Acupuncture

Three studies reviewed acupuncture therapy. In a randomized, controlled, double-blind study involving 45 patients, Fink and coauthors¹⁵ noticed a significant improvement in maximum grip strength, pain, and function based on the disability of the arm, shoulder, and hand (DASH) form at 2 weeks in the true acupuncture versus sham acupuncture. At 2 months, however, this benefit was no longer seen. In a single blinded, randomized, controlled trial involving 48 patients, Molsberger and Hille¹⁶ found acupuncture pain relief superior and longer lasting (20.2 vs. 1.4 hours average) than sham acupuncture. Davidson and coworkers¹⁷ compared acupuncture with ultrasound in a study enrolling 16 patients and found significant improvement in both treatments over baseline in DASH scores, pain, and pain-free grip strength but found no significant difference between the two modalities at 1 month. In conclusion, not enough evidence exists to support acupuncture definitively, but studies suggest (grade B) it may have a role in the short term (<6 weeks) only.

Nonsteroidal Anti-Inflammatory Medication

Three randomized, controlled trials on topical NSAIDs versus placebo were done, and meta-analyses were conducted by Green and researchers.¹⁸ The trials (Burnham et al., 1998,¹⁹ Burton 1988,²⁰ and Jenoure et al., 1997²¹) had a combined population of 130 subjects and assessed the effect of topical NSAIDs (2 trials used diclofenac, 1 trial used difflam) on pain. The pooled weighted mean difference (WMD) was 2 1.88 (95% confidence interval [CI], 2 2.54 to 2 1.21) in the short term (1–3 weeks). That is, those in the topical NSAID group had 1.88 of 10 points less pain on a visual analogue scale (VAS) at the end of the trial than those using placebo.¹⁸

No clinically significant differences were found between topical NSAIDs and placebo in the effect on strength, tenderness, range of motion, or physician’s opinion regarding effect. Side effects were reviewed in two of the three trials, and no significant differences were found.

Two trials examined oral NSAIDs versus placebo with contradictory results. Labelle and Guibert,²² examining diclofenac daily use for 1 month plus cast immobilization versus immobilization and placebo in a double-blinded, randomized study involving 129 patients, found a significant difference in pain reduction (1.4/10 less pain) but not in grip strength or function in the diclofenac group. The diclofenac group, however, had a greater incidence rate of abdominal pain (30% vs. 9%) and diarrhea (39% vs. 20%) than the placebo group. Hay and investigators,²³ examining naproxen use in a study involving 111 randomized patients, however, found no significant difference in improvement in pain after 1 month and 1 year. Side effects were not mentioned in the study.

In conclusion, good evidence (grade A) exists that topical NSAIDs have a role in short-term pain relief. Oral NSAIDs have conflicting (grade I) evidence supporting their use and are associated with more side effects. Further studies are needed.

Corticosteroid Injections

Three randomized studies were found on corticosteroid injections. Two of these studies involve a control group in which avoidance of aggravating activities is the only treatment, and they give us the best available data on the natural history of LE.

Altay and coauthors,²⁴ in a prospective, randomized trial involving 120 patients, compared injecting triamcinolone with lidocaine to lidocaine alone using the peppering technique (injecting 40–50 times through the same skin entry point) and found no differences in the two groups at 2 months in the number of excellent results defined as complete pain relief, completely satisfied and devoid of pain on resisted wrist dorsiflexion (93% in the triamcinolone arm vs. 95% in the control group). Results were not changed at 6 months or 1 year.

Smidt and researchers²⁵ compared physiotherapy with corticosteroid injection with a wait-and-see approach in a randomized, controlled trial involving 185 patients. Exclusion criteria included physiotherapy treatment or injections for the elbow complaint in the last 6 months. Corticosteroid injections consisted of 1 mL triamcinolone (10 mg/mL) and 1 mL 2% lidocaine injected into all painful areas until resisted dorsiflexion was no longer painful. A maximum of 3 injections over a 6-week period were allowed. Fifty-eight percent (58%) used one injection, 27% used two, and 15% used three. The physiotherapy consisted of nine treatments of pulsed ultrasound, deep friction massage, and a home exercise program over 6 weeks. The waitand-see group was educated on the disease and how to avoid provoking pain. All groups were told to avoid aggravating activities, discouraged from using other forms of treatment but allowed pain medication (naproxen [Naprosyn] and paracetamol). Patients kept a diary of other forms of treatments used, but activity level was not recorded. Success was defined as pain completely resolved or much improved. The researchers also reviewed patient self-rated general improvement, severity of the main complaint, pain, elbow disability, and patient satisfaction. Severity of elbow complaint, grip strength, and pressure pain threshold were assessed by a research physiotherapist.

As shown in Figure 22-1, at 6 weeks, corticosteroids had significantly greater success rate (92%) than both physiotherapy (47%) and wait-and-see (32%) approaches, but this was no longer true at 3 months. The corticosteroid injection group had a greater relapse rate (defined as no longer completely resolved or much improved) at 72% than the physiotherapy and wait-and- see groups (at 8% and 9%, respectively). At 1 year, the injection group had a lower success rate (69%) than both the physiotherapy (91%) and the wait-and-see (83%) groups.

FIGURE 22-1 Success rates of three treatment regiment.

(Adapted from Smidt N, van der Windt DA, Assendelft WJ, Deville WL, Korthals-de Bos IB, Bouter LM. Corticosteroid injections, physiotherapy, or a wait-and-see policy for lateral epicondylitis: a randomised controlled trial. Lancet 23;359[9307]:657-662, 2002 with permission of Elsevier.)

In the follow-up period (week 7 to 1 year), 76% of the wait-and-see group used no other forms of treatment compared with 57% in the corticosteroids group and 19% in the physiotherapy group. In the wait-and-see group, 20% used pain medication. In the corticosteroids group, 21% used physiotherapy, 34% used more injections, and 27% used pain medication. In the physiotherapy group, 66% used more physiotherapy. In 38 of the 42 patients who used more physiotherapy, the duration of the extra physiotherapy was less than 6 weeks.

Bisset and colleagues²⁶ also compared corticosteroid injections to physiotherapy to a wait-and-see approach in a single-blinded, randomized, controlled trial involving 198 participants with follow-up of 1 year. Participants had symptoms of LE for more than 6 weeks. Exclusion criteria included any treatment for LE by any health practitioner in the last 6 months. The corticosteroid injection group received 1 mL triamcinolone (10 mg/mL) and 1 mL of 1% lidocaine to painful elbow points. A second injection was allowed at 2 weeks if needed. Eighty-six percent received one injection, whereas 12% received two and 1.5% (one patient) received three. Physiotherapy consisted of eight sessions of elbow manipulations and exercises over a 6-week period. All patients were given an information booklet outlining the disease process and providing practical advice on selfmanagement and ergonomics. A diary recording any other forms of treatment was kept. Analgesics were allowed, but other treatment modalities were discouraged. Primary end points were global improvement scale where completely resolved or much improved were deemed successful, pain-free grip strength as a percentage of normal side, and assessor’s rating of severity. As shown in Figure 22-2, at 6 weeks, corticosteroid injections were significantly better in all 3 end points, having 78% success rate compared with 65% in the physiotherapy group and 27% in the wait-and-see group. At 3 months, however, this was reversed, with the injection group having only 45% success as opposed to 76% success with physiotherapy or 59% with the wait-and-see approach. This relation was maintained at 1 year, with the corticosteroid group having less success (68%) compared with the physiotherapy group at 94% success and the wait-and-see group at 90% success.

FIGURE 22-2 Primary outcome measure: mean assessor’s rating of severity (visual analog scale), mean pain-free grip (PFG— affected/unaffected expressed as a percentage), and percentage success. Significant differences between study arms at 6 and 12 weeks: ^†corticosteroid injection versus wait-and-see approach; ^‡physiotherapy versus wait-and-see approach; §corticosteroid injection versus physiotherapy; ¶significant difference between corticosteroid and wait-and-see approach on per protocol analysis. Bar graphs represent mean (99% confidence interval) area under curve (trapezium method²⁴) analysis of assessor severity, PFG, and global improvement. *Significant differences between groups (P <0.01).

Adapted from Bisset L, Beller E, Jull G, Brooks P, Darnell R, Vicenzino B. Mobilisation with movement and exercise, corticosteroid injection, or wait and see for tennis elbow: randomised trial. BMJ 333(7575):939, 2006 with permission of the BMJ Publishing Group.

Area under the curve analysis demonstrated a significant advantage in favor of physiotherapy over injection for all primary outcome measures and over wait and see for pain-free grip (mean difference [MD], 534; 99% CI, 3–1065) and assessor-rated severity (MD, 447; 99% CI, 137–758). It also demonstrated a significant advantage for wait and see over injection for global improvement (MD, 2 8.3; 99% CI, 2 15.0 to 2 1.5) and assessor-rated severity (2 337; 99% CI, 2 642 to 2 32).²⁶

The physiotherapy group had the most patients who used no other forms of treatment (79%), followed by the corticosteroid group (51%) and the wait-and-see group (45%).

In conclusion, although there is good evidence (grade A) supporting corticosteroid injection in the short term, there is also good evidence that corticosteroid injections are less successful than wait and see in the long term (grade A).

Physiotherapy

Two studies (see Corticosteroid Injections earlier in chapter) reviewed physiotherapy versus a waitand-see approach^25,26 and had conflicting results. Smidt and researchers²⁵ found nine treatments of pulsed ultrasound, deep friction massage, and a home exercise program over 6 weeks to be better than wait and see over the short and long term but not by a significant amount. Wait and see also had more patients not use any other forms of treatment (79%) than the physiotherapy group (19%). If one excludes physiotherapy done within 6 weeks of the end of the treatment period, however, both groups have similar percentages of patients not using any other forms of treatments (85% in the physiotherapy group and 83% in the wait-and-see group).

Bisset and colleagues²⁶ found eight sessions of elbow manipulations combined with an exercise program for 6 weeks was favorable over wait and see at 6 weeks in terms of the number of patients with complaints completely resolved or much improved, pain-free grip strength and assessor-rated severity but no longer at 1 year. Furthermore, more of the patients in the physiotherapy group used no other forms of treatment (79%) versus 45% in the wait-and-see group.

Evidence is contradictory on the benefits of physiotherapy in LE preventing a clear conclusion. Physiotherapy may help but benefits when found are modest at best (grade I).

Smidt and researchers’²⁵ and Bisset and colleagues’²⁶ studies give the closest approximation of the natural history of LE to date. In both studies, more than 80% of patients had symptoms that were much improved or completely resolved by 6 months with little or no treatment beyond activity modification and patient education.

Brace

Studies on braces are limited by small numbers, short follow-up, and heterogeneity of the braces studied. Conclusions on the role braces play are hard to make based on these data.

Faes and investigators²⁷ compared a dynamic extensor brace (Carp-x; Somas, St-Anthonis, Netherlands) with no brace in a randomized, controlled trial involving 63 patients and found significant differences favoring the brace group in pain, pain-free maximum grip strength, and function of the arm at 12 and 24 weeks. In that article, graphs are presented but no numbers are given.

Jensen and coworkers²⁸ compared an off-the-shelf orthotic (Rehband) with corticosteroid injections in 30 randomized patients over a period of 6 weeks and found no significant differences between the groups in pain, maximum grip strength, dumbbell test, function VAS, or global improvement.²⁹ Similarly, Erturk and colleagues³⁰ compared an epicondylitis bandage with a steroid injection in 36 randomized patients and found no significant difference at 3 weeks in pain on resisted wrist extension and no difference in maximum grip strength. Haker and Lundeberg,³¹