Sports physicians, orthopaedic surgeons, podiatrists, podiatric surgeons, and physiotherapists working in sports injury clinics who were known to the main 2 authors (NP, TAC) across the UK were contacted via email and invited to refer patients with painful, recalcitrant MTSS that had not responded to other conservative treatment modalities, whom they felt might benefit from the trial intervention (Figure.1.). The failed conservative treatments included, (a) rest, ice, compression, elevation (RICE), (b) assessing and addressing any lower limb functional factors (muscle strength & flexibility, proprioception & balance, (c) therapeutic ultrasound therapy, (d) ESWT, (e) acupuncture, (f) needling, (g) Graston fascia release, (h) improving neurodynamic, (i) non-steroidal anti-inflammatory drugs (NSAIDs), (j) correction of lower limb biomechanics (e.g., foot orthoses, brace, and taping for control of foot pronation), and (k) walking/running gait assessment and gait re-training.
Inclusion criteria were patients with persistent, painful MTSS assessed and confirmed by the lead clinician (NP). Exclusion criteria were previous periosteal surgery for MTSS, previous or current tibial stress fracture, or contraindications to the intervention such as pregnancy or anticoagulant therapy.
Patients were assessed by the lead clinician and the diagnosis of MTSS was confirmed by a history of exercise-induced pain over the posteromedial border of the middle to the distal third of the tibia, a positive Shin Palpation Test (palpation tenderness over the painful middle and distal thirds of the medial tibia) both at rest and following exercise, and MRI changes showing periosteal or bone marrow oedema.
Ethical approval was granted by the Queen Mary University of London Ethics Review Board (QMREC2009/22). Participants provided signed written informed consent. Subsequently, a detailed medical history and biographical data were taken and an examination was performed.
Intervention and injection procedure
The injection procedure is very simple and within the expertise of all clinicians who are used to injection techniques for musculoskeletal pathology. It is primarily performed in the out-patient department (OPD) and does not require a local anaesthetic even though local anaesthetic is used to dilute 50% glucose down to 15%. In some cases where pain tolerance is poor, it can be performed under a general anaesthetic. In this study, all subjects were treated in OPD. The target area for needle placement in the area anterior to the deep crural fascia (Figs. 1A and 1B.) along the medial tibia. The skin overlying the most painful area of the tibia was marked with an indelible marker pen and then cleansed using alcoholic chlorhexidine (2% chlorhexidine gluconate, 70% isopropyl alcohol). An ultrasound scanner (USS) (Siemens AG, Berlin, Germany) was used to guide and confirm needle position. The needle is introduced under real-time USS from the most proximal end of the site of pain (knee end) to the most distal (ankle end) (Figs 2A & 2B). The spinal needle (0.7 mm diameter x 90 mm length, Becton, Dickinson and Company LLC, Franklin Lakes, New Jersey, US) was positioned parallel along the medial tibia in the area anterior to the deep crural fascia (Fig.3). The needle introducer was removed with the needle in place. 15% dextrose solution was slowly infiltrated, approximately 1mL of solution per 1cm along the whole length of the area of pain. In some cases where the length was longer than the spinal needle, the second entry point was made following the same protocol as above. After the injection, the area was cleaned, dressed, ice packs applied for 2 minutes, and knee high compression socks (Fig.4.) (Bauerfeind AG, Zeulenroda-Triebes, Germany) were fitted. Patients were advised to continue wearing them for up to four weeks, removing them at night. Patients were advised to take relative rest for three days and advised for simple flexibility exercises and a graded return to physical activity. Patients were followed up one week later to monitor progress and address any concerns or questions. Patients were also given an emergency mobile number to call if they experienced any adverse reaction to the injection. This would include, extreme unrelenting pain, erythema, swelling, delayed allergic rash or sleep disturbance.
The patients' 'average pain', defined as the most pervasive severity of pain throughout 24 hours, was measured using a 10-cm visual analog scale (VAS). The VAS is sensitive (35), reliable, valid, and responsive for measuring pain in other common musculoskeletal conditions, such as patellofemoral pain syndrome (36). The pain was assessed in this way at 0 (baseline), 1, 2, and 4 weeks after the injection, and at medium-term (mean 18 weeks, range 13-36 weeks) and long-term (mean 52 weeks, range 47-74 weeks) follow-up to assess the patients' response to the intervention.
A Likert (37) symptom resolution scale was used to measure the subjective degree of recovery at medium-term and long-term follow-up compared to baseline. There are six possible outcome scores for the Likert scale: 1– completely recovered, 2– much improved, 3– somewhat improved, 4– no change, 5– worse, 6– much worse. Treatment was classed as a success in patients who rated themselves as 'completely recovered' or 'much improved', reflecting the method of previous authors (38). Other scores were considered a treatment failure. Categorical scales such as the Likert are sensitive indicators of clinical trial endpoints (37).
Return to the sport was assessed with a five-point activity scale at medium-term and long-term follow-up: 1– not active at all, 2– no return to sport, 3– returned to the sport at an unsatisfactory lower level, 4– returned at desired but not pre-injury level, 5– returned at the pre-injury level. Although not reported in other literature, treatment aimed to enable patients to return to their desired sports at pre-injury levels. All other activity scores were considered a treatment failure.
Statistical analyses were performed using SPSS for Mac version 19.0.0 (IBM, New York, NY, US). Statistical significance was set at a p-value less than 0.05. The Shapiro-Wilk test was used to evaluate the normality of the distribution of data. The distribution of data were negatively skewed and not normally distributed (Shapiro-Wilk = 0.009) therefore appropriate non-parametric tests were performed to evaluate the changes in pain levels.
Median values and interquartile ranges were calculated to compare baseline and follow-up data for VAS average pain scores, Likert symptom resolution, and return to sports scores.
Friedman's non-parametric test was used to compare the within-patient treatment response over time. Post-hoc Wilcoxon signed-rank tests with Bonferroni corrections were performed to determine VAS average pain response to treatment over five paired periods (baseline – 4 weeks, baseline – 18 weeks, 4 – 18 weeks, baseline – 52 weeks, 4 weeks – 52 weeks). The Bonferroni corrected alpha value was (p<0.01).