Efficacy of low-intensity pulsed ultrasound in healing of lower limb fractures

Abstract

Introduction, purpose and relevance Fracture healing represents a worldwide challenge to return to ordinary well-being preceding injury. Fracture healing encompasses multifaceted and progressive events due to the extent of the injury, factors associated with management and patient-related variables. The more complex the injury is, the higher the chances of developing infection and therefore the longer the healing will take. Fracture healing also depends on the surgical management of choice, which normally depends on the severity of the injury and consists of either open reduction internal fixation or external fixation. In addition, patient-related factors have an added impact as to whether the fracture heals or not. Patient-related factors, such as comorbidities including immuno-compromising diseases and health risk behaviours, are known to hinder fracture healing and result in non-union fractures. The prevalence of lower limb fracture non-union in South Africa is not well documented, as the medical and patient-related factors associated with lower limb non-union fractures in this context have not been reported. It was deemed useful to determine the clinical pathway and the patient outcomes of lower limb non-union fractures in South Africa. The literature revealed that the management of delayed and non-union fractures is problematic and alternative approaches to augment the management of fractures have been growing globally. One form of intervention is low-intensity pulsed ultrasound, which has been studied internationally and reported to be effective in fracture healing. Studies that have been done do not highlight the effect of ultrasound in patients with a high prevalence of comorbidities, including diabetes, hypertension, HIV and health risk behaviours such as smoking. The studies that have investigated the efficacy of LIPUS involved its daily application. The efficacy of low-intensity pulsed ultrasound in lower limb fractures in a South African population with a high prevalence of multiple comorbidities has not been determined. The relevance of the information gathered is to reduce the burden of fracture healing and non-union fractures by using non-invasive equipment such as low-intensity pulsed US that could help some patients with surgically managed fractures with either delayed or non-union fracture healing, which leads to prolonged stay in hospital. Therefore, the study sought to: vi ● To determine the prevalence and site of lower limb non-union fractures in hospitals in Johannesburg ● To determine the factors to inform a proposed clinical pathway for patients with lower limb non-union fractures at selected public hospitals in Gauteng ● To determine the effect of non-invasive low-intensity LIPUS on the healing of lower limb fractures The specific objectives the study were: 1. To establish the prevalence of lower limb non-union fractures at three selected public hospitals in Johannesburg. 2. To determine the factors in forming a proposed clinical pathway for patients with lower limb non-union fractures at the five selected public hospitals in Gauteng, South Africa. 3. To establish the quality of life of patients with lower limb non-union fractures. 4. To establish the effect of low intensity pulsed ultrasound on lower limb fractures including non-union fractures. 5. To establish the influence of demographic, health status variables and smoking on fracture healing. Methodology The study includes two phases. The first phase had two parts. Part 1 used a retrospective record review of 3847 records on patients with lower limb fractures, to establish the prevalence of non-union fractures in three academic state hospitals in Gauteng province, South Africa. In Part 2, a cross sectional prospective descriptive study was used to determine medical- and patient-related factors associated with non-union fractures. Data were collected on medical factors, patients’ health-related quality of life and impairment in everyday activities to contribute to a proposed clinical pathway for physiotherapy for the treatment of lower limb non-union fractures. In the second phase was a double-blinded randomised controlled trial, using a consecutive sampling method to recruit patients with lower limb fractures. One hundred and fifteen patients, including 17 with non-union fractures, were recruited on admission from two state academic hospitals and randomly assigned to the control or intervention group, and who received real or sham low-intensity pulsed vii ultrasound. The researcher confirmed the fracture site from the file and X-rays as diagnosed by the radiologist, before the treatment commenced. Treatment with low intensity pulsed ultrasound was administered every second day to attain 20 treatments. Both machines were pre-set at an intensity of 0.03 W/cm², frequency of 1.5 MHz and pulse width of 200 μs. The sham machine was set not to produce ultrasound. After the completion of low-intensity pulsed ultrasound treatment, fracture healing was assessed using X-rays for callus formation, cortices bridged, fracture gap and overall healing at six, 12 and 18 weeks. Both the researcher and research assistance providing the intervention and the radiologist who assessed the X-rays were blinded to patient group allocation. The fracture healing outcomes were observed by the radiologist on X-rays. The difference in fracture healing between the intervention and control group was analysed using Pearson’s Chi Square test. The percentage change in fracture healing within the intervention and control group was analysed using the McNemar Chi-square test at six, 12 and 18 weeks. Data to determine the association between fracture outcomes and personal and medical demographics, comorbidities and risky health behaviours were analysed using logistic regression and presented as odds ratios and 95% confidence intervals. The logistic regression was used to establish the effect of medical comorbidities on fracture healing in the context of the current study. Data to determine the association between fracture outcomes and personal and medical demographics, comorbidities as well as risk health behaviours were analysed using logistic regression and presented as odds ratios and 95% confidence intervals. Results and Discussion The results for Phase 1, Part 1 indicated a low prevalence of less than 1% (0.72%; n=28) for fracture non-union. Non-union was more prevalent in males than in females in the 41-60 age group. The low prevalence was attributed to the exclusion criteria used in this phase of the study. The exclusion criteria included patients with pathological fractures, and multiple fractures on the ipsilateral side. However, even in Phase 1, Part 2 of the study, where inclusion criteria were widened, only 22 participants were recruited at five state hospitals in Gauteng over an 11month period. Participants were between 20 and 59 years, unemployed or viii dependent on a social grant. A high percentage of participants in this study were HIV positive (41%), and 13% had diabetes, while all of them smoked. Medical factors determined a long stay in hospital, a number of re-admissions, medical and surgical procedures, medication, rehabilitation therapies, as well as assistive devices. All participants had impairment in everyday activities and decreased quality of life. A clinical pathway for physiotherapy treatment of patients with non-union fractures was created considering these factors. In Phase 2 of the study, the participants between the ages of 18 and 82 years, again presented with a high prevalence of comorbidities, HIV positive (43%), 27% had diabetes and 61% of them smoked. Results indicated no significant difference in fracture healing for patients with surgically managed lower limb fractures who received real and sham low-intensity pulsed ultrasound for any period up to 18 weeks. Some acceleration in fracture healing was found for a small percentage of the experimental group, and this group has less deterioration in fracture healing over 18 weeks than did participants in the control group. A significant difference was found for the participants with non-union fractures treated with real low-intensity pulsed ultrasound compared to those treated with sham low-intensity pulsed ultrasound. Fracture healing in the logistic regression for single factors was found to be associated with gender, the site of fracture, fracture management and non-union, while multivariate regression analysis showed that gender, type of fracture management, non-union and HIV were associated with fracture healing. Conclusion Fracture non-union and the outcomes of non-union fractures are of concern in patients treated in state hospitals in South Africa. There are considerable costs in terms of medical and rehabilitation care, as well as to the patient in terms of accessing care. Patients face a prolonged period in which no fracture healing is reported and present with decreased health-related quality of life and functional impairment. In view of the high prevalence of comorbidities associated with non union fractures, we proposed a clinical pathway for physiotherapy, which in turn suggested that the use of low-intensity pulsed ultrasound might be useful in facilitating fracture healing. ix While low-intensity pulsed ultrasound cannot be recommended for the treatment of surgically managed lower limb fractures in patients in state hospitals in Gauteng, it may be recommended for patients with non-union fractures. A significant difference in terms of fracture healing was seen for patients with non-union fractures who received low-intensity pulsed ultrasound every second day for 20 days of treatment. However, this recommendation must be viewed in light of the small sample size of patients with non-union fractures and the costs and challenges of providing this intervention.