To The Editor:
In "Single-Portal Endoscopic Carpal Tunnel Release Compared with Open Release. A Prospective, Randomized Trial" (2002;84:1107-15), Trumble et al. used questionnaires, standardized functional tests, and physical measurements to compare the results of single-portal endoscopic carpal tunnel release with those of open carpal tunnel release. The results of the studies were subjected to statistical analysis.
We have several comments and questions. Differences that are statistically significant are not necessarily clinically significant. Do the authors believe that the statistically significant differences between the endoscopic group and the open-release group are clinically significant? For example, do the authors believe that the statistically significant difference in the results of the Purdue pegboard test at four weeks is clinically significant?
The authors reported that patients who underwent endoscopic carpal tunnel release returned to work sooner than those who underwent open carpal tunnel release. Return to work is an inexact measure of success. Louis
l pointed out that "no objective, definable, measurable endpoint has been used to assess appropriateness for return-to-work status. The decision to return to work is largely physician-driven, and may be biased by patient motivation..." Other authors have reported no significant difference in the time to return to work between patients who underwent endoscopic carpal tunnel release and those who underwent open carpal tunnel release
2-4 . How did Trumble et al. decide when to advise a patient to return to work? What did Trumble et al. tell their patients about the nature of the study, and might this in any way have influenced return to work? Also, in the first paragraph of their study, Trumble et al. referenced a paper by Adams et al. in which only 67% of patients with occupational carpal tunnel syndrome were able to return to their original job
5 . The poor rate of return to work appears to have been offered by Trumble et al. as a rationale for the endoscopic carpal tunnel release procedure. How did patients with a Workers' Compensation claim fare in the study by Trumble et al.? How do results of carpal tunnel release, open and endoscopic, in patients with a Workers' Compensation claim in the study by Trumble et al. compare with the results reported by Adams et al.? If there are differences that appear meaningful, how do Trumble et al. account for those differences? Is it the procedure that is most likely to determine return to work or is it some other factor?
The authors reported that two weeks after surgery, grip strength had decreased significantly in both groups compared with preoperative measurements but was greater in the endoscopic group. The patients in the endoscopic group then recovered grip strength faster than did those in the open-release group until the three-month follow-up examination, when grip strength in both groups approached preoperative levels. If the endoscopic carpal tunnel release group started out ahead of the open carpal tunnel release group at two weeks after surgery and then went on to recover faster, how did they end up the same at about three months? When did the grip strength in the open-release group catch up?
In the second paragraph of their Discussion section, the authors stated: "Analysis of the primary outcomes in this study demonstrates that the patients who had undergone endoscopic release had greater relief of symptoms, improvement in function, and satisfaction for the first three months following the surgery." This statement, if taken out of context, has the potential to be misleading. As the authors stated in their Results section and in Table I, the only time that there was any statistically significant difference in patient satisfaction was at the two-week mark. After this, there was no statistically significant difference in patient satisfaction. Regarding patient satisfaction, one can argue that this is the best measurement of outcome (in this study or in any study). The objective measurements may help us to understand what our treatment is doing to the patient, but from the patient's point of view objective measurements may mean little or nothing. After all, a patient comes to us with subjective complaints and wants to be made better. From the patient's point of view, it appears that the differences between endoscopic and open carpal tunnel release are not as great as they are from the authors' point of view.
T.E. Trumble replies:
We appreciate the interest that Dr. Kuschner and Dr. Lane have expressed in our study. They bring up two key points that interest surgeons treating carpal tunnel syndrome: return to work and patient satisfaction. It is important to remember the power of a prospective, randomized, multicenter study. This helps to control many of the variables that Dr. Kuschner and Dr. Lane are concerned about. Because orthopaedic surgeons have so little experience with prospective, randomized studies, there is often confusion about how to interpret the data. Prospective, randomized studies provide the greatest statistical power. We would like to provide some background information regarding the study. We responded to a request for randomized studies issued by the Orthopaedic Research and Education Foundation. We generated a grant that was subjected to rigorous peer review. In addition, this grant was subjected to an equally rigorous review by our institutional review boards, and the article was subjected to a very thorough review by the editorial staff of The Journal of Bone and Joint Surgery. In order to obtain adequate patient numbers and follow-up, the clinical study required an investment of approximately four years for our research program.
With regard to return to work, all patients were advised that they could return to work as soon as their sutures were removed. A soft dressing was applied postoperatively, and the patients were allowed to change their dressing in forty-eight hours.
By randomizing the patients into two groups, we were able to control for many of the psychological and social factors that are involved in return to work, including issues related to Workers' Compensation claims.
Our study, involving 192 hands in 147 patients; the study by Brown et al.
6 , involving 169 hands in 145 patients; and the study by Agee et al.
7 , involving 127 hands in 122 patients, all demonstrated a significant difference in the time for return to work by patients treated with endoscopic surgery either with the single-portal technique (as in the study by Agee et al. and our study) or with the double-portal technique (as in the study by Brown et al.). Palmer et al.
8 compared single-portal endoscopic, two-portal endoscopic, and open carpal release and found that patients treated with endoscopic surgery had faster recovery of wrist motion, grip strength, and pinch strength than did patients treated with open surgery. Patients treated with single-portal endoscopic surgery returned to work sooner than patients treated with either the two-portal technique or the open technique. Palmer et al. mentioned persistent palmar scar tenderness as a significant factor in these findings.
Dr. Kuschner and Dr. Lane provided a number of references indicating that there is no significant difference in return to work between the two techniques. Of these references, we would like to note that Jacobsen and Rahme
3 evaluated only thirty-two hands in twenty-nine patients treated with the two-portal technique. Furthermore, in a letter to the editor, Burge
9 noted that the statistical analysis performed by Jacobsen and Rahme was flawed. Skoff and Sklar
4 compared twenty consecutive two-portal endoscopic releases with their last twenty open releases in a nonrandomized trial and found no significant difference in the time to return to work. Gibbs et al.
2 compared thirty-four patients treated with the open technique with only twelve patients treated with the endoscopic two-portal technique in a nonrandomized, nonprospective study. They also did not find any significant difference in the time to return to work.
In our study, we included the reference by Adams et al.
5 to highlight a concern regarding the ability of patients treated with carpal tunnel surgery to return to work. Those authors evaluated every patient treated in the state of Washington who had a Workers' Compensation claim for a given quarter and noted that only 67% returned to the job where they had sustained the injury. We do not suggest that endoscopic surgery alone can improve the percentage of patients who return to the job where they sustained the injury. A number of factors, including obtaining accurate electrodiagnostic studies prior to performance of the surgery, have helped to improve the percentage of patients returning to work. In our study, every patient treated with either an open or an endoscopic technique was able to return to the job where they sustained the condition. As we noted in the article, the carpal tunnel syndrome was confirmed by electrodiagnostic studies in all patients in our study. We would like to emphasize that it is important to compare our study with other studies of single-portal techniques. It is also important to compare our study with other prospective, randomized studies rather than with retrospective, nonrandomized studies. Finally, it is important to evaluate studies in which an adequate power analysis was performed.
With regard to patient satisfaction, as we pointed out in our article, we compared two techniques that are each highly successful. Patients treated with the endoscopic technique had higher satisfaction scores, higher scores as measured with validated outcome tools, and less scar sensitivity during the first three months after surgery than did those treated with the open technique. The patients in the endoscopic group also had better grip strength and pinch strength for the first two months after surgery. Not until the third month after surgery did patients with open carpal tunnel release make significant gains in grip strength and pinch strength. We specifically designed this study so that patients with bilateral carpal tunnel syndrome would have the same surgery on both sides. We did this after observing the phenomenon noted by Agee et al.
7 , who had difficulties in the early phases of their study because patients with bilateral carpal tunnel syndrome refused to have open carpal tunnel surgery if endoscopic surgery was performed first. In the states of Washington and California, where our study was carried out, patients preferred not to have simultaneous bilateral surgery. However, this is not the case in the United Kingdom. In a striking study reported to the British Hand Society by Giele et al.
10 , patients with bilateral carpal tunnel syndrome were randomly treated with open surgery in one hand and single-portal endoscopic surgery in the other. Sixty-two patients were enrolled in the study. Of forty-seven patients evaluated at an average of fifty-six days after the surgery, thirty-eight (81%) preferred the side treated with the endoscopic technique and nine preferred the side treated with the open technique. Furthermore, during the first twelve days after surgery, patients had significantly lower pain scores for the hand treated with the endoscopic technique.
In summary, we think that it is important to perform prospective, randomized trials to evaluate surgical procedures as well as rehabilitation techniques. It is also crucial to obtain adequate patient follow-up. Furthermore, our study is the only prospective, randomized trial of patients treated with carpal tunnel surgery evaluated with validated outcome tools over the course of one year.
To The Editor:
I was pleased to see that in their study, "Single-Portal Endoscopic Carpal Tunnel Release Compared with Open Release. A Prospective, Randomized Trial" (2002;84: 1107-15), Trumble et al. had performed a prospective, randomized trial to answer the question about the best technique for carpal tunnel release. Unfortunately, I was ultimately disappointed. Although most of the protocol for this study was carefully defined, the authors did not describe the occupations of their patients to ascertain if the two study groups were similar in this regard. More importantly, the authors did not indicate the criteria used for return to work or whether they measured return to modified duties, full duties, or either. Even though postoperative assessments were done by a research assistant who was blinded to the type of release performed, the surgeon and the patient were obviously not blinded when it came time to decide about returning to work. This imposes a bias if the surgeon is even unconsciously promoting one procedure over the other or if the patient has a preconceived notion about returning to work based on the experience of coworkers who have had carpal tunnel release. This bias is compounded by the fact that forty-five of the 147 patients in the study had bilateral carpal tunnel syndrome and had the same procedure (open or endoscopic) on the second hand. The patients with bilateral carpal tunnel syndrome (nearly a fourth of those studied) already had personal experience with postoperative recovery and return to work after carpal tunnel release, which would contribute to any opinion that the surgeon or coworkers might provide.
I hope that the authors can respond with a list of objective and strictly followed criteria that they used to determine when to advise their patients to return to work and that they will review their considerable data to exclude from their results workers with second-side surgery.
T.E. Trumble replies:
Clearly, Dr. Meals has a great deal of experience in hand surgery. As someone with experience with prospective, randomized, multicenter trials, I would like to point out that most of Dr. Meals' concerns were addressed by the statistical randomization that occurs during such a study. In one of our earlier submissions to
The Journal of Bone and Joint Surgery, we did describe the distribution of labor in each study group with use of standards established by the Washington State Labor and Industry Department. This included a classification regarding heavy, medium, and light labor. There was no significant difference between the distributions in the two groups. With regard to return to work, the patients received identical instructions in each study group. They were advised that they could return to work after the removal of their sutures. It was not the practice of any of the physicians in the study group to force patients to return to work against their wishes.
The study was completed in the states of Washington and California, with patients enrolled at the University of Washington Medical Center; the University of California, San Francisco, Medical Center; or the University of California, San Diego, Medical Center. In these states, the patients preferred to undergo unilateral rather than bilateral carpal tunnel surgery. Our experience was similar to that of Agee et al.
7 , who reported that patients preferred to have staged operations rather than bilateral operations. We specifically set up the research design so that the patients would receive identical surgical procedures in both hands. In the early phases of the study performed by Agee et al., the patients refused to have open surgery as the second surgery after they had had endoscopic surgery as the first surgery. Simultaneous, bilateral carpal tunnel surgery is common in the United Kingdom. In a dramatic study performed by Giele et al.
10 , the authors randomized patients with bilateral carpal tunnel syndrome to be treated with open surgery in one hand and with endoscopic surgery in the other. They found that, at fifty-six days after the surgery, 81% of the patients preferred the side treated with endoscopic surgery. They had significantly lower pain levels on the side of the endoscopic release during their first twelve days after surgery. However, because the patients had bilateral surgery, it was obviously impossible to determine which side would allow them to return to work sooner.
In our study, we noted that there was still a significant difference in return to work (p < 0.01) when patients with bilateral surgery were excluded. The patients treated with the open method returned to work at an average of forty days compared with eighteen days for the endoscopic group.
I hope that this information is useful to the readers of
The Journal. We underscore the importance of prospective, randomized, multicenter trials, especially with regard to surgical treatment.