Item ID

Description

Implementation

Patient selection

NOTE: All criteria were scored yes (+), no (‐) or don't know (?).

b1

Was an adequate method of randomisation applied?

A random (unpredictable) allocation sequence must have been applied. Methods of allocation using date of birth, date of admission, hospital numbers, or alternation are not considered to be appropriate.

b2

Was the treatment allocation concealed?

Allocation should have been performed by an independent person who is not responsible for determining eligibility for inclusion. This person has no information about the patients included in the trial and has no influence on the allocation sequence or the decision about eligibility for inclusion.

c

Were the groups similar at baseline with regard to the most important prognostic indicators?

Groups must be similar at baseline with regard to at least three of the four prognostic indicators of age, sex, duration of symptoms and value of main outcome measure(s).

Interventions

e

Were co‐interventions avoided or similar for all groups?

Co‐interventions should either have been avoided in the trial design or be similar in the groups.

f

Were the patients blinded for the intervention?

Adequate information about blinding must have been provided.

Outcome measurement

g

Was the outcome assessor blinded to the intervention?

Adequate information about blinding must have been provided.

j

Was the drop‐out/loss to follow‐up rate described and acceptable?

Included patients who did not complete the follow‐up period or were not included in the analysis, must have been described. If the percentage of drop‐outs and loss to follow‐up is < 20% for short‐term follow‐up and < 30% for long‐term follow‐up, and loss to follow‐up does not lead to substantial bias, a '+' is scored. (N.B. these percentages are arbitrary and not supported by empirical evidence).

l

Was the timing of the outcome assessment similar for all groups?

Timing of outcome assessment should have been started from the moment of treatment allocation and be identical for all intervention groups and for all important outcome assessments.

Statistics

n

Did the analysis include an intention‐to‐treat analysis?

For all randomised patients, the most important moments of effect measurement should have been reported/analysed (minus missing values), irrespective of non‐compliance and co‐interventions.

Reference

b1

b2

c

e

f

g

j

l

n

Adequate randomisation

Adequate allocation concealment

Groups similar at baseline

Co‐interventions similar or avoided

Adequate patient blinding

Adequate assessor blinding

Dropout rate described and acceptable

Assessment timing similar for all groups

Intention to treat analysis included?

ECTR versus OCTR

Agee 1992

‐

?

?

?

n/a

?

?

+

?

Atroshi 2006

+

+

+

+

n/a

‐

+

+

+

Benedetti 1996 + Sennwald 1995

+

+

+

+

n/a

?

+

+

‐

Brown 1993

+

‐

+

+

n/a

+

+

+

+

Dumontier 1995

+

‐

?

+

n/a

?

‐

+

‐

Eichhorn 2003

?

?

?

?

n/a

?

?

?

‐

Erdmann 1994

+

?

+

?

n/a

?

?

+

+

Ferdinand 2002

+

?

?

+

n/a

+

+

+

?

Foucher 1993

?

?

?

?

n/a

?

?

+

?

Hoefnagels 1997

?

?

+

+

n/a

?

+

+

‐

Jacobsen 1996

?

?

+

+

n/a

‐

+

+

+

Macdermid 2003

?

?

+

?

n/a

+

?

+

?

Saw 2003

+

+

+

+

n/a

+

+

+

+

Stark 1996

?

?

?

+

n/a

?

+

+

+

Trumble 2002

+

?

+

+

n/a

+

+

+

?

Westphal 2000

?

?

+

?

n/a

?

?

+

?

ECTR versus OCTR with modified incision

Eichhorn 2003

?

?

?

?

n/a

?

?

?

‐

Mackenzie 2000

?

?

?

+

n/a

?

?

+

‐

Rab 2006

+

?

+

+

n/a

?

+

+

+

Wong 2003

+

?

+

+

n/a

?

+

+

?

OCTR: modified versus standard incision

Brüser 1999 + Richter 1996

+

?

+

‐

n/a

?

+

+

+

Citron 1997

?

?

+

?

n/a

?

+

+

+

Eichhorn 2003

?

?

?

?

n/a

?

?

?

‐

Jugovac 2002

+

+

?

?

n/a

+

+

+

+

Nakamichi 1997

+

?

?

+

n/a

+

+

+

+

Nitzsche 1999

?

?

?

?

n/a

?

+

?

?

OCTR versus OCTR with lengthening of the flexor retinaculum

Dias 2004

+

+

+

+

+

+

+

+

+

OCTR with Knifelight versus OCTR alone

Bhattacharya 2004

+

+

+

+

n/a

‐

?

+

‐

Helm 2003

+

?

+

?

n/a

‐

+

+

‐

OCTR with Knifelight versus OCTR with limited incision

Cellocco 2005

????

?

+

+

n/a

?

?

‐

?

OCTR with internal neurolysis versus OCTR alone

Holmgren 1985 + Holmgren 1987

?

?

?

+

?

?

+

+

+

Lowry 1988

+

?

?

+

+

+

+

+

‐

Mackinnon 1991

+

+

+

+

+

+

+

+

?

OCTR with epineurotomy versus OCTR alone

Leinberry 1997

+

?

+

+

?

+

+

+

+

OCTR with tenosynovectomy versus OCTR alone

Shum 2002

+

?

?

?

?

?

?

?

?

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work/ADL

Complications

Agee 1992

Results of 97 adequately randomised patients with unilateral CTS not presented separately.

Results of 97 adequately randomised patients with unilateral CTS not presented separately.

Median 25 (ECTR) and 46.5 (OCTR) days (significant difference between the groups).

ECTR: re‐operation needed with OCTR in 2 of 82 patients; transient ulnar neurapraxia (2 patients). OCTR: injury to deep motor branch of ulnar nerve (1 patient); bowstringing of digital flexor tendons (1); wound dehiscence (2).

Atroshi 2006

Mean symptom severity score (Levine) after 3 months: ECTR 1.5; OCTR 1.5.
Mean functional status score (Levine) after 3 months: ECTR 1.3; OCTR 1.3.
Difference in mean pain scores (0‐100) after 3 months ‐13.3 (95%‐CI ‐21.3 to ‐5.3) in favour of ECTR.

Mean symptom severity score (Levine) after 12 months: ECTR 1.4; OCTR 1.4 (NS).
Mean functional status score (Levine) after 12 months: ECTR 1.3; OCTR 1.2 (NS).
Difference in mean pain scores (0‐100) after 12 months ‐5.8 (95%‐CI ‐13.3 to ‐1.7) in favour of ECTR.

Not on sick leave before surgery: mean difference ‐5 days (95%CI ‐11.5 to 1.5) in favour of ECTR.
On sick leave before surgery: mean difference 8 days (95%‐CI ‐62.5 to 78.5) in favour of OCTR.

Repeat surgery: ECTR 2/63 (3%); OCTR 1/65 (2%). No other complications.

Benedetti 1996 and Sennwald 1995

Not assessed.

Not assessed.

Mean 24 (ECTR) and 42 (OCTR) days (significant difference between the groups).

1 conversion to OCTR and 1 transient neurapraxia after ECTR. 1 painful hypertrophic scar and 1 reflex sympathetic dystrophy after OCTR.

Brown 1993

Improvement in symptoms (paraesthesiae, numbness) in 99% (ECTR) and 98% (OCTR) after 12 weeks (difference 1% [95%CI ‐3 to 5%] ).

Not assessed.

Median 14 (ECTR) and 28 (OCTR) days (significant difference between the groups).

Significantly more scar tenderness after OCTR versus ECTR after 12 weeks (no significant differences after 3 and 6 weeks). No significant differences between the groups in tenderness of the thenar eminance at 3, 6 and 12 weeks. 1 partial transection of the superficial palmar arch, 1 digital‐nerve contusion, 1 ulnar‐nerve neurapraxia and 1 wound hematoma after ECTR.

Dumontier 1995

Persisting paraesthesiae after 3 months: 7% (OCTR) versus 12% (ECTR). Persisting pain after 3 months: 43.3% (OCTR) versus 38.5% (ECTR).

Paraesthesiae completely disappeared in all patients after 6 months. Persisting pain after 6 months: 28% (OCTR) versus 25% (ECTR).

Percentage of patients returned to work (OCTR versus ECTR): 72% versus 45% after 1 month; 90% versus 72% after 3 months.

Transient reflex sympathetic dystrophy in 4 patients (2 in each group).

Eichhorn 2003

Overall severity score (scale 1‐6) after > 1 year: OCTR 2.2; ECTR 2.1.

Two postoperative infections after OCTR; none in the ECTR group.
Recurrences: ECTR 3/128 (2%), OCTR 4/60 (7%).
Need for repeated surgery: ECTR 2/128 (2%); OCTR 3/60 (5%).

Erdmann 1994

Significantly more improvement in carpal tunnel pain in favour of ECTR after 1, 2 and 4 weeks, but no significant difference between the groups after 3 months.

No significant difference in carpal tunnel pain between the groups after 6 and 12 months.

Mean 14 (ECTR) and 39 (OCTR) days (only patients not simultaneously operated on both hands) (significant difference between the groups).

1 ulnar nerve paraesthesiae and 1 incomplete release after ECTR. 1 wound infection, 1 scar tethering and 5 scar hypertrophy after OCTR.

Ferdinand 2002

After 12 weeks better endoscopic Jebson scores (75 vs 65).

After 12 months better endoscopic Jebson scores (59 vs 48).

Not applicable (all patients had bilateral CTS).

3 conversions to OCTR after ECTR. 1 persisting wound pain in each group. 1 persisting symptoms and 1 superficial nerve injury after OCTR.

Foucher 1993

No data presented.

No data presented.

No significant differences in time to return to work between the groups (all 17 days).

1 algodystrophy and 2 conversions to OCTR after ECTR.

Hoefnagels 1997

Mean symptom severity score after 3 months 1.6 ± 0.7 after ECTR; 1.5 ± 0.5 after OCTR (no significant difference).

Not assessed.

Longer than 4 weeks absence from work in 16% (ECTR); 13% (OCTR) (difference 3%, 95%CI ‐7 to 14).

Significantly less postoperative pain after ECTR versus OCTR after 1 week. 1 conversion to OCTR, 1 broken knife left in operation wound and 1 increased numbness in fingertips after ECTR.

Jacobsen 1996

Not assessed.

Not assessed.

Mean 17 (ECTR) and 19 (OCTR) days (no significant difference between the groups).

3 transient numbness on the radial side of the ring finger after ECTR. 1 prolonged wound secretion after OCTR.

Macdermid 2003

After 12 weeks no significant differences in pain (McGill) (8 vs 12), symptom severity score (Levine) (1.8 vs 2.0) and functional status (SF‐36) (47 vs 42).

After a mean of 3.2 years lower satisfaction scores after ECTR (85% vs 93%).

No significant differences (no quantitative data presented).

No complications reported. Within 4 year in 5% of the ECTR cases re‐operation needed.

Saw 2003

Area under the curve analysis of symptom severity score (Levine) after 3 months: ECTR 120 (IQR 21); OCTR 119 (IQR 19) (P = 0.70).
Area under the curve analysis of functional status score (Levine) after 3 months: ECTR 109 (IQR 22); OCTR 108 (IQR 24) (P = 0.98).

Mean (SD) days off work ECTR 18 (11); OCTR 26 (14). Mean difference ‐8 (95% CI ‐13 to ‐2).

ECTR: 1 transient numbness index finger, 1 superficial wound infection, 1 repeat surgery.
OCTR: 1 hyperaesthesia over scar area, 1 superficial wound infection, 1 superficial haematoma, 1 persistence of symptoms.

Stark 1996

(Matched pairs) Pain completely relieved in 20/20 (ECTR) vs. 15/20 hands (OCTR) after 3 months. Persisting paraesthesiae in 1/20 (ECTR) vs. 1/20 (OCTR) after 3 months.

(Matched pairs) Pain completely relieved in 20/20 (ECTR) vs 19/20 hands (OCTR) after 8 months. Persisting paraesthesiae in 1/20 (ECTR) vs 1/20 (OCTR) after 8 months.

Mean 20 (ECTR) versus 30 (OCTR) days (significant difference between the groups).

1 subcutaneous hematoma and 1 loss of strength and mobility in the wrist after ECTR. 2 loss of strength and 1 swollen/stiff fingers after OCTR.

Trumble 2002

After 12 weeks better scores for satisfaction (4.4 vs 4.0, non‐significant), symptom severity (Levine) (1.8 vs 2.5; significant) and functional status (Levine) (1.7 vs 2.4; significant.).

After 12 months no significant differences for satisfaction (4.6 vs 4.5), symptom severity score (Levine) (1.8 vs 1.8) and functional status score (Levine) (1.7 vs 1.7).

Median 18 (ECTR) and 38 (OCTR) days (significant difference between the groups).

After OCTR 2 reflex sympathetic dystrophy and 1 repeat procedure.

Westphal 2000

Symptom severity score (variant of Levine) after 3 months: ECTR 11.0 (3.7); OCTR 10.6 (2.6).
Mean functional status score (variant of Levine) after 3 months: ECTR 10.2 (4.5); OCTR 9.8 (4.4).

Mean 34.5 (ECTR) versus 36 (OCTR) days (no significant difference between the groups).

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Eichhorn 2003

Mean overall severity score (scale 1‐6) after > 1 year: ECTR 2.1; mini incision 2.2.

Not assessed.

None. Recurrences: ECTR 2%, mini‐incision 14%. Need for repeated surgery ECTR 2%, mini‐incision 9%.

Mackenzie 2000

No quantitative data presented.

Not assessed.

Not assessed.

1 pillar pain in each group.

Rab 2006

At 12 weeks: mean symptom severity score (Levine) ECTR 14.7; modified OCTR 16.8 (P = 0.27). Mean functional status score (Levine) ECTR 10.3; modified OCTR 12.3 (P = 0.16). Pain (VAS) ECTR 0.3; modified OCTR 1.7 (P = 0.10).

At 12 months: mean symptom severity score (Levine) ECTR 14.0; modified OCTR 12.8 (P = 0.49). Mean functional status score (Levine) ECTR 11.1; modified OCTR 9.9 (P = 0.39). Pain (VAS) ECTR 0.6; modified OCTR 0.2 (P = 0.43).

Not assessed.

None.

Wong 2003

Statistically significant difference in reduction of wound pain at 2 and 4 weeks in favour of ECTR, but not after 8 and 16 weeks.

At 12 months: complete relief or minimal symptoms ECTR 27/30 hands (90%); modified OCTR 27/29 hands (93%). Preference for ECTR 6, modified OCTR 13; no preference 10.

Not Assessed.

None.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Brüser 1999 and Richter 1996

No significant differences in night symptoms, tingling and numbness between the groups after 1, 2, 3, 6 weeks.

Not assessed.

Median (mean) 21 (17.6) days versus 18 (18.8) days (no significant difference between the groups).

No significant differences in scar tenderness and tenderness of the thenar and hypothenar eminances after 1, 2, 3, 6 weeks (differences ranging from 0.3 to 0.6 on 5 point scale).

Citron 1997

No suitable data presented.

96% (modified incision) and 89% (standard incision) completely free of symptoms after 12 months (difference 7%, 95%CI ‐9 to 22).

Not assessed.

No significant difference in pain of the thenar eminance between the groups. Significantly more recovery in scar sensitivity after modified incision compared with standard incision (differences ranging from 0.5 to 1.0 on a visual analogue scale). 1 delayed wound healing in both groups.

Eichhorn 2003

Mean overall severity scores (scale 1‐6) after > 1 year: OCTR 2.2; mini incision 2.2

Two postoperative infections after OCTR; none in the mini‐incision group.
Recurrences: OCTR 7% and mini‐incision 14%. Need for repeated surgery: OCTR 5% and mini‐incision 9%.

Jugovac 2002

No significant difference in relief (31/36 vs 31/36) between the groups after 3 months.

Not assessed.

Median 15 (5 to 45) days versus 30 (10 to 60) days (significant difference between the groups in favour of limited incision).

Scar tenderness: 3/36 (limited incision) versus 8/36 (standard incision).

Nakamichi 1997

No significant differences in symptoms (numbness and paraesthesiae) between the groups after 3, 6, 13 weeks.

No significant differences in symptoms (numbness and paraesthesiae) between the groups after 6, 12, 24 months. 86% (modified incision) and 89% (standard incision) completely free of symptoms after 24 months (difference ‐3%, 95% CI ‐16 to 10).

Not assessed.

Standard incision significantly more scar pain and tenderness compared with modified incision after 3, 6, 13 weeks (differences ranging from 0.5 to 0.9 on 4 point scale). No other complications.

Nitzsche 1999

No quantitative data reported.

No quantitative data reported.

No quantitative data reported.

No quantitative data reported.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Dias 2004

Mean symptom severity score (Levine) after 12 weeks: OCTR 1.2 (95% CI 1.1 ‐ 1.3); lengthening 1.2 (95% CI 1.1‐1.4). Mean functional status score (Levine) after 12 weeks: OCTR 1.2 (95% CI 1.1‐1.3); lengthening 1.2 (95% CI 1.0‐1.3).

Mean symptom severity score (Levine) after an average of 26 weeks: OCTR 1.3 (95% CI 1.2 ‐ 1.5); lengthening 1.3 (95% CI 1.1‐1.5). Mean functional status score (Levine) after an average of 26 weeks: OCTR 1.2 (95% CI 1.1‐1.3); lengthening 1.3 (95% CI 1.1‐1.5).

Not assessed.

None reported.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Knifelight versus OCTR

Bhattacharya 2004

Symptom relief after 6 weeks: Knifelight 17/26 (65%); OCTR 19/26 (73%).

Not assessed.

Median no. of weeks (range): Knifelight 2 (0‐6); OCTR 2 (0‐6) (P = 0.8).

Scar tenderness: Knifelight 8/26 (31%); OCTR 17/26 (65%). Postoperative numbness after Knifelight (1)

Helm 2003

After 6 weeks: symptoms cured 36/39 (Knifelight) vs 37/43 (standard); none or mild scar tenderness 35/39 vs 21/43.

Not assessed.

Significant difference in time to return to work (20 vs 28 days).

1 sympathetic reflex dystrophy, 2 conversions to OCTR (presumably analysed in the OCTR group), 1 superficial wound infection and 1 transient numbness of index finger after Knifelight. 1 partial wound dehiscence, 4 pillar pain, 1 unexplained thumb pain, 1 mild stiffness of fingers and 1 transient numbness of index finger after standard OCTR.

Knifelight versus OCTR (limited open)

Cellocco 2005

Not assessed.

Follow up 19 months (range 12 to 28 months): mean symptom severity score (Levine) 1.46 (Knifelight) versus 2.04 (OCTR limited open) (P < 0.001); mean functional status score (Levine) 2.02 (Knifelight) versus 2.53 (OCTR limited open) (P < 0.001). Follow up 30 months (range 24‐42 months): mean symptom severity score (Levine) 1.28 (Knifelight) versus 1.39 (OCTR limited open) (NS); mean functional status score (Levine) 1.87 (Knifelight) versus 1.73 (OCTR limited open) (NS).

Mean no. of days: Knifelight 16.6 (range 15 to 18); OCTR (limited open) 25.4 (range 23 to 29).

Recurrents disease Knifelight 5% vs OCTR (limited open) <1%.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Holmgren 1985 and Holmgren 1987

No significant differences in pain, paraesthesiae and hypoaesthesiae between the groups after 3‐4 weeks.

89% totally free of symptoms in both groups after 6 months and 86% (with neurolysis) versus 75% (without neurolysis) after 3 to 4 years (difference 11%, 95% CI ‐15 to 36 ).

Not assessed.

No complications ascribable to internal neurolysis.

Lowry 1988

Excellent or good clinical response (pain, sensory deficit, complications) according to neurologist in 67% (with neurolysis) and 65% (without neurolysis) after 3 months (difference 2%, 95% CI ‐26 to 29).

Not assessed.

Not assessed.

4 persistent incisional pain, 1 hand swelling and 1 adhesive capsulitis in the group with internal neurolysis. 3 persistent incisional pain and 1 causalgia in the group without internal neurolysis.

Mackinnon 1991

No data presented.

Improvement in symptoms in 81% (with neurolysis) and 88% (without neurolysis) after 12 months (difference ‐7%, 95% CI ‐25 to 11 ).

Not assessed.

No data presented.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work /ADL

Complications

Leinberry 1997

No data presented.

56% (with epineurotomy) and 60% (without epineurotomy) totally free of symptoms (pain, altered sensibility, paraesthesias, loss of manual dexterity) after 1 year (difference ‐4% 95% CI ‐31% to 23%).

Not assessed.

No data presented.

Reference

Symptoms <= 3 months

Symptoms > 3 months

Return to work / ADL

Complications

Shum 2002

Not assessed.

After at least 12 months of follow‐up no significant differences between the groups with respect to symptom severity score (Levine: 1.6 (0.68) vs 1.6 (0.70)) and functional status score (Levine: 1.7 (0.71) vs 1.6 (0.62)).

Not assessed.

No wound infections.