Surgical Treatment of Carpal Tunnel Syndrome: A Health Technology Assessment
Health technology assessment
|Published
We conducted a health technology assessment of surgical treatment compared to non-surgical treatments with splinting, combinations of non-surgical treatments, local corticosteroid injection and physical therapy (including manual therapy).
Key message
Carpal tunnel syndrome is a compressive neuropathic disorder, leading to typical symptoms of numbness, tingling and pain in the hand. Surgical treatment with release of the transverse ligament of the wrist is an established treatment. In Norway, a geographic variation in the provision of surgery has been observed.
We conducted a health technology assessment of surgical treatment compared to non-surgical treatments with splinting, combinations of non-surgical treatments, local corticosteroid injection and physical therapy (including manual therapy). For evaluation of efficacy and safety, we included 10 randomized controlled trials. We found that:
- Surgery may slightly improve symptoms and hand function compared to splinting and combinations of non-surgical treatment in patients with mild to moderate carpal tunnel syndrome (low-certainty evidence).
- The efficacy of surgery compared to corticosteroid injection and surgery compared to manual therapy is uncertain due to very low-certainty evidence.
- No trials reported outcomes separately for subpopulations with mild, moderate or severe carpal tunnel syndrome.
- Overall, few adverse events were reported after the surgical and non-surgical treatments.
- Surgery in patients with mild to moderate carpal tunnel syndrome is the most costly treatment alternative at about Norwegian kroner (NOK) 11,200. Non-surgical treatment alternatives with splinting and local corticosteroid injection cost around NOK 3,100.
Potential cost savings depend on the reduction of surgical procedures per health region. We estimated the potential cost savings at the national level to be between NOK 14.5 and 27.5 million over five years.
Summary
Background
Carpal tunnel syndrome is a clinical syndrome characterized by symptoms and signs of irritation or impairment of the median nerve at the level of the wrist. Typical symptoms are numbness, tingling and pain in the hand, sometimes also in the arm and shoulder. Although surgical treatment with release of the transverse carpal ligament is an established treatment, the relative benefits and risks compared with non-surgical treatments is unclear.
Objective
The objective of this health technology assessment is to summarize the current knowledge on the efficacy and safety of decompression surgery for carpal tunnel syndrome compared to non-surgical treatments with wrist splinting, combinations of non-surgical treatments, local corticosteroid injection (steroid injection) and physical therapy (including manual therapy). We planned to analyse outcomes based on pre-treatment severity of carpal tunnel syndrome (mild, moderate, and severe) if reported, to evaluate whether some subpopulations seem to benefit more from surgery than others.
Methods
We developed a project plan with input from the external experts and patient representatives. We searched for systematic reviews in the Cochrane Database of Systematic Reviews (Wiley), Epistemonikos (Epistemonikos Foundation), INAHTA (International Network of Agencies for Health Technology Assessment), MEDLINE (Ovid) and Embase (Ovid), and for randomized controlled trials (RCTs) in MEDLINE (Ovid), Embase (Ovid), and Cochrane Central (Wiley) up to December 2020. Additionally, we identified randomized controlled trials from the systematic reviews.
We included randomized controlled trials comparing surgery with the selected non-surgical treatments in individuals aged 18 years or older, regardless of other comorbidities or severity of carpal tunnel syndrome. Our primary outcomes were symptom severity including paraesthesia and pain, functional impairment, and health-related quality of life. Secondary outcomes included adverse events. Data from 6 months, 1 year, 2 years and 5 years were retrieved, and primary time point of interest was set at 1 year. Two researchers independently selected trials for inclusion and assessed risk of bias of the included randomized controlled trials according to the Cochrane Handbook for Systematic Reviews of Interventions (ROB1). One researcher extracted data, and one checked the accuracy of the data. We calculated measures of effect as mean difference (MD) or standardized mean difference (SMD) with 95% confidence interval (CI) for continuous outcomes, and risk ratio (RR) with 95% CI for dichotomous outcomes. We merged data into meta-analyses when possible, and we present data as forest plots if appropriate. We assessed certainty of evidence for the primary outcomes at 1 year with Grading of Recommendations Assessment, Development and Evaluation (GRADE).
Due to great uncertainty in the estimates of the relative efficacy, we conducted a simplified assessment of economic consequences in the form of a cost analysis where the costs of the relevant non-surgical treatment alternatives in Norway, i.e., steroid injection and splinting, were estimated and compared to the costs of surgery for patients with mild to moderate carpal tunnel syndrome. In addition, we conducted a simple budget impact analysis to show possible cost savings.
Results
We included 10 randomized controlled trials (13 publications) with 960 patients or wrists with carpal tunnel syndrome. Mean age ranged from 41 to 53 years, and 51 to 100% of the participants were females. Seven trials excluded patients with severe carpal tunnel syndrome. All trials were at high risk of bias for at least two domains; lack of blinding of participants and assessors. Some trials had serious methodological concerns. None of the trials reported outcomes according to pre-treatment severity with mild, moderate, and severe carpal tunnel syndrome.
Three trials compared surgery with splinting. At 1 year, low-certainty evidence from one trial (downgraded for bias and imprecision) suggested a superior, but small, effect of surgery on symptom severity, daytime paraesthesia, and function. For symptoms, which was considered as the most important outcome by the patient representatives and the project’s clinical experts, patients reported less severity after surgery than after splinting; the standardized mean difference between groups from the intention to treat (ITT) analyses was -0.47 (95% confidence intervals, CI -0.78 to -0.15). Notably, 38% of the patients allocated to splinting had undergone surgery at 1 year.
Three trials compared surgery with non-surgical treatments. At 1 year, low-certainty evidence from one trial (downgraded for bias and imprecision) suggested a superior, but small, effect of surgery on symptom severity and hand function, and little or no difference in effect on pain. Mean difference between groups in symptom severity was -0.33 points (95% CI -0.65 to -0.01). In this trial 44% of the patients allocated to non-surgical treatment had undergone surgery at 1 year. “As treated” analyses revealed a larger mean difference between groups; -0.84 (95% CI -0.55 to -1.13) points.
Two trials compared surgery with steroid injections and three trials compared surgery with manual therapy. We are uncertain of the efficacy of surgery compared to steroid injection and of surgery compared to manual therapy very low-certainty evidence (downgraded for bias and imprecision).
Overall, few serious adverse events were reported, but rare adverse events such as complex regional pain syndrome did occur after surgery.
The results of our economic evaluation showed that surgery is the most costly treatment at Norwegian kroner (NOK) 11,200 for treatment of patients with mild to moderate carpal tunnel syndrome. The non-surgical treatment alternatives splinting and local steroid injection cost approximately NOK 3,100. We estimated potential cost savings at the national level to be between 14.5 and 27.5 million NOK kroner over five years.
Discussion
Current evidence is insufficient to draw firm conclusions about the efficacy of surgery compared to non-surgical treatments. Low-certainty evidence demonstrated a superior effect of surgery compared to splinting and combinations of non-surgical treatments in patients with mild to moderate carpal tunnel syndrome, but the effect sizes were small, and the clinical relevance is therefore uncertain. An important finding from the trials was that a substantial proportion of patients allocated to non-surgical treatment had been treated with surgery at 1 year, suggesting that the patients were not satisfied with the conservative means. A planned Norwegian trial and other ongoing trials may further elucidate the relative efficacy of surgery to steroid injections. The trials of manual therapy were conducted at one single centre and other studies are needed to confirm the findings from these trials.
We have conducted a simple analysis of the costs associated with the treatments in the short term. Therefore, we have not included subsequent treatment after initial treatment. If more reliable evidence becomes available, the long-term effect of the different treatment alternatives should be evaluated in a model-based analysis.
There is moderate geographical variation in the provision of surgery for carpal tunnel syndrome in Norway. The need for surgery should be expected to be the same irrespective of geographic area. According to clinical experts, patients with mild to moderate carpal tunnel syndrome could initially have non-surgical treatments. We therefore performed a simple budget impact analysis to show possible cost-savings. However, there is uncertainty regarding this analysis. We did not have information about the distribution of disease severity in the surgically treated population in Norway and therefore we are uncertain about how many patients who could have been treated conservatively instead. This may have over- or underestimated the feasibility of the non-surgical treatments in the patient population used.
Conclusion
Overall, low-certainty evidence suggests that decompression surgery is slightly more effective than splinting and combinations of non-surgical treatments at 1 year in patients with mild to moderate carpal tunnel syndrome. There is insufficient evidence regarding the efficacy of surgery compared to steroid injections and manual therapy. Overall, few serious adverse events were reported, but small randomized trials are not suitable to make reliable comparisons of adverse events. Surgery is the most costly treatment alternative for patients with mild to moderate carpal tunnel syndrome. An observed regional variation in the provision of surgery in Norway suggests a potential for cost-saving per health region and at the national level if patients with mild to moderate carpal tunnel syndrome are initially treated with the non-surgical alternatives.
