Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Infectie, secundaire dislocatie, patient related outcome measure (PROM)

Comments

Dynamic versus static fixation

Kortekangas, 2015

Type of study:

RCT

Setting:

Single centre

Country:

Finland

Source of funding:

The study was supported by Oulu University Central Hospital.

The funding source had no influence or involvement in the study.

Medical Physicist, PhD Jani Katisko has a consultant service

agreement with Medtronic Finland Oy. All other authors declare

that they have no financial or personal relationships that could

influence this study.

Inclusion criteria:

-skeletally mature patients (16 year of older with Lauge-Hansen pronation-external rotation (PER), AO/OTA Weber C type ankle fracture

-pre- or intra-operative evidence of syndesmotic disruption based on plain radiographs or on the manual external rotation test under fluoroscopy

Exclusion criteria:

-previous ankle fracture,

-concomitant tibia fracture

-diabetic or other neuropathy

-a delay from trauma to surgery of more than 7 days

-pathological fracture

-inadequate cooperation.

N total at baseline:

Intervention: 21

Control: 22

Important prognostic factors²:

For example

age ± SD:

I:46.0 (14.8)

C: 43.5 (15.7)

P=0.591

Sex:

I: 61,9% M

C: 63,6% M

Weber C type fracture:

I: 57%

C: 54,5%

Maisonneuve fracture

I: 42,9%

C: 40,9%

Comorbidities (arteriosclerosis obliterans, diabetes mellitus, alcoholism or mental illness)

I: 14,2%

C: 18,2%

p>0.9

Groups comparable at baseline?

Yes, baseline characteristics were comparable.

One TightRope device

One 3.5-mm cortical screw purchasing three cortices

Both groups: fixation following standard AO principles (Arbeitsgemeinschaft fur Osteosynthesefragen).

Both groups: similar post-operative treatment protocol:

Immobilization in a below-the-knee cast with the ankle joint at 90 degrees for 6 weeks with partial weight bearing. At 6 weeks, the cast was removed, the ankle was examined and a physiotherapist instructed the patient in rehabilitation exercises. No additional bracing was used and weight bearing was allowed as tolerated.

Only screw removal if local irritation occurred.

Length of follow-up:

>2 years, ranging from 27-47 months

Loss-to-follow-up:

Intervention:

None

Control:

N (%) 2 (9%)

Reasons: two patients declined to participate in follow-up visits

Incomplete outcome data:

Intervention:

None

Control:

N (%) 1 (4,5%)

Reasons: low quality of intraoperative CT

Radiological outcomes

Malreduction (>2mm side-to-side difference in the mean width of the syndesmosis:

Intraoperative CT:

I: 1/21 (5%)

C: 1/21 (5%)

p>0.9

after 2 years on CBCT

I: 1/21 (5%)

C: 3/19 (16%)

P=0.33

Osteoarthritis based on Morrey – Wiedeman (0 points: no OA; 3 points: total degeneration) after 2 years

Injured; uninjured ankle (n)

Grade O

I: 0; 7

C: 4; 11

Grade I:

I: 7; 10

C: 8; 7

Grade II:

I: 11; 3

C: 6; 1

Grade III

I: 3; 1

C: 1; 0

OA one or more grades serious in the injured than in the uninjured ankle:

I: 13/21

C: 11/19

p>0.9

Functional outcomes

Foot and Ankle Outcome Score FAOS after 2 years. Mean (SD)

Symptoms

I: 81 (18)

C: 78 (24) P=0.64

Pain

I: 92 (15)

C: 89 (15) P=0.59

Daily living

I: 93 (14)

C: 94 (12) P=0.84

Sports and free time

I: 85 (28)

C: 82 (27) p=0.75

Quality of life

I: 78 (27)

C: 78 (24) p=0.95

Range of motion degree injured ankle

I: 83 (10)

C: 82 (9) p=0.81

OMAS

1 year

I: 75; C: 80

2 years

I: 82; C: 84

VAS pain 1 year (scale 0-100):

I: 16; C: 13

2 years:

I: 12; C: 12

Complications

Complications:

I:

C: 3 patients had broken screws and 13 patients had loosened screws after 2 years

Second surgery needed:

I: 1 due to new accident and pilon fracture in the same ankle

C: 3 removed syndesmotic screws within 3 months due to local irritation; 1 infection (S. Auerus): reoperation

Author’s conclusion:

Syndesmotic screw and TightRope fixation had similar postoperative malreduction rates.

After at least 2 years of follow-up, analysis of bilateral CBCT data suggested that the rate of syndesmosis malreduction may have slightly increased when syndesmotic screw fixation was used, but reduction was well maintained after fixation with TightRope. Osteoarthritis and functional outcome revealed insignificant differences between the groups.

Laflamme, 2015

Type of study:

RCT

Setting:

Multi-centre

Country:

Canada

Source of funding:

Not reported

Inclusion criteria:

-age between 18-65 years

-recent ankle fracture impairing syndesmotic stability

-time from trauma to surgery less than 8 days

Exclusion criteria:

-polytrauma

-neurologic impairment

-obesity (BMI≥40)

-poor medical condition limiting rehabilitation or participation in the study’s tests.

N total at baseline:

Intervention: 34

Control:36

Important prognostic factors²:

Mean (SD)

Age:

I: 40.1 (14.8)

C: 39.3 (12.4)

P=0.41

Sex:

I: 73% M

C:72 % M

P=1.0

BMI:

I:28.96 (4.6)

C: 28.6 (4.3)

P=0.49

AO/OTA fracture classification n (%)

44-B2:

I: 3 (9%)

C: 6 (17%)

44-B3

I: 2 (6%)

C: 1 (3%)

44-C1:

I: 18 (53%)

C:18 (50%)

44-C3:

I: 11 (32%)

C: 11 (31%)

P=0.73

Groups comparable at baseline?

Yes, no differences between groups based on smoking, diabetes, fracture type, age, sex, bmi.

Intervention:

One TightRope implant

Control:

One 3.5 mm quadricortical screw

Both groups:

No routine screw removal; standardized rehabilitation process: posterior splint for 2 weeks, then replaced by a below-knee cast for 4 more weeks. No ankle brace nor weight bearing was allowed during this period. After 6 weeks, the cast was removed and progressive rehabilitation authorized without formal physiotherapy (progressive weightbearing without protection and restoration of ankle motion).

Length of follow-up:

3, 6, and 12 months

Loss-to-follow-up:

Intervention:

N (%) 1 (2.9%)

Reasons: not described. Lost to follow up immediately after surgery

Control:

N (%) 4 (11.1%)

Reasons: Reasons: not described. Lost to follow up immediately after surgery

Incomplete outcome data:

See loss to follow up.

Radiological outcomes after 12 months: mean (SD) in mm

loss of reduction

Medial clear space

I: 0.05 (0.32)

C: 0.41 (0.37) p=0.02

Lateral TF clear space

I: 0.32 (0.55)

C: 1.34 (1.36) p=0.0005

Medial clear space in mm, mean (SD)

I: 2.9 (0.5)

C: 3.1 (0.6) p=0.38

Lateral clear space in mm, mean (SD)

I: 3.9 (0.9)

C: 4.8 (1.5) p=0.008

Functional outcomes after 12 months Mean (SD)

OMAS

I: 93.3 (10.2)

C: 87.7 (12.2) P=0.046

AOFAS score

I: 93.1 (9.3)

C: 89.9 (12.7) P=0.255

Range of motion

Plantarflexion degrees

I: 33.6 (7.3)

C: 32.0 (9.8) p=0.45

Dorsiflexion degrees

I: 13.6 (4.5)

C: 14.8 (7.6) p=0.43

Pain (VAS, scale 0-10)): mean

I: 0.6 (range 0-3)

C: 1.1 (range 0-5) p=0.12

Complications and reoperations

I: 2 patients required removal (after 14 weeks)

C: 1 patient required early revision; 11 patients required screw removal for discomfort after 6.3 months

P=0.0006

Author’s conclusion:

The dynamic fixation of acute ankle syndesmosis rupture gives better clinical and radiographic outcomes than the static fixation with one 3.5mm quadricortical screw.

Comments:

Surgery was performed by general orthopaedic surgeons

Seyhan, 2015

Type of study:

Case-control

Setting:

Retrospective file study

Country:

Turkey

Source of funding:

‘’ there is no grants and/or financial support for the work’’

Inclusion criteria:

-patients who had undergone surgical treatment between January 2007-June 2011

-syndesmotic fracture

Exclusion criteria:

- >60 years

-diabetes mellitus

-open fracture

-poly trauma

-insufficient follow-up

N total at baseline:

Intervention:15

Control:17

Important prognostic factors²:

Age, mean:

I:33.2

C:32

Sex:

I: 86.7% M

C: 82.4% M

Weber B:

I: 3 patients (20%)

C: 4 patients (23.5%)

Weber C:

I: 12 patients (80%)

C: 13 patients (76.5%)

Groups comparable at baseline?

Based on the available data, the groups seem comparable. No information available about co-morbidities or smoking.

Intervention:

Single level elastic fixation TightRope

Control:

Quadricortical 4.5mm screw

At the end of the 3^rd month, the screws were removed and complete weight-bearing was permitted.

Both groups:

Plaster splints, mobilization with a walker or double crutches without weight-bearing on the operated side. After 2 weeks, the splint was removed and pressure-socks were applied with the foot kept in the neutral position. After the 6^th week, partial weight bearing was initiated on the operated site suing double crutches.

Length of follow-up:

1, 2, 3, 6 and 12 months

Loss-to-follow-up:

None, one patient with insufficient follow-up was excluded based on the exclusion criteria

Incomplete outcome data:

Not applicable

Complications

I: 6 removals (one after 8 months, the other after 12 months, the other 4 ‘at a later stage’)

C: removal was part of protocol.

Functional outcomes after 12 months

AOFAS, mean (SD)

I: 93.35 (6.93)

C: 93.73 (7.38) p=0.889

AOFAS subscore for pain, mean (SD)

I: 38.00 (4.14)

C: 38.23 (3.93) p=0.87

Range of motion, mean difference between injured versus uninjured foot (SD)

Dorsiflexion

I: 1.4 (1.14)

C: 3.65 (1.54) p=0.001

Plantarflexion:

I: 2.67 (1.49)

C: 8.41 (2.63) p=0.001

Comments:

Surgeon was performed by one surgeon

Thornes, 2005

Type of study:

Retrospective cohort

Setting:

Single centre

Country:

Ireland

Source of funding:

‘one of the authors may receive funding from a patent/licensing agreement arising from this technique’.

Inclusion criteria:

-patients with Weber C type ankle fractures who required syndesmosis fixation

Exclusion criteria:

-multiple injuries

-patients pursuing compensation litigation arising from the injury (worker compensation) were excluded because of known adverse effect on assessing outcome.

N total at baseline:

Intervention: 16

Control: 16

Important prognostic factors²:

Average age (range):

I:32 (19-74)

C:31 (17-51)

Sex:

I: 81 % M

C: 75 % M

Groups comparable at baseline?

Based on the available data, the groups seem comparable.

Intervention (prospective data collection)

Polyester suture button (brand: Ethibond Excel; Ethicon, Johnson & Johnson)

Immobilization in a below-knee cast for 6 weeks post-operatively.

Early weight bearing was allowed (after 2 weeks)

Control (retrospective data collection)

One four-cortical syndesmotic screw. Immobilization in a below-knee cast for 6 weeks post-operatively.

Length of follow-up:

3 and 12 months

Loss-to-follow-up:

No loss to follow up

Incomplete outcome data:

Not applicable

Complications

Need for second operation to remove implant:

I: 0 patients

C: 12 patients, p=0.001

Functional outcomes after 12 months:

AOFAS score: mean

I: 93

C: 83, p=0.04

Standard deviation was presented in the figure as error bar, these were not described separately as numbers

Return to work

I: 2.8 months

C: 4.6 months, p=0.02

Quality of Life

Patient satisfaction (type of questionnaire not mentioned)

I: 8 patients excellent rating

7 patients good rating

1 patient fair rating

C: 5 patients excellent rating

7 patients good rating

2 patients fair rating

3 patients poor rating

No p-value reported

Comments:

One of the authors might have received funding due to a patent/licensing agreement arising from the technique.

Different post-operative treatment protocols were followed: mean time non-weight bearing:

I: 4.1 weeks

C: 6.3 weeks, p=0.01

CT-scans were only performed in the intervention group.

Removal versus retention of screws

Kaftandziev, 2015

Type of study:

Retrospective cohort

Setting:

University hospital

Country:

Macedonia

Source of funding:

Not reported

Inclusion criteria:

patients who had undergone open reduction and internal fixation of the ankle fracture with screw fixation of the disrupted

syndesmosis that was confirmed at the operation

-syndesmotic screw fixation from January 2011 to December 2012

Exclusion criteria:

postoperative infection, postoperative hardware failure, additional surgery due to postoperative complications, delayed syndesmosis repair, open fracture, direct crush injury and age under 18 years. Patients in whom the syndesmotic screw was retained because of advanced age, limited mobility or sedentary lifestyle were also excluded.

Patients were excluded from the study if their charts or radiographs were unaar aanleiding vanailable or incomplete.

N total at baseline:

Intact screw: 46

Removed screw: 23

Broken screw: 13

Important prognostic factors²:

Mean age: 49 (range: 19 to 71)

Sex:

66 % M

Danis-Weber B fractures:

33 patients

Danis-Weber C fractures: 49 patients

In 66 patients (80%) three cortices were engaged

In 16 patients (20%) quadricortical fixation was performed. ‘’The distribution of three- and quadricortical fixation in all three groups was almost equal’’

71 patients (86%) received one screw, 14% 2 screws.

Groups comparable at baseline? Yes, ‘’ There were no statistically significant differences in demographic features between the three defined groups.’’

However, the underlying data were not specified in the article. The removal of screws was based on the surgeon’s preference.

Removal of screw

Intact screw

Broken screw

Both not removed.

All patients: 3.5 mm stainless steel screws were used to fix the injured syndesmosis.

Length of follow-up:

12 months

Loss-to-follow-up:

Not applicable

(136 patients met the inclusion criteria and were invited for the study. 82 patients (60%) agreed to take part.)

Incomplete outcome data:

Not applicable

All outcomes were measured after 12 months.

Functional outcomes

AOFAS (mean, SD)

Intact: 83 (12.2)

Removed: 85.5 (11.6)

Broken: 92.5 (6.02)

Intact versus broken: p=0.01

Intact versus removed: p=0.48

Broken versus removed: p=0.15

Patient satisfaction

VAS satisfaction (scale 0-10, 10 denoting maximum satisfaction)

Mean, SD

Intact: 8.43 (1.2)

Removed: 8.78 (1)

Broken: 9.92 (1.1)

P=0.34

Author’s conclusion:

Our study does not support routine removal of the syndesmosis screw. Overall, the patients with broken screw had the best clinical outcome and the most favourable VAS results.

Comments

The decision for screw removal was based upon individual surgeon preference.

Boyle, 2014

Type of study:

RCT

Setting:

Country:

New-Zealand

Source of funding:

None: ‘’ No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.’’

Inclusion criteria:

an age between 16 and 65 years, a displaced distal fibular fracture with associated tibiofibular

diastasis (Orthopaedic Trauma Association

(OTA) type 44-B1, 44-B2, 44-C1 or 44-C2), and a fracture that had occurred < 48 hours prior to presentation to

Hospital

Exclusion criteria:

an open fracture, a pre-existing

abnormality of the ankle, a neurovascular injury, a posterior malleolar fracture involving > 10% of the tibial articular surface on lateral radiographs, a tibial plafond fracture (although those with a medial malleolar fracture

and/or a posterior malleolar fracture were included), a diaphyseal tibial fracture, a pathological fracture, polytrauma,

cognitive impairment and pregnancy

N total at baseline:

Intervention: 37

Control: 35

Important prognostic factors²:

Age, mean (SD)

I:30.8 (12.8)

C:36.2 (14.1)

P=0.154

Sex:

I: 73% M

C:76% M

P=0.485

Smoking:

I: 35%

C: 32%

P=0.843

Time from injury to surgery: mean days, (SD)

I: 4.8 (4.7)

C: 4.0 (3.2)

P=0.499

Syndesmosis screw location (mm from tibial plafond) mean, (SD)

I: 16.4 (8.4)

C: 19.4 (8.3)

P=0.223

Groups comparable at baseline?

Yes, groups were comparable at baseline

Screw removal three months post-operatively

(mean: 116 days, 95%CI: 81 to 177 days)

Screw retention

Both groups:

Using prophylactic antibiotics, general anaesthesia, a lower limb tourniquet and intra-operative fluoroscopy, the fibular fracture was reduced and stabilised through a longitudinal lateral incision using 3.5 mm lag screw and plate fixation. A total of four patients (two in each group) also underwent 3.5 mm lag screw fixation of a medial malleolar fracture. Disruption of the syndesmosis was then confirmed by observing diastasis of > 2 mm on a

fluoroscopic mortise view of the ankle, with a bone hook

generating direct lateral stress on the fibula. Closed reduction of the diastasis was then performed without direct visualisation

using a reduction clamp applied at the level of

the tibial plafond with the ankle held at 10° of dorsiflexion and fluoroscopy confirming restoration of syndesmotic and

articular congruence. A single, non-cannulated, fully

threaded, 4 mm cortical screw (Synthes/DePuy, West Chester, Pennsylvania), engaging four cortices, was then introduced for fixation of the syndesmosis. This screw was introduced without compression, parallel to and 15 mm to

20 mm proximal to the tibial plafond, through the fibular

plate when possible and angled anteriorly from the fibula to tibia. A below-knee Plaster of Paris back-slab was applied. All patients commenced touch weight-bearing immediately

post-operatively and were reviewed in clinic ten to 14 days later with removal of the sutures and the application of a below-knee resin cast.

Length of follow-up:

1 year

Loss-to-follow-up:

Intervention:

N (%) 11 (29.7%)

Reasons: Patient unable to be contacted

Control:

N (%) 10 (28.6%)

Reasons: patient unable to be contacted

Incomplete outcome data:

Same as loss-to-follow-up

All outcomes were measured after 12 months

Functional outcomes

OMAS, mean (SD)

I: 86.7 (13.6)

C: 82.4 (19.8)

Mean difference between groups (95%CI)

4.3 (-5.2 to 13.9)

AOFAS

I: 90.1 (13.6)

C: 88.6 (13.3)

Mean difference between groups (95%CI)

1.5 (-6 to 9.1)

AAOS, mean (SD)

I: 91.8 (13.3)

C: 87 (16.4)

Difference between groups (95%CI)

4.8 (-3.5 to 13.2)

Pain

VAS scale, mean (SD)

I: 0.66 (0.85)

C: 1.03 (1.35)

Mean difference between groups (95%CI)

-0.38 (-1.01 to 0.26)

Range of motion

Ankle dorsiflexion

Mean degrees (SD)

I: 13.0 (9.0)

C: 10.2 (5.2)

Mean difference between groups (95%CI)

2.7 (-1.4 to 6.9)

Ankle plantar flexion

Mean degrees (SD)

I: 31.2 (14.7)

C: 33.6 (9.3)

Mean difference between groups (95%CI)

-2.3 (-9.3 to 4.6)

Radiological outcomes

Tibiofibular clear space

Mean mm (SD)

I: 5.3 (1.1)

C: 5.0 (1.0)

Mean difference between groups (95%CI)

0.34 (-0.28 to 0.95)

There were no intra-operative complications. During

the first post-operative year, one patient in the screw

retention group developed recurrent diastasis with a broken syndesmotic screw four months post-operatively,

which was managed with revision surgery. Two patients in the screw removal group developed a wound infection

after removal of the screw. This was a superficial infection in one managed with antibiotics with resolution, and a deep infection in the other managed with antibiotics and

debridement, followed by removal of the fibular plate with resolution.

No further reasons for high follow-up reported (what has been done to contact the patients?)

Tucker, 2013

Type of study:

Retrospective cohort

Setting:

Single centre

Country:

Ireland

Source of funding:

No funding or conflicts of interest

Inclusion criteria:

-adult patients requiring ankle and syndesmotic fixation (ORIF performed) between 1 January 2008 and 31 December 2010

-all syndesmotic screw fixations by any consultant surgeon in the hospital

Exclusion criteria:

-ankle fixation not involving a syndesmosis screw

-no available address or contact details on clinical records

N total at baseline:

Intervention: 43

Control: 20

Important prognostic factors²:

age (range):

I:41.84 (16-80)

C:45.3 (19-78)

Sex:

I: 60.5% M

C: 45% M

SER injuries: 15 patients (23.8%)

PER injuries: 48 patients (76.2%)

There were no Danis-Weber Type A injuries.

Groups comparable at baseline?

Based on available data, there were more males in the removal group. There were not enough factors reported to conclude whether groups were comparable.

Removal of the screw (10-12 weeks post ORIF), mean time 83.44 days, range from 47 to 178 days).

Retaining screw

In both groups, the operating surgeon decided upon the size of the screw, screw position and number of cortices to be engaged. Postoperatively, all patients were kept strictly non-weight-bearing for 6 weeks in cast. After this period, partial weight-bearing was allowed and the choice to retain or remove the syndesmotic screw at 10-12 weeks post ORIF was based on consultant preference (no standardised protocol).

Length of follow-up:

Minimum of 10 months

Mean (range) in days

I: 31.45 (10 – 43)

C: 28.57 (26 – 42)

Loss-to-follow-up:

Not applicable

Incomplete outcome data:

(Of the 91 identified patients, 22 had no up-to-date contact information on their medical file and six patients refused to participate. Total: n=43.)

Functional outcomes

OMAS, mean (SD)

I: 75 (12.9)

C: 81.5 (19.3)

P=0.107

Mean difference: -6.5 (95%CI: -15.8 to 2.8)

Multivariate analysis:

Equal variances assumed (due to impact of ceiling effect OMAS score) I versus C:

-7.2 (-15.1 to 1.1) p=value 0.088

For the removal group, time from surgery to removal was not statistically significant (p=0.148)

Costs:

Average costs for removal of metalwork:

£1372

Avoidance of routine screw removal, savings of £54.880 for 40 patients(not specified). Additional non-financial benefits included reduced morbidity from general anaesthesia and general surgery procedure as well as maximising operating theatre efficiency

Comments:

the choice to retain or remove the syndesmotic screw at 10-12 weeks post ORIF was based on consultant preference (no standardised protocol).

Study reference

(first author, publication year)

Describe method of randomisation¹

Bias due to inadequate concealment of allocation?²

(unlikely/likely/unclear)

Bias due to inadequate blinding of participants to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of care providers to treatment allocation?³

(unlikely/likely/unclear)

Bias due to inadequate blinding of outcome assessors to treatment allocation?³

(unlikely/likely/unclear)

Bias due to selective outcome reporting on basis of the results?⁴

(unlikely/likely/unclear)

Bias due to loss to follow-up?⁵

(unlikely/likely/unclear)

Bias due to violation of

intention to treat analysis?⁶

(unlikely/likely/unclear)

Kortekangas, 2015

‘’A computer-generated randomisation list was created by a

biostatistician. Randomisation was performed in randomly

varying blocks, with the block size varying among 4, 6 and 8.’’

Unlikely. ‘’a research assistant who was not involved in patient care sealed

the randomisation lists into numbered, opaque envelopes to, ensuring concealment. After repair of the bony injuries, in the

operating room an assistant nurse opened a numbered envelope

containing the information of the method of syndesmosis

fixation’’

Unlikely, patients did not know which intervention they received

Likely, the surgeon could not be blinded to the treatment

Likely, the radiologist assessing intraoperative and follow-up CT and CBCT was not blinded to the method of syndesmosis fixation Orthopaedic surgeon who carried out the latest follow-up visit was blinded to the interventions and earlier functional and radiological results, although it is not clear whether this was a different surgeon than the surgeon who had performed the operating procedure.

Unlikely, outcome measures reported at clinicaltrials.gov in December 2012 were described in the study

Unlikely, low loss to follow up, 2 people in the control group did not want to participate in the follow-up visits

Unlikely; although ITT was not completely followed (not all patients who received the intervention were analysed), it is not likely that this has influenced the results.

Laflamme, 2015

‘’Concealed randomization was obtained with the randomizer.org software and stratified by centres’’

Unlikely, sealed opaque envelopes were used.

Unlikely, patients were blinded to the type of fixation until the end of the study

Likely, the surgeon could not be blinded to treatment

Unlikely, evaluators were blinded to the type of fixation until the end of the study.

Unclear, no registration of study protocol

ITT was not followed.

Boyle, 2014

‘’ patients in the study were randomized … using a computer random number generator with random permuted blocks and an allocation ratio of one-to-one and six patients per block’’

Unclear, about concealment: ‘’Randomisation was performed independently and was concealed from those managing the patients’’

Likely: ‘’allocation was revealed at the end of the first post-operative visit’’.

Likely: not possible to blind for screw removal

Likely: there was no blinding of outcome assessors

Unclear, no registration of study protocol

Likely, high loss to follow-up (approximately 28% in both groups). Study was terminated when one-year follow-up data were available for 25 patients in each group. This was part of the protocol.

Likely, no ITT followed

Study reference

(first author, year of publication)

Bias due to a non-representative or ill-defined sample of patients?¹

(unlikely/likely/unclear)

Bias due to insufficiently long, or incomplete follow-up, or differences in follow-up between treatment groups?²

(unlikely/likely/unclear)

Bias due to ill-defined or inadequately measured outcome ?³

(unlikely/likely/unclear)

Bias due to inadequate adjustment for all important prognostic factors?⁴

(unlikely/likely/unclear)

Seyhan, 2015

Likely, in the control group (syndesmotic screw) all screws were removed after three months

Unlikely, only one patient excluded from baseline due to incomplete information

Likely, outcome assessors were not blinded due to retrospective characteristic of the study

Likely, no adjustment for potential confounding factors

Cottom, 2009

Unlikely

Unclear, not reported at which time-point the assessments were done.

A shortened AOFAS-score was used, but they did not explain why.

Likely, no adjustment for potential confounding factors

Thornes, 2005

Unlikely

Unlikely, no loss to follow up.

Range of motion was mentioned in methods but no results were reported; different post-operative weight bearing protocols. Only CT-scan in intervention group.

Likely, no adjustment for potential confounding factors

Tucker, 2013

Unclear which reasons for screw removal were applicable in the removal group. It could be that these patients differed from the ‘retain’ group.

Unclear,

Unlikely

Unlikely, adjustment for potential confounding factors

Kaftandziev, 2015

Likely, the decision for screw removal was based upon individual surgeon preference

Unlikely

Unlikely

Likely, no adjustment for potential confounding factors