Lateral epicondylitis of the elbow is characterised by pain and tenderness of the lateral humeral epicondyle and pain on resisted dorsiflexion of the wrist, the 3. digit or both. There is also often pain on resisted radial deviation of the wrist. The condition is a frequent complaint with an overall prevalence of 1-3% . The highest incidence is found in persons 40-60 years old. For women, the incidence increases to 10% between the ages of 42 - 46 [2, 3]. The incidence in general practice is estimated to be 4 - 7 per 1000 per year [2, 4, 5]. The aetiology has been assumed to be over-use damage to the forearm extensor muscles - either minor or non - recognised traumas. There is little evidence of inflammation . Most authors attribute the condition to a lesion in the extensor apparatus at the lateral humeral epicondyle, specifically the short radial extensor muscle [2, 7]. Cyriax  and Ombregt et al  identify four subgroups of lateral epicondylitis depending on the exact location of the lesion. In their experience, in 90% of the cases, the lesion is situated in the anterior part of the lateral epicondyle at the origin of the short radial extensor muscle. The second most frequent lesion accounts for 8% of the cases and is localised at the muscle body itself. Lateral epicondylitis usually is a self-limiting condition, but complaints may last up to 2 years or longer . A study from general practice shows that 80% heal within one year on wait-and-see treatment (rest, paracetamol or NSAIDs taken orally) even when initial symptoms had lasted more than 4 weeks . In many countries, treatment guidelines recommend a wait-and-see policy.
Many treatments have been proposed leading to a number of trials, but reviews including several recent meta-analyses have led to no conclusions as to which is the best. This is due to low statistical strength, low internal validity and insufficient study data reporting [12, 13]. Schmidt et al 2003  reviewed literature on physical therapy prior to 1999 and found no evidence of effect, with the exception of ultrasound, where a minor effect was shown. Bisset et al 2005  published a meta-analysis of 28 randomised studies published before 2003 of different physical therapies for lateral epicondylitis satisfying at least 15 out of 28 criteria (PEDro rating scale). Most studies had a small number of subjects, and only eight had long term follow-up of effect of therapy. Extra corporeal shock wave therapy was found to have no effect, and manipulation and exercise were found to have only a short-term effect.
A meta-analysis by Smith et al 2002  on the effect of corticosteroid injections found evidence of short-term pain relief, but no effect beyond the initial 6 weeks. There was however some uncertainty due to few and small studies.
The Cochrane Library has several reviews of treatment for lateral epicondylitis: acupuncture , deep transverse friction massage , NSAIDS , orthosis , extra corporeal shock wave therapy  and surgery . These reviews all conclude that there is insufficient evidence to draw firm conclusions as to which methods of treatment are effective. However, there are indications that topical NSAIDs and manipulation and exercise have a short term effect. As to NSAIDs taken orally, there is probably a short-term effect, although it is impossible to either recommend use or not. For extracorporeal shockwave therapy, there is evidence to conclude that this treatment has no effect. Ultrasound has a possible short-term effect based on one meta-analysis . In fact, there is scant support for any long-term treatment in the literature.
Looking for a better treatment for epicondylitis, we have found two studies to be of special interest. Both were carried out in primary care settings with one year follow-up. One study compared corticosteroid injection with physical therapy (ultrasound, manipulation and exercise) and a wait-and-see group . The other compared corticosteroid injection with naproxen orally and placebo medication . Both concluded that corticosteroid injection is safe and effective for pain relief during the first 6 weeks, and the effect of this treatment is better than physiotherapy, wait-and-see and naproxen orally within the same time-frame. Smidt et al  found that physiotherapy gave some, but not significantly better long-term effect than wait-and-see treatment. A more recent study comparing physiotherapy and corticosteroid injection  concluded that the significant short term benefits of corticosteroid injection are paradoxically reversed after six weeks, with high recurrence rates and that combining elbow manipulation and exercise has superior benefit to wait and see in the first six weeks and to corticosteroid injections after six weeks. Comparing corticosteroid injection with placebo injection, four clinical trials found no significant effect of corticosteroids at 6 months [26–29] and at 12 months, although one trial reported improvement at 8 weeks .
We find there is a good reason to investigate the long-term effects of physiotherapy - this is also recommended in a recent meta-analysis . At the same time, it would be interesting to see whether the good initial response from corticosteroid injection [11, 24] may be extended if combined with physiotherapy. We have only found two studies that have evaluated this combination, one found no added effect of corticosteroid injection at 6 months , the other had only a 7 week follow up . A protocol for a larger study has recently been published .
The objective of this study is to compare the clinical effect of physiotherapy alone or physiotherapy combined with corticosteroid injection in the initial treatment of acute lateral epicondylitis in a primary care setting. Also, to find the short and long term effect of physiotherapy and to ascertain whether this outcome is influenced by corticosteroid injection.