash University; 2006. [13] Morshed S, Corrales L, Genant H, Miclau III T. Outcome assessment in clinical trials of fracture-healing. J Bone Joint Surg 2008;90(S1):62–7 (American). [14] Whelan DB, Bhandari M, McKee MD, Guyatt GH, Kreder HJ, Stephen D, et al. Interobserver and intraobserver variation in the assessment of the healing of tibial fractures after intramedullary fixation. J Bone Joint Surg 2002;84B(1): 15–18 (British). [15] Lenert LA. The reliability and internal consistency of an internet-capable computer program for measuring utilities. Qual Life Res 2000;9(7):811–7. [16] Resnick B, Parker R. Simplified scoring and psychometrics of the revised 12-Item Short-Form Health Survey. Outcomes Manag Nurs Pract 2001;5(4): 161–166. [17] Salyers MP, Bosworth HB, Swanson JW, Lamb-Pagone J, Osher FC. Reliability and validity of the SF-12 health survey among people with severe mental illness. Med Care 2000;38(11):1141–50. [18] Ware Jr JE, Kosinski M, Keller SD. A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity. Med Care 1996;34(3):220–33. [19] Sanderson K, Andrews G. The SF-12 in the Australian population: crossvalidation of item selection. Aust N Z J Public Health 2002;26(4):343–5.
The patient-reported outcomes of lower limb long bone shaft fractures do not return to normal at one year. Even patients whose
* Funded by the Transport Accident Commission (TAC) through the Institute for Safety, Compensation and Recovery Research (ISCRR).
1658
W.-H. Tay et al. / Injury, Int. J. Care Injured 45 (2014) 1653–1658
[20] Kiely JM, Brasel KJ, Guse CE, Weigelt JA. Correlation of SF-12 and SF-36 in a trauma population. J Surg Res 2006;132(2):214–8. [21] Ware Jr JE, Kosinski M, Keller SD. SF-12: How to score the SF-12 Physical and Mental Health Summary scales. 3rd ed. Lincoln: QualityMetric; 1998 . [22] Avery J, Dal Grande E, Taylor A. Quality of life in South Australia as measured by the SF12 health status questionnaire: population norms for 2003, trends from 1997–2003. Adelaide: South Australian Department of Human Services; 2004. [23] Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47(11):1245–51. [24] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40(5):373–83. [25] Gould B. Stata Release 13. College Station, TX, USA: StataCorp LP; 2013. [26] Harris I, Hatfield A, Donald G, Walton J. Outcome after intramedullary nailing of femoral shaft fractures. Aust N Z J Surg 2003;73:387–9. [27] Harris I, Lyons M. Reoperation rate in diaphyseal tibia fractures. Aust N Z J Surg 2005;75:1041–4.
[28] Dogra AS, Ruiz AL, Marsh DR. Late outcome of isolated tibial fractures treated by intramedullary nailing: the correlation between disease-specific and generic outcome measures. J Orthopaed Trauma 2002;16(4):245–9. [29] Spencer RF. The effect of head injury on fracture healing. A quantitative assessment. J Bone Joint Surg 1987;69B(4):525–8 (British). [30] Bhandari M, Guyatt GH, Swiontkowski MF, Tornetta III P, Sprague S, Schemitsch EH. A lack of consensus in the assessment of fracture healing among orthopaedic surgeons. J Orthopaed Trauma 2002;16(8):562–6. [31] Bhandari M, Guyatt GH, Tornetta III P, Schemitsch EH, Swiontkowski MF, Sanders D, et al. Randomized trial of reamed and unreamed intramedullary nailing of tibial shaft fractures. J Bone Joint Surg 2008;90:2567–78 (American). [32] Skoog A, So¨derqvist A, To¨rnkvist H, Ponzer S. One-year outcome after tibial shaft fractures: results of a prospective fracture registry. J Orthop Trauma 2001;15(3):210–5. [33] Swiontkowski MF, Agel J, McAndrew MP, Burgess AR, MacKenzie EJ. Outcome validation of the AO/OTA fracture classification system. J Orthopaed Trauma 2000;14(8):534–41.