Page 64 SA Orthopaedic Journal Spring 2014 | Vol 13 • No 3 Effect of additional ultrasound therapy to analgesics in treatment of acute low back pain A randomised control trial conducted at the Aga Khan University Hospital, Nairobi (AKUH, N) Dr S Mohammedali* MBChB, MMed(Surg) Dr SK Mutiso** BSc MBChB Dr P Oroko* MBChB, MMed(Surg), FRCS(Ed), FRCS(Tr&Orth) Dr B Ombachi* MBChB, MMed, AO Spine Fellow Dr H Saidi*� BSc, MBChB, MMed(Surg), FCS(ECSA), FACS *Department of Surgery, Aga Khan University Hospital, Nairobi, Kenya **Medical Officer at the Tigoni Hospital, Limuru, Kenya � Department of Human Anatomy, University of Nairobi, Kenya Correspondence: Dr S Mohammedali Department of Surgery Aga Khan University Hospital Nairobi, Kenya PO Box 030270-00100 Nairobi, Kenya Tel: +254711092115 Fax: +254203743935 Email: shamshuza@yahoo.com Abstract Background: Acute low back pain (LBP) is a common condition that is encountered by many physicians with varied treatments instituted in its management. Ultrasound physiotherapy is a common modality used in its management, although its effectiveness and its role in management of acute LBP is not well known. Methods: A randomised controlled trial was conducted to compare the effect of the addition of ultrasound therapy to a defined analgesia protocol in patients presenting with acute LBP at the Aga Khan University Hospital, Nairobi. The main outcomes were reduction in disability and pain which were evaluated using the mean change in Oswestry Disability Index (ODI) and Visual Analogue Scale (VAS) score for pain, respectively. Patients were followed up for a minimum of four weeks with assessment using the ODI and VAS at weekly clinic visits. Results: Seventy-four patients were included in the study. Thirty-six patients were allocated to the analgesia with additional ultrasound group and 38 to the analgesia alone group. The Minimal Clinically Important Difference (MCID) of 10% points at four weeks after initiation of treatment was achieved in the ultrasound group but not in the analgesia only group. This difference was not statistically significant though, 10.35% vs 8.44%; p=0.36. There was no difference in the mean change in the ODI score between the two groups at any of the four follow-up visits after initiation of treatment. Mean difference in change of ODI (95% confidence interval) was −3.2(−7.0 to 0.6) after the first week, 2.96(−1.3 to 7.2) after the second and 1.90(−2.3 to 6.1) after the third week, p=0.36, 0.17 and 0.096 respectively. There was also no difference detected in the mean change of VAS score between the two groups at the first and fourth visit. Mean difference in VAS between the first and fourth weeks was 0.2 with a 95% confidence of −0.85 to 1.2 (p=0.72). Conclusion: The addition of ultrasound therapy to the treatment of acute LBP improved patient outcomes as assessed by an MCID of a disability index, but which was not statistically significant. No outcome difference was noted in the two groups using the VAS pain score. Key words: acute low back pain, analgesia, ultrasound therapy SAOJ Spring 2014_BU_Orthopaedics Vol3 No4 2014/07/30 3:04 PM Page 64 SA Orthopaedic Journal Spring 2014 | Vol 13 • No 3 Page 65 Introduction Low back pain (LBP) is a common condition worldwide with associated significant disability and economic implications.1,2 The lifetime prevalence of LBP is over 70% in industrialised countries with peak prevalence occurring between ages 35 and 55 years.3 It is the second most common reason for absence from work in adults aged <55 years, with work-related LBP estimated to cause 818 000 disability-adjusted life years lost annually.4,5 In African countries, a lifetime prevalence of 62% among adults has been reported and this was associated with significant disability, loss of produc- tivity and working man hours.6 Most patients with LBP have no serious pathologic aetiology, with most having their LBP attributed to an unidentified cause, either due to mechanical strain or related to work/posture.7-9 The natural history of acute LBP is resolution of symptoms in the majority of patients by the fourth week of onset with 90% returning to work within three months.10-12 With minimal intervention such as analgesia and minimising activity to tolerable levels only, most patients improve in the first few weeks.13,14 Analgesic therapy alone without additional treatment has been shown to be effective in terms of pain control and reduction of disability.15 The use of paracetamol, ibuprofen and mefenamic acid has been recommended, with opiates and muscle relaxants being disapproved due to their sedative effects, drug dependence and lack of additional benefit.12,16,17 Interventions such as physio- therapy, bed rest and back exercises that are frequently used in clinical practice, have no strong evidence to support their routine use.12,14 Moreover, massage and transcutaneous electrical nerve stimulation (TENS) therapy are not recommended as modalities of therapy because of a lack of proven clinical benefit.12 Ultrasound waves are often used as an adjunct to medical therapy for their thermal effect.18,19 However, although its use is expensive to the patient there is no proven clear benefit in the management of acute LBP.20-22 There is no published randomised trial to assess this. The present study aimed to assess the effectiveness of ultrasound in addition to analgesia as compared to analgesia alone as interventions to reduce the morbidity of LBP with regard to pain control and disability. Methods Trial design The study was a randomised, controlled trial of patients with acute LBP comparing an intervention group, who received the standard analgesia protocol and ultrasound therapy, with a control group, who received only the standard analgesia protocol (Figure 1). Participants Patients presenting with acute LBP, as defined in the inclusion and exclusion criteria, aged 18–60 years. The patients were all seen at the Aga Khan University Hospital, Nairobi (AKUH, N), Kenya. The inclusion criteria constituted: Patients aged 18–60 years with acute onset LBP. An acute episode of pain was defined as pain that lasts for more than 24 hours but less than six weeks. Moreover, the pain had to be in an area bounded superiorly by T12 and inferiorly by the buttock crease. The exclusion criteria constituted the following: • Patients with indicators of serious pathology from history and examination findings, also referred to as ‘red flags’.12 Patients with red flags were further investigated with imaging and referred for further management. • Patients who were vomiting – we wished to administer the analgesic prescription by the oral route. • Patients who were already on other forms of analgesia, e.g. opioids and muscle relaxants. • Patients with contraindications to ultrasound treatment such as skin allergies, dermatitis or a skin haematoma, and those already on other forms of physiotherapy. Interventions Study protocol A study guideline and unbiased randomisation was adhered to. This guideline was circulated and a copy for reference was available at all study recruitment sites. Informed consent was obtained by the treating physician after discussing the study with patients. Training was held for the doctors and physiotherapists who were involved in patient recruitment and to maximise protocol adherence. The ODI was given to the patients and was reported by patient self-response on study entry and at intervals of 1, 2 and 3 weeks after initiation of therapy. Following an initial visit, they were provided with a mobile phone contact for reporting non-improvement of their pain and instructions for escalation of pain management were given as per the protocol. Treatment protocol Control arm: Patients were commenced on 1 g of parac- etamol by mouth stat, then 1 g 6 hourly for seven days. All patients were managed as outpatients unless they had severe pain, i.e. ODI score of 50% or more or a VAS score of more than 8/10; were not ambulatory due to pain; requested to be managed as an inpatient; fulfilled the three criteria above and could afford the admission fee or opted to go to another hospital. Intervention arm: Patients were commenced on 1 g of parac- etamol by mouth stat, then 6 hourly for seven days. Concurrent physiotherapy (ultrasound only) was commenced with the first session starting the same day of recruitment unless they presented after midnight, in which case it was commenced on the earliest available session. Pulsed ultrasound therapy was administered with a total of three sessions in a week for two weeks (a total of six sessions). All patients were managed as outpatients unless they met the criteria stated above. With minimal intervention such as analgesia and minimising activity to tolerable levels only, most patients improve in the first few weeks SAOJ Spring 2014_BU_Orthopaedics Vol3 No4 2014/07/30 3:04 PM Page 65 Page 66 SA Orthopaedic Journal Spring 2014 | Vol 13 • No 3 Follow-up: In both arms, patients managed on an in/outpa- tient basis were reviewed at intervals of 1, 2, and 3 weeks with the ODI administered at every contact. If at subsequent review the reported pain was persistent or had increased, the treating physician was to add a non-steroidal anti- inflammatory drug, i.e. Ibuprofen 400 mg twice daily for seven days.17,21,23 All patients had access to a central help line number for enquiries or to report any problems during their management. Direct questioning about medication compliance was used to confirm patient compliance with the protocol. Outcome Oswestry Disability Index (ODI) Outcomes were measured using a patient-administered questionnaire, the Oswestry Disability Index (ODI). The ODI is a validated and widely used tool for assessing disability from back pain. It is a ten-question, self-adminis- tered questionnaire that is convenient to use. It has been translated for use into various languages around the world, and has been shown to be indicative of the respondent’s disability from LBP. We used the English version to assess disability from acute back pain in this study.23,24 The Minimal Clinically Important Difference (MCID) for the ODI in the present study was taken to be 10 percentage points.23 Sample size determination Malimvaara and colleagues14 showed that at three weeks, the mean ODI index for patients who were randomised to exercise as compared to no treatment, was significantly different (mean difference=6.6) to that for patients on bed rest (mean difference=3.9). In the same study at 12 weeks, the mean ODI index was about 6 for all treatments. In a meta-analysis of acute LBP treatment, Machado and colleagues compared the analgesic effects of treatments for non-specific LBP using a 100-point scale and reported that analgesics had a moderately significant treatment effect (10–20 points) when compared to placebo, unlike exercise, radiotherapy, traction, physiotherapy or prolotherapy, which showed a minor treatment benefit (<10 points).24 The estimated pooled standard deviation (SD) of the ODI for the bed rest-control and exercise-control comparisons was 10.5 and 11.71 at three weeks, and 9.44 and 10.07 at 12 weeks, respectively. Using a level of significance of 5% and 80% power, the study needed 37 patients in each study arm assuming a clini- cally meaningful difference (δ) in the ODI index of 10, an observed difference (Δ) between the control and treatment arm of 3.5 and a standard deviation of 10. The formula for calculating the sample size in the control and treatment group for a comparative study is given by: Figure 1. Study flow chart Pain improved Pain persistent/ worsened Pain improved Pain persistent/ worsened Administer ODI at 1, 2 and 3weeks Add Ibuprofen 400mg BD X 7days Administer ODI at 1, 2 and 3weeks Add Ibuprofen 400mg BD X 7days Paracetamol 1gm PO QDS X 7days Paracetamol 1gm PO QDS X 7days + Ultrasound therapy Improved Not improved Administer ODI at 1, 2 and 3 weeks Exclude, start opioid Improved Not improved Administer ODI at 1, 2 and 3 weeks Exclude, start opioid ACUTE LOW BACK PAIN SAOJ Spring 2014_Orthopaedics Vol3 No4 2014/08/06 11:36 AM Page 66 SA Orthopaedic Journal Spring 2014 | Vol 13 • No 3 Page 67 where: Z1-a and Z1-b are the critical values from standard normal tables associated with Type I and II errors as a and b respectively, Δ = μE – �μC (difference in treatment effect) s is the standard deviation showing variation in the ODI index in each group. It was hypothesised that at three weeks the absolute difference in the mean ODI in the control arm (μC) and the mean of the index from the other arm (μE) is not clinically important. Thus the hypotheses to be tested are: H0: | μE – μC | ≥d� HA: | μE – μC |