Industrial Medicine and Acute Musculoskeletal Rehabilitation. 6. Upper- and Lower-Limb Injections for Acute Musculoskeletal Injuries and Injured Workers

6.1 Clinical Activity: To analyze the risks and benefits of corticosteroid injections for a 40-year-old right-handed carpenter with lateral epicondylitis 

A RECENT LITERATURE REVIEW1 of complications of corticosteroid injections for musculoskeletal conditions including lateral epicondylitis concluded that, among dozens of randomized controlled trials (RCTs), only minor complications were reported. Overall, approximately 15% of patients reported side effects, with 10% reporting postinjection pain (the most common side effect), about 2% reporting skin atrophy, and less than 1% reporting either skin depigmentation, localized erythema with warmth, or facial flushing.

Postinjection exacerbations seem to be of short duration. A recent RCT2 of patients with lateral epicondylitis compared 3 treatments: corticosteroid injection, naproxen, and oral placebo. The study showed that although 62% of patients may initially have increased pain after the injection, by day 4 their pain scores were significantly lower and they were less likely to be taking acetaminophen than patients who received either naproxen or oral placebo. The researchers concluded that patients would perceive the “modest” increased pain postinjection “to be acceptable, given the clear benefits of localized corticosteroid injection treatment over the following few weeks.”2(p333)

Although serious complications were not reported within the recently reviewed RCTs, published case reports have described concerning side effects such as tendon rupture, including rupture at the lateral epicondyle tendon origin.1 However, the literature review1 concluded that the case reports could not prove that corticosteroid injections actually cause tendon rupture, because the underlying tendon pathology that prompts injection may also be the underlying cause (or at least a contributing factor) to subsequent tendon rupture. Even the case reports of tendon rupture after multiple injections at a particular site could possibly reflect cases with more severe preinjection pathology (and thus the failure to respond adequately to a single injection). It still may be prudent, however, to consider multiple injections as a source of increased risk for rupture. Overall, the literature does not provide clear evidence to confirm what constitutes a safe maximal number of corticosteroid injections, although various investigators have opined on this.

Different corticosteroid agents have had varying impacts on posttreatment tendon strength. Mechanical structural defects and tendon rupture occurred more commonly in tendons treated with triamcinolone acetonide than in those treated with methylprednisolone, betamethasone, or hydrocortisone.1

A recent literature review3 concluded that these injections were associated with statistically significant and clinically relevant improvements in pain, global improvement, and grip strength compared with placebo, local anesthetic injections (without steroid), and noninjection treatments. In fact, among the published studies that met the criteria for that review, almost all studies concluded that corticosteroid injections provided more favorable outcomes for all measured parameters (eg, pain, grip strength) in the short term (≤6wk). Longer-term benefit (>6wk) was more difficult to assess because of variable study designs and the lack of high-quality studies. Thus, the researchers concluded that the current literature does not provide a basis for firm conclusions regarding benefit beyond 6 weeks.3

In the context of workers’ compensation injuries, the initial faster onset of improvement clearly provided by the injections may presumably translate into quicker return to work and/or better tolerance for work-related activities during those initial weeks.

Thus, the literature indicates that corticosteroid injections are very effective in the treatment of lateral epicondylitis (particularly during the initial months after injection) and that serious complications of injections are uncommon.

Substances other than corticosteroids have also been injected to treat lateral epicondylitis. There are a few published studies4, 5 on the potential role of botulinum toxin in treating chronic lateral epicondylitis, with mixed results. There is a single, modest-sized study6 with impressive outcomes after autologous blood injection at the lateral epicondyle region. Of 28 patients with refractory lateral epicondylitis despite various nonsurgical treatments who were injected with 2mL of autologous blood at the lateral elbow, 22 (79%) had complete pain relief even with strenuous activity.6 The researchers speculated that the local injection of blood may induce inflammatory changes that improve recovery, although overall the mechanism remains unclear. Further research is needed before any definitive consensus can be reached on these treatment approaches.

6.2 Clinical Activity: To assess the advantages and disadvantages of administering focal corticosteroid injections in a right-handed male secretary with de Quervain’s tenosynovitis at the right wrist. His symptoms failed to resolve after ergonomic modifications of his computer workstation, use of nonsteroidal anti-inflammatory drugs, and use of a thumb spica splint 

A recently published, pooled, quantitative evaluation7 assessed the medical literature for treatment outcomes for de Quervain’s tenosynovitis. The researchers excluded from review any articles that did not meet their diagnostic criteria (documentation of pain at the radial wrist, tenderness at the first dorsal wrist extensor compartment, a positive Finkelstein test) and also excluded articles that did not re-evaluate these same 3 findings after treatment. Only 7 studies met the criteria to be included in this review. These 7 studies included 459 wrists subjected to 1 or several therapeutic modalities. The results were pooled for analysis. The most common intervention was corticosteroid injection alone. Analysis showed that a symptomatic cure was achieved in 83% of the 226 wrists that received injection alone, 61% of the 101 wrists that received injection and splint immobilization, and 14% of the 76 wrists that received splinting alone. Conversely, no such symptomatic cure was achieved in any of the 17 wrists treated with rest alone or in any of the 39 wrists treated with nonsteroidal anti-inflammatory drugs (NSAIDs) alone. Thus, the pooled results seem to indicate that corticosteroid injection alone (without splinting or other treatment) is by far the most effective. The investigators concluded that such injections are a simple, cost-effective, and definitive treatment.7

The complication rate of corticosteroid injection to treat de Quervain’s tenosynovitis is very low. A literature review by Richie and Briner7 of 327 wrists that received 1 or multiple injections (either alone or in combination with splinting) reported no cases of tendon rupture. The reported side effects of injections were 18 cases of skin color changes (eg, hypopigmentation), 16 cases of subcutaneous fat atrophy (all of which were among those same 18 cases of skin color changes), 5 cases of flare, 2 nontender nodules, and a single case of superficial thrombophlebitis.7 Because that article was not explicit on the total number of injections given among the subset of patients receiving multiple injections, it is unfortunately not possible to calculate the exact complication rate per injection. However, there were at least 357 injections, and thus the calculated complication rate would be, at most, 5% for skin color changes, 1.4% for flare, 0.6% for nontender nodules, and 0.3% for superficial thrombophlebitis. The literature contains only 1 case report8 of corticosteroid injection into a first dorsal compartment causing persistent cheiralgia paresthetica (painful dysesthesias in the distribution of the superficial radial nerve).

Thus, overall, the literature strongly supports the use of first dorsal compartment corticosteroid injection for de Quervain’s tenosynovitis, with very favorable outcomes (ie, most patients achieved a symptomatic cure) and very low rates of complications.

6.3 Clinical Activity: To discuss the use of corticosteroid injection into the carpal tunnel for an assembly worker with carpal tunnel syndrome in whom the use of a wrist splint has given inadequate relief 

Injecting corticosteroids into the carpal tunnel can be done by injecting just ulnar to the palmaris longus tendon at the wrist into the ulnar bursa (tenosynovium surrounding the deep and superficial flexors of digits 2–5).

There are some discrepancies in the medical literature about the effectiveness of corticosteroid injection to treat carpal tunnel syndrome (CTS). A recent meta-analysis9 of RCTs concluded that compared with placebo injection, local injection of corticosteroids provided symptomatic improvement of CTS 1 month after injection. Further, the same meta-analysis concluded that compared with oral steroids, local injection of corticosteroids provides significantly greater clinical improvement up to 3 months after treatment.9

Even more recently, a 1-year, prospective RCT in 163 wrists with CTS compared surgical decompression versus local injection of corticosteroids into the carpal tunnel. All clinical diagnoses of CTS were electrodiagnostically confirmed. Subjects with clinically visible thenar atrophy were excluded. The study concluded that over the short term (3mo), local injection of corticosteroids is better than surgical decompression for the symptomatic relief of CTS. At 1 year, local injection is as effective as surgical decompression for the symptomatic relief of CTS.10

Thus, the literature suggests that corticosteroid injection provides at least short-term (1–3mo) symptomatic relief for CTS patients, and may rival surgical treatment even at 1 year. These studies may suggest that, at least for patients with just mild or moderate CTS, corticosteroid injection may be the most appropriate initial treatment (eg, combined with wrist splinting), with surgery reserved for patients with severe or nonresponsive CTS. The caveat may be to provide close medical monitoring to make sure that patients being treated nonsurgically do not progress to the point of irreversible nerve damage and muscle atrophy.

The literature has also indicated that injecting corticosteroids into the carpal tunnel can improve median nerve function, as measured by motor and sensory nerve conduction studies. A recent study11 of CTS treatments showed that corticosteroid injection is superior to iontophoresis and phonophoresis and found that that the most sensitive neurophysiologic parameters in follow-up were the difference between the median and ulnar distal sensory latencies to the fourth digit and also the difference between median distal sensory latency to second digit and ulnar distal sensory latency to the fifth digit.

Other studies have also shown electrodiagnostic improvements in objective median nerve conduction parameters after injection into the carpal tunnel. For example, 1 prospective study12 evaluated clinically mild CTS (defined as intermittent symptoms without thenar atrophy, thenar weakness, or absence of sensations). These 48 clinically mild CTS cases included electrodiagnostically mild, moderate, and severe CTS but excluded the electrically most severe cases (needle electromyography showing abnormal spontaneous activity in the thenar muscles and/or a median nerve distal motor latency greater than 7.5ms). The study showed that at 3 months after the injection, 93.7% of the patients reported marked improvement in their symptoms, with significant improvement in distal motor latencies, distal sensory latencies, symptom severity, and functional scores. Improvements were seen among patients with electrically mild, moderate, and severe CTS. Significant improvement was still present for median distal motor latency at 12 months. Almost 50% of patients achieved normalization in the electrophysiologic study. At an average follow-up of 16 months, 79% of patients continued to have improvement in symptoms. Of patients studied, 16.6% relapsed clinically after an initial response that lasted 7 to 15 months. The researchers concluded that local injection of corticosteroid results in long-term improvement in nerve conduction parameters, symptom severity, and functional scores in patients whose CTS is clinically mild and electrodiagnostically mild, moderate, or somewhat severe.12

One RCT13 examined intracarpal insulin injection to treat CTS. In patients with non–insulin-dependent diabetes mellitus and mild to moderate CTS, 20mg of methylprednisolone was injected into the carpal tunnel. A week after receiving a corticosteroid injection, patients were randomized to receive additional injections into the carpal tunnel, with placebo or isophane insulin (12U), weekly for 7 weeks. Those patients who received insulin injections showed a more significant improvement in mean median nerve distal motor latency, median nerve sensory velocity, and global symptom score. The insulin injections did not appreciably change the overall glucose control, and the mechanism of CTS improvement was unclear. Further research is needed to establish what role insulin injections into the carpal tunnel may have in the nonsurgical treatment of CTS.

6.4 Clinical Activity: To critique the risks and benefits of focal corticosteroid injections in the treatment of Achilles’ tendinitis in a heavy laborer 

The medical literature currently provides no definitive consensus regarding the benefits and risks of focal corticosteroid injections for Achilles’ tendinitis.

Regarding risks, a 1% overall incidence of side effects (including subcutaneous atrophy and depigmentation) was reported by a systematic review14 of 145 published articles related to Achilles’ tendinitis and corticosteroid injections. The reviewers noted that animal studies of intratendinous injections showed decreased tendon strength, thus suggesting increased risk of tendon rupture. Although they found multiple published instances of Achilles’ tendon rupture after injection, most of these were isolated case reports or small case series. One inherent problem with case reports of complications is that they generally fail to have any control group for comparison, thus making it impossible statistically to conclude that the risk of that complication was increased. The reviewers concluded that no published rigorous studies existed evaluating the risk of Achilles’ tendon rupture. Finally, they concluded that insufficient published data exist to determine the comparative risks and benefits of corticosteroid injections for Achilles’ tendinitis.14

Another study15 looked retrospectively at 64 consecutive patients with Achilles’ tendon pain (excluding, eg, ruptures and excluding patients with concomitant peroneal tendonitis) who specifically had Achilles’ pain at rest that improved with activity. Of these 64 patients, 35 chose to undergo corticosteroid injection and 29 chose not to have it not randomized. Subsequently, 2 patients in each group developed tendon rupture, representing 6% of patients who had undergone injection and 8% of those who had not undergone injection. The researchers concluded that injections did not increase the rupture rate.

A recent review article analyzed 9 randomized or quasirandomized trials involving 697 patients with Achilles’ tendinitis.16 There was weak evidence of a modest benefit of NSAIDs for alleviation of acute symptoms. There was weak evidence of no difference (compared with no treatment) of low-dose heparin, heel pads, topical laser therapy, and peritendinous injection of corticosteroids. The reviewers concluded that evidence from RCTs was insufficient to determine the most appropriate treatment of acute or chronic Achilles’ tendinitis.

A number of recent studies have examined the potential role of ultrasound (eg, Doppler) or fluoroscopy for image guidance during injections for Achilles’ tendon pathology. Theoretically, better outcomes (improved benefits and/or decreased side effects) could be obtained by more precise peritendinous placement of the corticosteroid along the tendon sheath, rather than inadvertent injection into the substance of the tendon itself. A recent retrospective cohort study17 examined fluoroscopically guided injections along the anterior aspect of the Achilles’ tendon for Achilles’ tendinopathy: 43 patients had undergone injections, with a collective total of more than 80 injections (including 1 patient having 14), and all subjects had at least 2 years of follow-up after injection. There was 1 minor complication (persistent purplish skin discoloration after 2 injections), but no tendon ruptures or other major complications were reported. Of the patients, 17 (40%) reported improvement after the procedure(s), 23 (53%) thought that their condition was unchanged, and 3 (7%) thought that their condition was worse than it had been before the injection.

Although the literature is increasingly reporting the use of such image guidance for Achilles’-related injections, the literature currently lacks any substantial studies comparing the outcomes with and without it, thus leaving it to each individual clinician to decide whether to use such imaging techniques.

Other published articles have examined treating Achilles’ tendinitis with injections of agents other than corticosteroids. These potential alternatives include injection of sclerosing agents into newly proliferated blood vessels (neovessels),18 injection of viscosupplementation agents (eg, Hylan G-F 20),19 and a variety of other biologic substances. Further research is needed to determine what role these agents have in the management of Achilles’ tendinitis.

6.5 Clinical Activity: To evaluate the role of corticosteroid injections in the management of persistent plantar fasciitis in a salesperson working at a home improvement warehouse. The proximal plantar pain has progressed to the point where it is painful not only with the first steps in the morning but also throughout the workday, which involves prolonged standing and walking on concrete floors 

Overuse and microtears of the plantar fascia may contribute to plantar fasciitis, because histologic studies show collagen degeneration and necrosis.20 Because the histologic findings do not universally show inflammatory changes, some clinicians prefer the term plantar fasciopathy or fasciosis rather than plantar fasciitis.20 The possible lack of inflammation is germane to the discussion of injecting anti-inflammatory agents such as corticosteroids. Still, corticosteroid injections are commonly used as part of the nonsurgical treatment for plantar fasciitis. More than 80% to 90% of patients respond to nonsurgical care, including corticosteroid injections.20 It has been suggested that corticosteroids may be beneficial because of their antinociceptive and membrane-stabilizing properties rather than just their anti-inflammatory properties.21

A recent RCT of plantar fasciopathy treatment compared extracorporeal shockwave therapy (ECSWT) with corticosteroid injection, with 3- and 12-month follow-ups. The investigators concluded that the injections were more efficacious (significantly lower pain scores, per visual analog scale) and were much more cost effective than ECSWT.22

One potential drawback to corticosteroid injection for plantar fasciitis is the pain of the injection. A recent RCT of fluoroscopically guided corticosteroid injection for plantar fasciitis compared performing the injection with versus without an antecedent posterior tibial nerve block with 5mL of 1% lidocaine.23 Except for a mild burning sensation, pain during posterior tibial block was negligible in all cases. Among patients who had undergone the posterior tibial nerve block, the plantar injection was considered painless in 90% (9/10) and just mildly painful in 10% (1/10). Among patients who had not received the nerve block, the plantar injection was considered severely painful in 70% (7/10) and moderately painful in 30% (3/10). Among those who received the nerve block, no patients had complications (despite an increased number of injections per patient, because patients receiving nerve blocks were more likely to allow the injection to be repeated). Among those who did not receive the nerve block, 10% (1/10) of patients developed fat necrosis and 10% (1/10) developed pressure periostitis.

Plantar fascia rupture has been reported as a possible complication of corticosteroid injections. In a retrospective, nonrandomized study of 122 patients who received 1 or more injections, 12 (10%) patients developed plantar fascia ruptures.20 An overlapping patient population in that study showed that of 51 patients with plantar fascia rupture, 43 (84%) had received 1 or more injections into the calcaneal origin of the fascia.20 The retrospective, nonrandomized nature of this and similar studies makes it impossible to conclude that the corticosteroid injection actually causes plantar fascia rupture, because perhaps patients with more severe pathology (and hence a higher likelihood for rupture) were more likely to receive injections in the first place. Still, based on the number of case reports of plantar fascia rupture after injection, it seems prudent during informed consent to advise patients of this possible risk.

Plantar fat pad necrosis has been reported as another possible serious complication of corticosteroid injection for plantar fasciitis.22

Until further research clarifies the role of image guidance (eg, via fluoroscopy or ultrasound) in the performance of plantar fascia injections, clinicians must use their own discretion in deciding whether to use such guidance in a given injection.