- Email: [email protected]
Early Outcome and Cost-effectiveness of Endoscopic versus Open Carpal Tunnel Release: a Randomized Prospective Trial
EARLY OUTCOME AND COST-EFFECTIVENESS OF ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE: A RANDOMIZED PROSPECTIVE TRIAL N. L. B. SAW, S. JONES, L. SHEPSTONE, M. MEYER, P. G. CHAPMAN and A. M. LOGAN From the Departments of Orthopaedic and Plastic Surgery, Norfolk and Norwich University Hospital, Norwich, UK
Proponents of endoscopic carpal tunnel release have been advocating the technique for more than 10 years but there is still debate about its efficacy, safety and cost-effectiveness. We have performed a randomized, prospective, blind trial to compare early outcome after single portal endoscopic or open carpal tunnel surgery and to assess the cost-effectiveness of the procedures. There were no significant differences in symptom and functional activity scores, grip strength or anterior carpal pain in the first 3 months. For those in employment, we found a statistically significant difference between the two treatment groups with the endoscopic group returning to work, on average, 8 (95% CI, 2–13 days) days sooner than the open group. This translates into a cost saving to industry. There were no major neurovascular complications in either group. On the basis of these findings, we recommend that endoscopic carpal tunnel release should be considered in the employed as a cost-effective procedure, but perhaps not in the general population as a whole. Journal of Hand Surgery (British and European Volume, 2003) 28B: 5: 444–449 Keywords: endoscopic carpal tunnel release; cost-effectiveness
Between 1998 and 2001, 162 patients presenting to the Norfolk and Norwich or West Norwich Hospitals with a diagnosis of carpal tunnel syndrome were invited to take part in the trial. Of these, 123 patients gave informed consent and were enrolled into the study. The diagnosis of carpal tunnel syndrome was made on clinical grounds. Nerve conduction tests were performed only when there was clinical doubt. Exclusion criteria included those with previous wrist fractures and surgery, patients with rheumatoid arthritis and other inflammatory conditions, a suspected mass lesion or any patients requiring additional procedures such as biopsy and those less than 16 years of age. The patients were randomized using block randomization and sealed envelopes to either the endoscopic (ECTR) or open (OCTR) groups. All procedures were performed under local anaesthetic and tourniquet control by Consultant Hand Surgeons who are enthusiasts of their technique. OCTR was performed with a standard open approach with a 2 cm palmar incision (Fig 1) by one surgeon and ECTR by two surgeons of similar experience according to the extrabursal technique described by Agee (Agee et al., 1992) using the MicroAire CTRS single-portal system (MicroAire, USA). Post-operatively, dressings were applied and the patients advised to use the hand as comfort allowed. The patients were assessed immediately pre-operatively and at 1, 3, 6 and 12 weeks post-operatively. At each assessment, Levine et al.’s. (1993) Symptom Severity and Functional Status Scales questionnaires were completed to assess pre-operative severity, the subjective response to surgery and its impact on
INTRODUCTION Carpal tunnel syndrome is one of the most common conditions to affect the hand, with an incidence of carpal tunnel decompression of between 43 and 74 per 100,000 per year in the United Kingdom (Burke, 2000). Techniques of endoscopic carpal tunnel release (ECTR) have been in use for more than 10 years (Agee et al., 1992; Chow, 1993) and may result in a shorter postoperative recovery period and an earlier return to work. However, there are concerns that ECTR carries a higher risk of nerve injury than open carpal tunnel release (OCTR) and some doubt remains as to whether the benefits outweigh the risks (Evans, 1994). It is also important to consider the cost-effectiveness of any intervention (Burke, 2000). Although there are numerous articles published about ECTR, a search through the Cochrane Controlled Trials Register showed only two studies out of 20 listed that were performed in the United Kingdom; one (Erdmann, 1994) used a twoportal technique and the other (Ferdinand and MacLean, 2002) used a single-portal technique. No attempt has been made to establish the cost-effectiveness of ECTR in the UK. We have therefore compared the single-portal and open techniques and determined the cost-effectiveness of the procedures.
PATIENTS AND METHODS The Norwich Local Research and Ethics Committee granted ethical approval for this trial. 444
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE
445
Cost-effectiveness analysis The costing of each technique was predicated on the assumption that resource utilization in terms of surgeon, theatre staff, anaesthetic and theatre equipment was the same for both procedures except for the cost of the endoscopic equipment. The most expensive item was the disposable blade at d82 (h123). Provided the time taken to perform the procedures was similar, a costminimization analysis could be done. The main economic benefit of ECTR is a shorter time to return to work, and this was our primary efficacy endpoint. Activities of daily living (ADL) are always restricted in the post-operative period and the Levine functional questionnaire assessed this subjectively. We also used specific questions to detect increased dependence on the family, friends and community health services during the recovery from carpal tunnel release (CTR). Statistical methods
Fig 1 Skin markings for incision used for OCTR.
activities of daily living. A visual analogue scoring system was used to assess subjective anterior carpal tenderness in the heel of the hand. Further questions enquired whether any input in terms of domestic support and social services were required in the recovery period. Grip strength was measured using a Jamar Hand Dynamometer (Clifton, NJ, USA). The assessor was blinded to the procedure by placing a stockinet over the wrist of the patient. Patients were advised to return to work when they themselves felt that the discomfort in their hand had improved sufficiently to allow safe practice at their workplace. Sick notes were obtained by the patients from their General Practitioners. The dates of surgery and return to work as reported by the patient were used to calculate the number of days off work. Patients with bilateral carpal tunnel syndrome had their releases sequentially with the more affected hand first. Once these patients felt that they could use the operated hand normally they underwent the second procedure. The shortest interval between the two procedures was 7 months.
The primary endpoint was days until return to work. An intention-to-treat (ITT) analysis was used. Patients were analysed as randomized to the study and not with respect to the surgical procedure actually used. With respect to measures repeated over time, a ‘‘last observation carried forward’’ strategy was used to impute missing values. An area-under-the-curve (AUC) analysis was performed for the repeated measures. Two-sample t-tests or Wilcoxon–Mann–Whitney tests were carried out to test for between group differences, depending on whether the data were parametric or not. A significance level of 5% was considered statistically significant. Ninety-five per cent confidence intervals were constructed for estimated mean or median differences between the two groups.
RESULTS Seventy-four patients were randomized to the endoscopic group and 76 to the open group. Of the former, 9 (12%) were converted to open. No individual randomized to the open group received endoscopic surgery. A comparison of characteristics is summarized in Table 1. The patients in the endoscopic group were, on average, about 5 years older than those in the open group and were more likely to have the left hand operated on. Apart from these differences, the two groups appeared similar pre-operatively. The efficacy analysis is presented in Table 2. Fortythree patients (58%) in the endoscopic group and 42 (55%) in the open group were in employment at the time of the study. Analysis of the primary endpoint could only be carried out on these individuals, and this showed a statistically significant difference between the two groups (P ¼ 0:005; two-sample t-test). Those in the
446
THE JOURNAL OF HAND SURGERY VOL. 28B No. 5 OCTOBER 2003
endoscopic group returned to work, on average, 8 days (95% CI, 2–13 days) quicker than the open group. A general linear model was constructed to test for a between-group difference in the number of days taken to return to work whilst adjusting for age, side of hand operated (left or right) and occupation (manual and Table 1—Comparision of baseline characteristics Randomization group Endoscopic (n ¼ 74)
Open (n ¼ 76)
Age Mean (Std. dev.)
54 (15)
50 (15)
Sex Males
21 (28%)
19 (25%)
Hand side Left
30 (41%)
26 (34%)
Hand side Dominant
43 (58%)
48 (63%)
Employed
43 (58%)
42 (55%)
7 (16%)
6 (15%)
Self-employed Grip strength (kg) Mean (Std. dev.)
26 (12)
27 (11)
Levine functional score1 Median (S-IQR)
13.5 (4.0)
15.0 (4.8)
Levine symptom score2 Median (S-IQR)
19.5 (2.5)
20.0 (3.0)
1
non-manual), but these adjustments made no difference to our findings. The operation time for the endoscopic group was, on average, 2 min (95% CI, 1–4 min) longer. Data imputation in the remaining endpoints was required for four patients in the endoscopic group (2.7%) and two patients in the open group (1.3%). None of these endpoints were statistically significantly different between groups or near statistical significance. A plot of these secondary endpoints over time is provided in Fig 2. Complications There were no cases of permanent neurovascular injury in either group. One patient in the ECTR sustained transient numbness affecting the index finger that settled after 3 weeks. One patient in the OCTR group developed hyperaesthesia over the scar area that had improved by 3 months. There was one superficial wound infection in each group requiring antibiotics and one superficial haematoma in the OCTR group following a fall in the first week which resolved with no intervention. There was one case of incomplete release of the flexor retinaculum in the ECTR group that required open release at 3 months because there was no improvement in his symptoms. There was also one patient in the OCTR group with persistence of symptoms after surgery. Cost-effectiveness analysis
Levine functional score: range 8 (no impact on function) to 40 (unable to perform activities). 2 Levine symptom score: range 11 (asymptomatic) to 55 (severe symptoms).
Based on our assumptions as described earlier, the increased cost of ECTR was due to the initial capital expenditure for the basic ECTR equipment (d6482) and
Table 2—ITT efficacy analysis Randomization group Endoscopic (n ¼ 74)
Open (n ¼ 76)
Estimated difference
95% CI1
Significance2
Primary outcome Days off work Mean (Std. dev.).
18 (11)
26 (14)
8
13, 2
P=0.005
Secondary outcomes Operation time Mean (Std. dev.).
15.8 (3.5)
13.4 (3.4)
1.3, 3.5
Po0.001
22 (7)
24 (6)
2
4, 1
P=0.176
AUC analysis Anterior carpal tenderness3 Mean (Std. dev.). Grip strength (kg) Mean (Std. dev.). Levine functional score Median (S-IQR) (Range 88–440) Levine symptom score Median (S-IQR) (Range 121–605) 1
253 (121)
231 (102)
22
14, 60
P=0.213
109 (22)
108 (24)
1
10, 10
P=0.984
120 (21)
119 (19)
2
9, 13
P=0.701
95% confidence interval either for the difference in means or median difference as appropriate. Resulting from a two-sample t-test or the Wilcoxon–Mann–Whitney test as appropriate. 3 Range 11 (painless) to 55 (severe pain). 2
2.4
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE
447
Fig 2 (a) Grip strength (kg) by time; (b) Levine functional score by time; (c) Levine symptom scores by time; (d) anterior carpal tenderness by time.
to the cost of the non-reusable blade assembly (d82, h123). With the servicing costs of the equipment included, we estimated the incremental cost per case of ECTR due to extra equipment at d98 (h147). We found a small, but statistically significant, average difference of 2 min between the two procedures (see above), and our recorded operating time averaged 15 min from inflation of the tourniquet to completed application of the dressing. However, the time taken to infiltrate the local anaesthetic and transfer the patients meant that each procedure took over 30 min. In practical terms, we were not able to perform more than eight of either procedure per day-case list, so the time difference noted between the two techniques did not made an impact on our practice. Therefore, for our analysis we have assumed the time taken for both procedures was similar. According to the Confederation of British Industry (2002), it costs d67 (h101) per day for an employee to be off work (Counting the Costs, 2002). Therefore, as a result of returning to work 8 days sooner after ECTR than OCTR there would be a cost saving to industry of d536 (h805) per patient. This is offset by the incremental cost increase of ECTR of d98 (h147), resulting in a net
saving of d438 (h658) per patient for those in employment. We also assessed the impact of the surgery from the social perspective. Although patients found the first post-operative week in particular difficult in terms of pain and performing ADLs, in practice they were able to manage with no extra help from any community service or family or friends staying home from work. This was true even for the elderly living alone. As no input was required, no costing of any services was performed. Therefore, for those not in employment, there was nothing to offset the incremental cost of d98 (h147) of ECTR.
DISCUSSION ECTR is established as a method of treating patients with carpal tunnel syndrome (Chow, 1993; Erhard et al., 1999), but many feel that the risk of complications outweigh its benefits (Evans, 1994). The advantages of ECTR are reported as less immediate post-operative pain and a quicker return to work (Brown et al., 1993)
448
although some authors (Jacobsen and Rahme, 1996) have found little difference between the two methods of carpal tunnel release. Our study has found no evidence of one technique being superior to the other for provision of symptomatic relief and they appear to have similar post-operative courses. However, ECTR does lead to a quicker return to work, with an average reduction in sick leave of 8 days. The similarity between the two techniques may reflect the fact that recovery from either form of CTR is relatively quick, and though uncomfortable, the impact on the practical aspects of daily living is small. There was quite a high conversion rate in our ECTR group (12%) which is higher than in other reported series (Agee et al., 1995; Bande et al., 1994). This can be attributed to two main factors. Firstly, four of nine patients were converted to OCTR because of fogging of the camera which occurred on two consecutive theatre lists and was due to moisture affecting the coupling between the handpiece and the camera lead. Once this was remedied, there were no further problems. Secondly, in five of nine cases, the flexor retinaculum was not visualized completely so they were converted to open. It appears to be generally perceived that ECTR has a higher complication rate, especially for major neurovascular injury, and though ECTR is acknowledged to have a steeper learning curve than open techniques, a metaanalysis of published data showed irreversible nerve damage rates of 0.3% and 0.2% for ECTR and OCTR, respectively (Boeckstyns and Sorensen, 1999). The surgeons performing ECTR in our study had very strict criteria about whether to proceed with ECTR: unless the surgeon was certain he could see the whole of the flexor retinaculum with no intervening soft tissue the procedure was converted to an open one. This may have resulted in a higher conversion rate but almost certainly ensured that there were no lacerations of neurovascular structures. The economic evaluation of any form of intervention is increasingly important (Maniadakis and Gray, 2000). Attempts at calculating the cost-effectiveness of ECTR have been made in the United States (Chung et al., 1998; Vasen et al., 1999) and have shown ECTR to be a costeffective procedure, provided the complication rates of ECTR and OCTR are the same. We have performed a cost-effectiveness analysis predicated on what we feel are reasonable assumptions of practice in the UK but do not consider the following four factors. First, we did not perform bilateral CTR. If we had, then the findings in terms of social input required and time off work would probably have been different, and so the cost-effectiveness of the procedures would have altered. Secondly, the impact of any major nerve injury that resulted in permanent disability was not considered. If ECTR had a higher rate of nerve injury then this would have a huge impact on cost-effectiveness as the extra expense of attempts at nerve repair and the cost of
THE JOURNAL OF HAND SURGERY VOL. 28B No. 5 OCTOBER 2003
compensation and lost income far outweigh the saving from ECTR. However, we and others (Boeckstyns and Sorensen, 1999) have found ECTR a safe procedure. Thirdly, we did not consider whether ECTR results in a higher recurrence rate of carpal tunnel syndrome. The literature suggests recurrence rates after ECTR of up to 7% (Concannon et al., 2000) but there is also a recognized recurrence rate after OCTR of up to 12% (Botte et al., 1996). Our follow-up period was relatively short and it is possible that over time we will find a higher recurrence rate in those patients treated with ECTR. If this is the case, we will have to amend our calculations of cost-effectiveness. Fourthly, the small, but significant, time difference for the two techniques was not considered as administrative constraints prevent us from performing more than eight cases per day-case list. Without this limitation, use of OCTR would provide an extra 19 min at the end of one session compared to ECTR. This would not give us sufficient time to perform another OCTR procedure but could allow another small procedure to be performed. This time is therefore not utilized efficiently and may result in a loss to the overall productivity of the theatre suite. Should changes in the organizations of our day-case theatre practice occur, then the cost of this loss or productivity will have to be factored into our calculations. We conclude that ECTR has the advantage of a quicker return to work for those in employment and provides an economic benefit of ECTR of d438 (h658) per employed patient over OCTR.
References Agee LM, McCarroll HR, Tortosa RD et al. (1992). Endoscopic release of the carpal tunnel: a randomized prospective multicenter study. Journal of Hand Surgery, 17A: 987–995. Agee JM, Peimer CA, Pyrek JD et al. (1995). Endoscopic carpal tunnel release: a prospective study of complications and surgical experience. Journal of Hand Surgery, 20A: 165–171. Bande S, De Smet L, Fabry G (1994). The result of carpal tunnel release: open versus endoscopic technique. Journal of Hand Surgery, 19B: 14–17. Boeckstyns MEH, Sorensen AI (1999). Does endoscopic carpal tunnel release have a higher rate of complications than open carpal tunnel release? Journal of Hand Surgery, 24B: 9–15. Botte MJ, von Schroeder HP, Abrams RA et al. (1996). Recurrent carpal tunnel syndrome. Hand Clinics, 12: 731–743. Brown RA, Gelbermann RH, Seiler JC et al. (1993). Carpal tunnel release: a prospective randomized assessment of open and endoscopic methods. Journal of Bone and Joint Surgery, 75A: 1265–1275. Burke FD (2000). Carpal tunnel syndrome: reconciling ‘‘demand management’’ with clinical need. Journal of Hand Surgery, 25B: 121–127. Chow JCY (1993). The Chow technique of endoscopic release of the carpal ligament for carpal tunnel syndrome: four years of clinical results. Arthroscopy, 9: 301–314. Chung KC, Walters MR, Greenfield MLVH et al. (1998). Endoscopic versus open carpal tunnel release: a cost-effectiveness analysis. Plastic and Reconstructive Surgery, 102: 1089–1099. Concannon MJ, Brownfield ML, Puckett CL (2000). The incidence of recurrence after endoscopic carpal tunnel release. Plastic and Reconstructive Surgery, 105: 1662–1665. Confederation of British Industry (2002). Counting the costs: 2002 absence and labour turnover survey. CBI Publications, London. Erdmann M (1994). Endoscopic carpal tunnel decompression. Journal of Hand Surgery, 19B: 5–13.
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE Erhard L, Ozalp T, Citron N et al. (1999). Carpal tunnel release by the Agee endoscopic technique. Journal of Hand Surgery, 24B: 583–585. Evans D (1994). Endoscopic carpal tunnel release – the hand doctors dilemma. Journal of Hand Surgery, 19(B): 3–4. Ferdinand RD, MacLean JGB (2002). Endoscopic versus open carpal tunnel release in bilateral carpal tunnel syndrome. Journal of Bone and Joint Surgery, 84B: 375–379. Jacobsen MB, Rahme H (1996). A prospective, randomized study with an independent observer comparing open carpal tunnel release with endoscopic carpal tunnel release. Journal of Hand Surgery, 21B: 202–204. Levine DW, Simmons BP, Koris MJ et al. (1993). A self-administered questionnaire for the assessment of severity of symptoms and functional status in carpal tunnel syndrome. Journal of Bone and Joint Surgery, 75A: 1585–1592. Maniadakis N, Gray A (2000). Health economics and orthopaedics. Journal of Bone and Joint Surgery, 82B: 2–8.
449 Vasen AP, Kuntz KM, Simmons BP et al. (1999). Open versus endoscopic carpal tunnel release: a decision analysis. Journal of Hand Surgery, 24A: 1109– 1117.
Received: 9 September 2002 Accepted after revision: 3 March 2003 Mr. N.L.B. Saw, 54 Department of Orthopaedic Surgery, Norfolk & Norwich University Hospital, Colney Lane, Norwich NR4 7UY, UK. Tel.: +44-07747 777 082; E-mail: n [email protected].
r 2003 The British Society for Surgery of the Hand. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0266-7681(03)00097-4 available online at http://www.sciencedirect.com
Proponents of endoscopic carpal tunnel release have been advocating the technique for more than 10 years but there is still debate about its efficacy, safety and cost-effectiveness. We have performed a randomized, prospective, blind trial to compare early outcome after single portal endoscopic or open carpal tunnel surgery and to assess the cost-effectiveness of the procedures. There were no significant differences in symptom and functional activity scores, grip strength or anterior carpal pain in the first 3 months. For those in employment, we found a statistically significant difference between the two treatment groups with the endoscopic group returning to work, on average, 8 (95% CI, 2–13 days) days sooner than the open group. This translates into a cost saving to industry. There were no major neurovascular complications in either group. On the basis of these findings, we recommend that endoscopic carpal tunnel release should be considered in the employed as a cost-effective procedure, but perhaps not in the general population as a whole. Journal of Hand Surgery (British and European Volume, 2003) 28B: 5: 444–449 Keywords: endoscopic carpal tunnel release; cost-effectiveness
Between 1998 and 2001, 162 patients presenting to the Norfolk and Norwich or West Norwich Hospitals with a diagnosis of carpal tunnel syndrome were invited to take part in the trial. Of these, 123 patients gave informed consent and were enrolled into the study. The diagnosis of carpal tunnel syndrome was made on clinical grounds. Nerve conduction tests were performed only when there was clinical doubt. Exclusion criteria included those with previous wrist fractures and surgery, patients with rheumatoid arthritis and other inflammatory conditions, a suspected mass lesion or any patients requiring additional procedures such as biopsy and those less than 16 years of age. The patients were randomized using block randomization and sealed envelopes to either the endoscopic (ECTR) or open (OCTR) groups. All procedures were performed under local anaesthetic and tourniquet control by Consultant Hand Surgeons who are enthusiasts of their technique. OCTR was performed with a standard open approach with a 2 cm palmar incision (Fig 1) by one surgeon and ECTR by two surgeons of similar experience according to the extrabursal technique described by Agee (Agee et al., 1992) using the MicroAire CTRS single-portal system (MicroAire, USA). Post-operatively, dressings were applied and the patients advised to use the hand as comfort allowed. The patients were assessed immediately pre-operatively and at 1, 3, 6 and 12 weeks post-operatively. At each assessment, Levine et al.’s. (1993) Symptom Severity and Functional Status Scales questionnaires were completed to assess pre-operative severity, the subjective response to surgery and its impact on
INTRODUCTION Carpal tunnel syndrome is one of the most common conditions to affect the hand, with an incidence of carpal tunnel decompression of between 43 and 74 per 100,000 per year in the United Kingdom (Burke, 2000). Techniques of endoscopic carpal tunnel release (ECTR) have been in use for more than 10 years (Agee et al., 1992; Chow, 1993) and may result in a shorter postoperative recovery period and an earlier return to work. However, there are concerns that ECTR carries a higher risk of nerve injury than open carpal tunnel release (OCTR) and some doubt remains as to whether the benefits outweigh the risks (Evans, 1994). It is also important to consider the cost-effectiveness of any intervention (Burke, 2000). Although there are numerous articles published about ECTR, a search through the Cochrane Controlled Trials Register showed only two studies out of 20 listed that were performed in the United Kingdom; one (Erdmann, 1994) used a twoportal technique and the other (Ferdinand and MacLean, 2002) used a single-portal technique. No attempt has been made to establish the cost-effectiveness of ECTR in the UK. We have therefore compared the single-portal and open techniques and determined the cost-effectiveness of the procedures.
PATIENTS AND METHODS The Norwich Local Research and Ethics Committee granted ethical approval for this trial. 444
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE
445
Cost-effectiveness analysis The costing of each technique was predicated on the assumption that resource utilization in terms of surgeon, theatre staff, anaesthetic and theatre equipment was the same for both procedures except for the cost of the endoscopic equipment. The most expensive item was the disposable blade at d82 (h123). Provided the time taken to perform the procedures was similar, a costminimization analysis could be done. The main economic benefit of ECTR is a shorter time to return to work, and this was our primary efficacy endpoint. Activities of daily living (ADL) are always restricted in the post-operative period and the Levine functional questionnaire assessed this subjectively. We also used specific questions to detect increased dependence on the family, friends and community health services during the recovery from carpal tunnel release (CTR). Statistical methods
Fig 1 Skin markings for incision used for OCTR.
activities of daily living. A visual analogue scoring system was used to assess subjective anterior carpal tenderness in the heel of the hand. Further questions enquired whether any input in terms of domestic support and social services were required in the recovery period. Grip strength was measured using a Jamar Hand Dynamometer (Clifton, NJ, USA). The assessor was blinded to the procedure by placing a stockinet over the wrist of the patient. Patients were advised to return to work when they themselves felt that the discomfort in their hand had improved sufficiently to allow safe practice at their workplace. Sick notes were obtained by the patients from their General Practitioners. The dates of surgery and return to work as reported by the patient were used to calculate the number of days off work. Patients with bilateral carpal tunnel syndrome had their releases sequentially with the more affected hand first. Once these patients felt that they could use the operated hand normally they underwent the second procedure. The shortest interval between the two procedures was 7 months.
The primary endpoint was days until return to work. An intention-to-treat (ITT) analysis was used. Patients were analysed as randomized to the study and not with respect to the surgical procedure actually used. With respect to measures repeated over time, a ‘‘last observation carried forward’’ strategy was used to impute missing values. An area-under-the-curve (AUC) analysis was performed for the repeated measures. Two-sample t-tests or Wilcoxon–Mann–Whitney tests were carried out to test for between group differences, depending on whether the data were parametric or not. A significance level of 5% was considered statistically significant. Ninety-five per cent confidence intervals were constructed for estimated mean or median differences between the two groups.
RESULTS Seventy-four patients were randomized to the endoscopic group and 76 to the open group. Of the former, 9 (12%) were converted to open. No individual randomized to the open group received endoscopic surgery. A comparison of characteristics is summarized in Table 1. The patients in the endoscopic group were, on average, about 5 years older than those in the open group and were more likely to have the left hand operated on. Apart from these differences, the two groups appeared similar pre-operatively. The efficacy analysis is presented in Table 2. Fortythree patients (58%) in the endoscopic group and 42 (55%) in the open group were in employment at the time of the study. Analysis of the primary endpoint could only be carried out on these individuals, and this showed a statistically significant difference between the two groups (P ¼ 0:005; two-sample t-test). Those in the
446
THE JOURNAL OF HAND SURGERY VOL. 28B No. 5 OCTOBER 2003
endoscopic group returned to work, on average, 8 days (95% CI, 2–13 days) quicker than the open group. A general linear model was constructed to test for a between-group difference in the number of days taken to return to work whilst adjusting for age, side of hand operated (left or right) and occupation (manual and Table 1—Comparision of baseline characteristics Randomization group Endoscopic (n ¼ 74)
Open (n ¼ 76)
Age Mean (Std. dev.)
54 (15)
50 (15)
Sex Males
21 (28%)
19 (25%)
Hand side Left
30 (41%)
26 (34%)
Hand side Dominant
43 (58%)
48 (63%)
Employed
43 (58%)
42 (55%)
7 (16%)
6 (15%)
Self-employed Grip strength (kg) Mean (Std. dev.)
26 (12)
27 (11)
Levine functional score1 Median (S-IQR)
13.5 (4.0)
15.0 (4.8)
Levine symptom score2 Median (S-IQR)
19.5 (2.5)
20.0 (3.0)
1
non-manual), but these adjustments made no difference to our findings. The operation time for the endoscopic group was, on average, 2 min (95% CI, 1–4 min) longer. Data imputation in the remaining endpoints was required for four patients in the endoscopic group (2.7%) and two patients in the open group (1.3%). None of these endpoints were statistically significantly different between groups or near statistical significance. A plot of these secondary endpoints over time is provided in Fig 2. Complications There were no cases of permanent neurovascular injury in either group. One patient in the ECTR sustained transient numbness affecting the index finger that settled after 3 weeks. One patient in the OCTR group developed hyperaesthesia over the scar area that had improved by 3 months. There was one superficial wound infection in each group requiring antibiotics and one superficial haematoma in the OCTR group following a fall in the first week which resolved with no intervention. There was one case of incomplete release of the flexor retinaculum in the ECTR group that required open release at 3 months because there was no improvement in his symptoms. There was also one patient in the OCTR group with persistence of symptoms after surgery. Cost-effectiveness analysis
Levine functional score: range 8 (no impact on function) to 40 (unable to perform activities). 2 Levine symptom score: range 11 (asymptomatic) to 55 (severe symptoms).
Based on our assumptions as described earlier, the increased cost of ECTR was due to the initial capital expenditure for the basic ECTR equipment (d6482) and
Table 2—ITT efficacy analysis Randomization group Endoscopic (n ¼ 74)
Open (n ¼ 76)
Estimated difference
95% CI1
Significance2
Primary outcome Days off work Mean (Std. dev.).
18 (11)
26 (14)
8
13, 2
P=0.005
Secondary outcomes Operation time Mean (Std. dev.).
15.8 (3.5)
13.4 (3.4)
1.3, 3.5
Po0.001
22 (7)
24 (6)
2
4, 1
P=0.176
AUC analysis Anterior carpal tenderness3 Mean (Std. dev.). Grip strength (kg) Mean (Std. dev.). Levine functional score Median (S-IQR) (Range 88–440) Levine symptom score Median (S-IQR) (Range 121–605) 1
253 (121)
231 (102)
22
14, 60
P=0.213
109 (22)
108 (24)
1
10, 10
P=0.984
120 (21)
119 (19)
2
9, 13
P=0.701
95% confidence interval either for the difference in means or median difference as appropriate. Resulting from a two-sample t-test or the Wilcoxon–Mann–Whitney test as appropriate. 3 Range 11 (painless) to 55 (severe pain). 2
2.4
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE
447
Fig 2 (a) Grip strength (kg) by time; (b) Levine functional score by time; (c) Levine symptom scores by time; (d) anterior carpal tenderness by time.
to the cost of the non-reusable blade assembly (d82, h123). With the servicing costs of the equipment included, we estimated the incremental cost per case of ECTR due to extra equipment at d98 (h147). We found a small, but statistically significant, average difference of 2 min between the two procedures (see above), and our recorded operating time averaged 15 min from inflation of the tourniquet to completed application of the dressing. However, the time taken to infiltrate the local anaesthetic and transfer the patients meant that each procedure took over 30 min. In practical terms, we were not able to perform more than eight of either procedure per day-case list, so the time difference noted between the two techniques did not made an impact on our practice. Therefore, for our analysis we have assumed the time taken for both procedures was similar. According to the Confederation of British Industry (2002), it costs d67 (h101) per day for an employee to be off work (Counting the Costs, 2002). Therefore, as a result of returning to work 8 days sooner after ECTR than OCTR there would be a cost saving to industry of d536 (h805) per patient. This is offset by the incremental cost increase of ECTR of d98 (h147), resulting in a net
saving of d438 (h658) per patient for those in employment. We also assessed the impact of the surgery from the social perspective. Although patients found the first post-operative week in particular difficult in terms of pain and performing ADLs, in practice they were able to manage with no extra help from any community service or family or friends staying home from work. This was true even for the elderly living alone. As no input was required, no costing of any services was performed. Therefore, for those not in employment, there was nothing to offset the incremental cost of d98 (h147) of ECTR.
DISCUSSION ECTR is established as a method of treating patients with carpal tunnel syndrome (Chow, 1993; Erhard et al., 1999), but many feel that the risk of complications outweigh its benefits (Evans, 1994). The advantages of ECTR are reported as less immediate post-operative pain and a quicker return to work (Brown et al., 1993)
448
although some authors (Jacobsen and Rahme, 1996) have found little difference between the two methods of carpal tunnel release. Our study has found no evidence of one technique being superior to the other for provision of symptomatic relief and they appear to have similar post-operative courses. However, ECTR does lead to a quicker return to work, with an average reduction in sick leave of 8 days. The similarity between the two techniques may reflect the fact that recovery from either form of CTR is relatively quick, and though uncomfortable, the impact on the practical aspects of daily living is small. There was quite a high conversion rate in our ECTR group (12%) which is higher than in other reported series (Agee et al., 1995; Bande et al., 1994). This can be attributed to two main factors. Firstly, four of nine patients were converted to OCTR because of fogging of the camera which occurred on two consecutive theatre lists and was due to moisture affecting the coupling between the handpiece and the camera lead. Once this was remedied, there were no further problems. Secondly, in five of nine cases, the flexor retinaculum was not visualized completely so they were converted to open. It appears to be generally perceived that ECTR has a higher complication rate, especially for major neurovascular injury, and though ECTR is acknowledged to have a steeper learning curve than open techniques, a metaanalysis of published data showed irreversible nerve damage rates of 0.3% and 0.2% for ECTR and OCTR, respectively (Boeckstyns and Sorensen, 1999). The surgeons performing ECTR in our study had very strict criteria about whether to proceed with ECTR: unless the surgeon was certain he could see the whole of the flexor retinaculum with no intervening soft tissue the procedure was converted to an open one. This may have resulted in a higher conversion rate but almost certainly ensured that there were no lacerations of neurovascular structures. The economic evaluation of any form of intervention is increasingly important (Maniadakis and Gray, 2000). Attempts at calculating the cost-effectiveness of ECTR have been made in the United States (Chung et al., 1998; Vasen et al., 1999) and have shown ECTR to be a costeffective procedure, provided the complication rates of ECTR and OCTR are the same. We have performed a cost-effectiveness analysis predicated on what we feel are reasonable assumptions of practice in the UK but do not consider the following four factors. First, we did not perform bilateral CTR. If we had, then the findings in terms of social input required and time off work would probably have been different, and so the cost-effectiveness of the procedures would have altered. Secondly, the impact of any major nerve injury that resulted in permanent disability was not considered. If ECTR had a higher rate of nerve injury then this would have a huge impact on cost-effectiveness as the extra expense of attempts at nerve repair and the cost of
THE JOURNAL OF HAND SURGERY VOL. 28B No. 5 OCTOBER 2003
compensation and lost income far outweigh the saving from ECTR. However, we and others (Boeckstyns and Sorensen, 1999) have found ECTR a safe procedure. Thirdly, we did not consider whether ECTR results in a higher recurrence rate of carpal tunnel syndrome. The literature suggests recurrence rates after ECTR of up to 7% (Concannon et al., 2000) but there is also a recognized recurrence rate after OCTR of up to 12% (Botte et al., 1996). Our follow-up period was relatively short and it is possible that over time we will find a higher recurrence rate in those patients treated with ECTR. If this is the case, we will have to amend our calculations of cost-effectiveness. Fourthly, the small, but significant, time difference for the two techniques was not considered as administrative constraints prevent us from performing more than eight cases per day-case list. Without this limitation, use of OCTR would provide an extra 19 min at the end of one session compared to ECTR. This would not give us sufficient time to perform another OCTR procedure but could allow another small procedure to be performed. This time is therefore not utilized efficiently and may result in a loss to the overall productivity of the theatre suite. Should changes in the organizations of our day-case theatre practice occur, then the cost of this loss or productivity will have to be factored into our calculations. We conclude that ECTR has the advantage of a quicker return to work for those in employment and provides an economic benefit of ECTR of d438 (h658) per employed patient over OCTR.
References Agee LM, McCarroll HR, Tortosa RD et al. (1992). Endoscopic release of the carpal tunnel: a randomized prospective multicenter study. Journal of Hand Surgery, 17A: 987–995. Agee JM, Peimer CA, Pyrek JD et al. (1995). Endoscopic carpal tunnel release: a prospective study of complications and surgical experience. Journal of Hand Surgery, 20A: 165–171. Bande S, De Smet L, Fabry G (1994). The result of carpal tunnel release: open versus endoscopic technique. Journal of Hand Surgery, 19B: 14–17. Boeckstyns MEH, Sorensen AI (1999). Does endoscopic carpal tunnel release have a higher rate of complications than open carpal tunnel release? Journal of Hand Surgery, 24B: 9–15. Botte MJ, von Schroeder HP, Abrams RA et al. (1996). Recurrent carpal tunnel syndrome. Hand Clinics, 12: 731–743. Brown RA, Gelbermann RH, Seiler JC et al. (1993). Carpal tunnel release: a prospective randomized assessment of open and endoscopic methods. Journal of Bone and Joint Surgery, 75A: 1265–1275. Burke FD (2000). Carpal tunnel syndrome: reconciling ‘‘demand management’’ with clinical need. Journal of Hand Surgery, 25B: 121–127. Chow JCY (1993). The Chow technique of endoscopic release of the carpal ligament for carpal tunnel syndrome: four years of clinical results. Arthroscopy, 9: 301–314. Chung KC, Walters MR, Greenfield MLVH et al. (1998). Endoscopic versus open carpal tunnel release: a cost-effectiveness analysis. Plastic and Reconstructive Surgery, 102: 1089–1099. Concannon MJ, Brownfield ML, Puckett CL (2000). The incidence of recurrence after endoscopic carpal tunnel release. Plastic and Reconstructive Surgery, 105: 1662–1665. Confederation of British Industry (2002). Counting the costs: 2002 absence and labour turnover survey. CBI Publications, London. Erdmann M (1994). Endoscopic carpal tunnel decompression. Journal of Hand Surgery, 19B: 5–13.
ENDOSCOPIC VERSUS OPEN CARPAL TUNNEL RELEASE Erhard L, Ozalp T, Citron N et al. (1999). Carpal tunnel release by the Agee endoscopic technique. Journal of Hand Surgery, 24B: 583–585. Evans D (1994). Endoscopic carpal tunnel release – the hand doctors dilemma. Journal of Hand Surgery, 19(B): 3–4. Ferdinand RD, MacLean JGB (2002). Endoscopic versus open carpal tunnel release in bilateral carpal tunnel syndrome. Journal of Bone and Joint Surgery, 84B: 375–379. Jacobsen MB, Rahme H (1996). A prospective, randomized study with an independent observer comparing open carpal tunnel release with endoscopic carpal tunnel release. Journal of Hand Surgery, 21B: 202–204. Levine DW, Simmons BP, Koris MJ et al. (1993). A self-administered questionnaire for the assessment of severity of symptoms and functional status in carpal tunnel syndrome. Journal of Bone and Joint Surgery, 75A: 1585–1592. Maniadakis N, Gray A (2000). Health economics and orthopaedics. Journal of Bone and Joint Surgery, 82B: 2–8.
449 Vasen AP, Kuntz KM, Simmons BP et al. (1999). Open versus endoscopic carpal tunnel release: a decision analysis. Journal of Hand Surgery, 24A: 1109– 1117.
Received: 9 September 2002 Accepted after revision: 3 March 2003 Mr. N.L.B. Saw, 54 Department of Orthopaedic Surgery, Norfolk & Norwich University Hospital, Colney Lane, Norwich NR4 7UY, UK. Tel.: +44-07747 777 082; E-mail: n [email protected].
r 2003 The British Society for Surgery of the Hand. Published by Elsevier Ltd. All rights reserved. doi:10.1016/S0266-7681(03)00097-4 available online at http://www.sciencedirect.com