OPTIONS

Symptomatic patients may first present to primary care providers, who must decide which patients are unlikely to respond to conservative therapy and refer them to a surgeon. For the remaining patients, options include activity modifications, oral anti-inflammatory agents, splinting, and steroid injection.⁵ For splinting regimens, the choice of joint for immobilization in the affected digit and length of treatment need to be determined. For injection therapy, decisions include which steroid preparation to use, what volume and location to inject, how many injections to give, and at what time intervals repeat injections should be given. Although division of the A1 pulley is the widely accepted surgical procedure, the surgeon has a choice between open versus percutaneous technique. Ideally, systematic review of the published evidence should allow development of a treatment algorithm that can be followed reliably to achieve the safest, most effective results with the greatest level of patient satisfaction, whereas keeping costs at a minimum. In an effort toward reaching this goal, discussion in this chapter is limited to those comparative studies that provide the highest levels of evidence (Levels I-III) identified after systematic review of the published literature. Higher quality clinical studies are more likely to provide convincing evidence for influencing medical treatment decisions.

EVIDENCE

Few controlled studies comparing splinting with steroid injection have been published. In their prospective study, Patel and Bassini⁴ compared 50 consecutive patients who were treated with metacarpophalangeal (MP) joint splinting with 50 patients who were treated with intrasheath injection of betamethasone (Level of Evidence II). Severity of triggering was graded, and patients with minimal pain, “uneven” movements that did not interfere with hand function, or who were symptom free at the end of the 1-year study period were considered successfully treated. One fourth of the injection group required a single repeat injection 2 weeks after the initial injection. Success was achieved in 66% of the splinted group and 84% of the injected group. Symptoms recurred in 12% of the splinted patients during the 1-year follow-up period; none recurred in the injection group. Patients with symptoms for longer than 6 months before treatment, multiple digital involvement, or triggering at time of presentation faired the worst. The authors comment that splinting was the least effective for trigger thumbs, and that more than 6 weeks of splinting was necessary to achieve success in one third of the patients.⁴

Although splinting is painless and noninvasive, it interferes with normal hand use, prolonged treatment may be necessary, and success relies directly on patient compliance. If a single steroid injection can provide comparable or better success rates, then it might be preferred over splinting, despite the discomfort associated with injection. Controlled prospective trials have compared a single injection of steroid into the flexor tendon sheath with placebo injection of local anesthetic.^7,8 In separate studies by Murphy and colleagues⁷ and Lambert and coworkers⁸ (Level II), a single intrasheath injection of betamethasone or methylprednisolone, respectively, cured 64% and 60% of the patients. Both studies were properly blinded, with the intervention and evaluations for each patient performed by different clinicians. Although the results were statistically significant, it is interesting to note that placebo injection was effective in 20% and 16% of patients in the two studies. No recurrences were reported within the relatively short (one⁸ and four⁷ months) follow-up periods.

Initial published results in the 1950s for corticosteroid injection in the treatment of trigger fingers demonstrated short-term efficacy,^9,10 with more lasting results achieved after the development of longer acting corticosteroid preparations.^10,11 Choice of steroid is usually based on physician preference, and no good comparison studies of steroid preparations have been done. Betamethasone, methylprednisolone, and triamcinolone are common choices. Injection volume and concentration of steroid have also varied widely among studies, with essentially all studies advocating injection into the flexor tendon sheath just distal to the A1 pulley. An interesting controlled prospective study by Taras and researchers¹⁰ (Level II) explored this recommendation by comparing intrasheath with subcutaneous injection. The steroid was mixed with radio-opaque contrast dye. After their assigned injections, the actual location of each injection was verified radiographically. Symptom resolution was achieved in 71% of the subcutaneous group and 52% of the intrasheath group after 27-month average follow-up. Overall, 62% of the patients were treated successfully. Only 30% of the attempted injections into the tendon sheath were completely accurate. Recalculation of the results based on true location of steroid delivery had the following outcome: intrasheath, mixed, or subcutaneous showed no change in success rates. Although the results suggest that subcutaneous injection was as effective as intrasheath injection, the study lacked adequate power (β = 0.92). The study designs for these studies on efficacy of steroid injection are summarized in Table 15-1.

Although the physiologic explanation for greater incidence of trigger finger in patients with diabetes mellitus is unclear, it is usually reported as 4% to 10%, with some reports as high as 23%.¹² Steroid injections have been believed to be less effective in the treatment of diabetic trigger fingers. Stahl and investigators¹³ prospectively compared the efficacy of steroid injections for trigger finger in 60 patients with diabetes with 60 nondiabetic control patients, matched for age, sex, digit involvement, occupation, and available follow-up (Level II). A maximum of three injections at 3-week intervals were allowed, with 12% of the patients with diabetes requiring three injections, whereas only 3% of the patients without diabetes did. Overall, 59% were treated successfully with steroid injection, but the response rate was significantly lower in patients with diabetes (49%) than in those without (76%). Insulin dependence and duration of symptoms before treatment did not affect the results. Long-term follow-up beyond 9 weeks was not provided. Likewise, retrospective review of 66 patients treated for trigger finger by Nimigan and coauthors¹² (Level III) showed 52% response rate to up to three steroid injections, with significantly lower response in patients with diabetes (26%) than in those without diabetes (57%).¹² Again, duration of symptoms and insulin dependence did not affect the results. Nearly all of the patients who underwent surgery either after did not respond to injection therapy or because of personal preference were successfully treated (97%), although 36% experienced complications. Patients with diabetes did not have greater complication rates, and stiffness was the most common postoperative problem. In contrast, Griggs and colleagues¹⁴