Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Yang et al., 2014

PS., study charac-teristics and results are extracted from the SR (unless stated otherwise)

SR and meta-analysis of RCTs

Literature search up to July 2013

A: Hedbeck, 2011

B: Calder, 1996

C: Raia, 2003

D: Davison, 2001

E: Stoffel, 2013

F: Jeffcote, 2010

Source of funding:

No source of funding mentioned.

No conflict of interest reported.

Setting and Country:

Sweden, UK (2x), USA, Australia (2x).

Inclusion criteria SR:

- Patients with displaced femoral neck fractures

- Patients aged ≥65 years

- Follow-up ≥12 moths

Exclusion criteria SR:

- non-displaced femoral neck fractures, pathologic fractures or rheumatoid arthritis

- non-first time artificial joint replacement

- Previous osteosynthesis

-Patients with severe cognitive dysfunction

6 studies included

Important patient characteristics at baseline:

A total of 982 patients were included.

N, mean age

A: 120 patients, 86.4 yrs

B: 250 patients, 85 yrs

C: 115 patients, 82.1 yrs

D: 187 patients, 74 yrs (extracted from article itself)

E:259 patients, 82.0 yrs

F:51 patients, 80.1 (bipolar) – 81.4 (unipolar) yrs

Groups are comparable at baseline.

Describe intervention:

Bipolar hemiarthroplasty

N=487

Approach used:

A: Anterolateral approach

B: Lateral approach

C: Posterolateral approach

D: Unclear

E: Lateral approach

F: Anterolateral approach

Describe control:

Unipolar hemiarthroplasty

N=495

Approach used:

A: Anterolateral approach

B: Lateral approach

C: Posterolateral approach

D: Unclear

E: Lateral approach

F: Anterolateral approach

End-point of follow-up:

A: 12 months

B: > 12 months

C: 12 months

D: 3 years

E: 12 months

F: 2 years

For how many participants were no complete outcome data available?

(intervention/control)

A: (1/0)

B: Not mentioned

C: (13)

D: Not mentioned

E: (3)

F: (10)

Outcome measure-1

Defined as return to pre injury rate

Effect measure: OR [95% CI]:

B: 1.64 [0.97, 2.77] 21 mths

C: 0.91 [0.29, 2.87] 12 mths

D: 1.20 [0.67, 2.13] 36 mths

Pooled effect (fixed effects model):

1.36 [95% CI 0.94 to 1.96] favoring Bipolar hemiarthroplasty.

Heterogeneity (I²): 0%

Outcome measure-2

Defined as Harris Hip Score

Effect measure: standard mean difference [95% CI]:

A: -0.04 [-0.40, 0.31] 12 mths

E: -0.02 [-0.27, 0.22] 12 mths

Pooled effect (fixed effects model):

-0.03 [95% CI -0.23 to 0.17]

Heterogeneity (I²): 0%

Outcome measure-3

Defined as Dislocation rate at 1 year or end of follow-up

Effect measure: OR [95% CI]:

A: 2.03 [0.18, 23.06] 12 mths

B: 0.56 [0.05, 6.21] 21 mths

C: 1.09 [0.07, 17.90] 12 mths

D: 0.46 [0.04, 5.14] 12 mths

E: 0.95 [0.06, 15.30] 12 mths

Pooled effect ( fixed effects model): 0.87 [95% CI 0.29 to 2.60] favoring Bipolar hemiarthroplasty.

Heterogeneity (I²): 0%

Facultative:

Author’s conclusion:

“This review shows that both bipolar and unipolar hemiarthroplasty for the treatment of elderly patient suffering

displaced femoral neck fracture achieve similar and clinically satisfactory outcome in short-term follow-up.”

There are methodological limitations, but these limitations are not considered enough to downgrade.

Level of evidence: GRADE

Return to pre injury rate: LOW (-2 for imprecision: No OIS and wide CI’s)

Harris Hip Score :VERY LOW (-2 for imprecision: No OIS and wide CI’s, -1 for inconsistency: heterogeneity of I² 95%)

Dislocation Rate: LOW (-2 for imprecision: No OIS and wide CI’s)

No sensitivity analyses performed.

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Li et al., 2013

PS., study characteristics and results are extracted from the SR (unless stated otherwise)

Hemiarhtroplastie

SR and meta-analysis of RCTs

Literature search up to December 2012

A: DeAngelis, 2012

B: Emery, 1991

C: Figved, 2009

D: Parker, 2010

E: Sadr and Arden, 1977

F: Sonne-Holm, 1982

G: Taylor, 2012

Study design: RCT parallel design.

Setting and Country: not mentioned

Source of funding:

No support or funding to report for SR.

Conflict of interest:

No competing interests declared for SR.

Inclusion criteria SR:

- Patients aged ≥70 years

- Underwent primary hemiarthroplasty for unilateral femoral neck fractures with or without cement (CH/UCH)

- RCTs

Exclusion criteria SR:

- non-randomized trials

- quasi-randomized trials

7 studies included

Important patient characteristics at baseline:

A total of 1125 patients were included.

N, (CH/UCH)

A: 130 patients (66/64)

B: 53 patients (27/26)

C: 230 patients (115/115)

D: 400 patients (200/200)

E: 40 patients (20/20)

F: 112 patients (57/55)

G: 160 patients (80/80)

mean age (CH/UCH)

A: 81.8/82.8 yrs

B: 78/79.6 yrs

C: 83.4/83.0 yrs

D: 83/83 yrs

E: 77/78.4 yrs

F: 76 yrs

G: 85.3/85.1 yrs

Sex (CH/UCH):

A: 76.9% Female

B: 88.9/84.6% Female

C: 78/74% Female

D: 77% Female

E: 65/85% Female

F: 75% Female

G: 69% Female

Type of prosthesis:

A: LD/Fx/ Beaded FullCoat

B: Thompson/Austin Moore

C: Spectron/HA-coated

D: Thompson/Austin Moore

E: Thompson/Thompson

F: Austin Moore/Austin Moore

G: Exeter/Alloclassic

Groups comparable at baseline

Describe intervention:

Cemented primary hemiarthroplasty

Describe control:

Uncemented primary hemiarthroplasty

End-point of follow-up:

A: 12 months

B: 18 months

C: 24 months

D: 60 months

E: 17 months

F: 12 months

G: 24 months

For how many participants were no complete outcome data available?

(intervention/control)

A: 5 (2/3)

B: 0

C: 10 (3/7)

D: 0

E: 0

F: 37 (22/15)

G: 0

Total: 52

Outcome measure-Postoperative hip function

Defined as Number of patients requiring assistance with ambulation at 1 year.

Effect measure: OR [95% CI]:

A: 0.64 [0.32, 1.31]

B: 0.26 [0.08, 0.83]

C: 0.54 [0.15, 2.00]

E: 0.76 [0.15, 3.86]

G: 0.36 [0.14, 0.93]

Pooled effect (fixed effects model):0.48 [95% CI 0.31 to 0.76] favoring cemented hemiarthroplasty

Heterogeneity (I²): 0%

Outcome measure-Residual Pain

Defined as pain measured with various scoring instruments with a follow-up ranging from 12-60 months.

Effect measure: OR [95% CI]:

B: 0.12 [0.03, 0.50]

C: 1.52 [0.72, 3.21]

D: 0.47 [0.27, 0.82]

E: 0.43 [0.12, 1.57]

F: 0.47 [0.16, 1.39]

G: 0.55 [0.20, 1.50]

Pooled effect (random effects model):

0.52 [95% CI 0.29 to 0.95] favoring cemented hemiarthroplasty

Heterogeneity (I²): 58%

Outcome measure-Reoperation rate

Defined as reoperation rate with a follow-up ranging from 12-24 months

Effect measure: OR [95% CI]:

A: 2.95 [0.12, 73.87]

C: 1.07 [0.43, 2.63]

D: 0.59 [0.27, 1.28]

G: 0.49 [0.09, 2.74]

Pooled effect (fixed effects model):

0.76 [95% CI 0.44 to 1.30] favoring cemented hemiarthroplasty

Heterogeneity (I²): 0%

Outcome measure-Implant related complications

Includes complications as intraoperative and postoperative periprosthetic fractures, prosthesis loosening and dislocation.

Effect measure: OR [95% CI]:

A: 0.64 [0.10, 3.93]

C: 0.41 [0.14, 1.23]

D: 0.10 [0.02, 0.45]

E: 0.04 [0.00, 0.42]

F: 0.14 [0.04, 0.54]

G: 0.08 [0.03, 0.22]

Pooled effect (fixed effects model):

0.15 [95% CI 0.09 to 0.26] favoring cemented hemiarthroplasty

Heterogeneity (I²): 42%

Facultative:

Conclusion of author:

Compared with uncemented hemiarthroplasty, the existing evidence indicates that cemented hemiarthroplasty

can achieve better hip function, lower residual pain and less implant-related complications with no increased risk of

mortality, cardiovascular and cerebrovascular complications, general complications, local complications and reoperation

rate in treating elderly patients with femoral neck fractures.

Personal remarks: In general there is a low risk of bias in the SR and incorporated studies.

Level of evidence: GRADE

Postoperative hip function: MODERATE (-1 imprecision, No OIS)

Residual Pain: VERY LOW (-2 inconsistency, because heterogeneity I² 58% and multiple definitions of residual pain; -1 (-1 imprecision, No OIS)

Reoperation rate: MODERATE (-1 imprecision, No OIS)

Implant related complications: MODERATE (-1 imprecision, No OIS)

Remarks: All kinds of prosthesis were included, without discrimination for this review.

Kannan, 2012

PS., study characteristics and results are extracted from the SR (unless stated otherwise)

Hemiarhtroplastie

SR of National Registries

National Registry of:

A: Australia

B: England and Wales

C: Italy

D: Sweden

Study design: National registry.

Source of funding:

Not mentioned

Conflict of interest:

The authors declare that they have no conflict of

interest.

Inclusion criteria SR:

- National registries

Describe intervention:

Uncemented THR

Describe control:

cemented THR

End-point of follow-up:

Up to 9 years

Outcome measure-Failure of prostheses 0-1 month

Australia: 7,29 p<0,001

Outcome measure-Failure of prostheses 1 month – 7 years

Australia HR 2,04 p=0,004

Outcome measure-Failure of prostheses 3 years

England and Wales: HR 2,9

Facultative:

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Chan, 1975

Anterior - Posterior

Type of study:

Retrospective cohort

Setting:

Single Centre

Country:

UK

Source of funding:

Not mentioned. Remark: decision was made to exclude Austin-Moore prostheses, reason was not mentioned.

Inclusion criteria:

- Displaced subcapital fractures treated by prosthetic replacement;

- Patients with Thompson prostheses;

Exclusion criteria:

- Austin Moore prostheses;

N total at baseline:

Intervention: 136

Control: 107

Important prognostic factors²:

age:

I: 77,4

C: 76,9

Previous operation on hip:

I: 7%

C: 6%

Sex:

I: 8 % M

C: 7 % M

Urinary incontinence

I: 28%

C: 22%

Groups comparable at baseline.

Describe intervention:

Posterior approach

The posterior (southern) approach is that described by Moore (1959). The lateral rotators are divided and the hip is most stable in a position of extension and lateral rotation

Describe control:

Anterior approach

The anterior approach is basically that described by Smith-Petersen with minor modifications (Henry 1957). The hip capsule is divided anteriorly and the prosthesis is most stable with the hip in flexion and medial rotation.

Length of follow-up:

6 weeks

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

3

Outcome measures and effect size:

Wound infection

I: 14/136

C: 3/107

Dislocation

I: 19/136

C: 1/107

Appointment to control- or intervention group was based on preference of surgeon.

No P-values provided.

Tsukada 2010

MIS Anterior - Posterior

Type of study:

Prospective cohort

Setting:

Single Centre

Country:

Japan

Source of funding:

The authors did not receive and will not receive any benefits or funding from any commercial party related directly or indirectly to the subject of this article.

Inclusion criteria:

- Displaced femoral neck fractures;

Patients aged > 70 years;

Exclusion criteria:

N total at baseline:

Intervention: 39

Control: 44

Important prognostic factors²:

age ± SD:

I: 81,9 ± 7,9

C: 80,4 ± 6,9

P=0,31

Sex:

I: 18 % M

C: 18 % M

P=0,97

ASA 1 or 2

I:84,6%

C:79,5%

P=0,54

Preinjury HHS

I: 33,7 ± 5,9

C: 33,5 ± 6,3

P=0,85

Groups comparable at baseline.

Describe intervention:

Posterior approach

Our operative technique for the conventional BHA was the posterior approach without a mini-incision technique. The incised short external rotators were repaired back to the major trochanter.

Identical clinical paths were used for both groups. The protocol included drain removal 1 day after surgery, after which full weight bearing was allowed. No deep venous thrombosis prophylactic agents were used in either group, but patients wore elastic stockings for 10 days after the operation.

Describe control:

MIS Anterior approach

In the MIS-BHA group, we used a modified musclepreserving anterior approach via the Smith-Petersen interval with a Judet’s table. The patients were placed in the supine position on a standard orthopedic table. The skin incision was made distally and laterally to the anterosuperior iliac spine. The anterior aspect of the hip joint was exposed by holding apart the rectus femoris muscle medially and the gluteus minimus muscle laterally. To obtain the femoral exposure without Judet’s table, we released the superolateral capsule and elevated the femur anteriorly.

Identical clinical paths were used for both groups. The protocol included drain removal 1 day after surgery, after which full weight bearing was allowed. No deep venous thrombosis prophylactic agents were used in either group, but patients wore elastic stockings for 10 days after the operation.

Length of follow-up:

12 months

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Dislocation

I: 1/39

C: 0/44

P=0,28

HHS at 12 months

I: 27,2 ± 7,2

C: 29,9 ± 7,4

P=0,10

Renken 2012

MIS Anterior - anterolateral

Type of study:

RCT

Setting:

Single Centre

Country:

Germany

Source of funding:

No competing interests to declare, this research was funded by the

University Lübeck.

Inclusion criteria:

- patients with an indication for bipolar

hemi-endoprosthesis of the hip for proximal femoral fractures (Garden 3 and 4);

- Age above 60;

- Ability to give informed consent.

Exclusion criteria:

- Active infections;

- History of infection in the injured hip joint;

- immobility;

- current treatment for malignant disease;

- suspicion for malignant disease;

- neurological deficits of the lower extremity;

- Skin diseases in area of incision.

N total at baseline:

Intervention: 27

Control:30

Important prognostic factors²:

Median age ± SD:

I: 87,5 ± 7,0

C: 84 ± 5,8

Not significant

Sex:

I: 11 % M

C: 13 % M

Not significant

Groups comparable at baseline.

Describe intervention:

Anterolateral approach (WJA)

The antero-lateral approach initially described by Watson-Jones is well published. The WJA procedure is performed in a supine

position.

Describe control:

MIS anterior approach (DAA)

the DAA procedure is performed in a supine position. The skin incision starts 2–3 cm distal and lateral to the anterior superior iliac spine along the medial border of the M. tensor fascia lata. This approach uses an inter-muscular and inter-nerval plane between the artorius, rectus femoris, and tensor fasciae latae [34].

Length of follow-up:

40 days

Loss-to-follow-up:

Intervention:

N=3 (10%)

Reasons:

- 2 converted to THP

- 1 Stem size not available

Control:

N=0 (0%)

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Postoperative pain at day 5 (VAS):

I:4 ±1,6

C: 2 ± 1,4

P=0,14

Postoperative pain at day 16 (VAS):

I:1 ±1,33

C: 2 ± 1,53

P=0,035

Postoperative pain at day 40 (VAS):

I: 1 ± 0,82

C: 0 ± 0,31

P<0,001

Modified Four-item Barthel (Functioning):

I: 30 ± 11,9

C:42,5 ± 14,6

P=0,013

For the femoral shaft preparation specific retractors and broach handles where used (Stryker MIS Set, Duisburg, Germany).

In all cases, a cemented anatomical ABGW II stem was used (Stryker, Duisburg, Germany) with 40 g of PMMAbone cement with Gentamicin (Palacos-RG, Heraeus Medical GmbH, Wehrheim, Germany). The bipolar head used was a UHRW (Universal Head Bipolar System, Stryker, Duisburg, Germany) in all cases.

Auffarth 2011

Anterior - lateral

Type of study:

RCT

Setting:

Single Centre

Country:

Austria

Source of funding:

No support or funding of any kind has been received for this work.

Inclusion criteria:

- Patients with femoral neck fractures.

Exclusion criteria:

- prior hip surgery on the affected side;

- preexisting severe neurodegenerative disorders;

- previously immobile before the injury.

N total at baseline:

Intervention: 24

Control:24

Important prognostic factors²:

Median age ± 95% CI:

I: 83,7 (69,7 – 94,4)

C: 82,6 (73,5 – 93,9)

Sex:

I: 17 % M

C: 25 % M

ASA class

I: 3,0

C:3,2

Groups comparable at baseline.

Describe intervention:

Direct Lateral approach (Hardinge)

The patient is placed in supine position, and the operative limb is flexibly draped in a sterile fashion. From the upper margin of the greater trochanter, a 10-cm to 12-cm a skin incision is made following the axis of the femur and ending at approximately the same height with the anterior superior iliac spine. The fascia lata is incised and split both proximally and distally. Trochanteric bursa tissue is excised, and the tendons of the gluteus medius and vastus lateralis muscles are exposed. At the junction between anterior and middle third of the greater trochanter, a musculoperioseteal sleeve is separated from the greater trochanter using electrocautery. Next, the ventral fibers of the gluteus medius and the vastus lateralis muscles are retracted together as a unit. This exposes the hip capsule that is subsequently excised

Describe control:

MIS Anterior approach (Smith Peterson)

The patient is positioned on the operating table in the supine position. Both legs are draped in a sterile fashion. A straight skin incision 10 cm to 12 cm in length is performed; the starting point is 1 cm distal to and 1 cm lateral to the anterior superior iliac spine, and the end point is at the same height in line with the upper aspect of the greater trochanter. The anterior rim of the tensor fascia lata muscle is identified, and the fascia is dissected in line with the skin incision. Using blunt dissection, crossing vessels are cauterized and dissected around the medial side of the tensor muscle. The rectus femoris muscle is then mobilized medially, whereas the gluteus medius and gluteus minimus muscles are held laterally. This exposes the hip capsule that is subsequently excised. For improved access to the femur, the surgical limb is positioned in external rotation under the contralateral leg, and the operating table is lowered by a 30-degree angle.

Length of follow-up:

6 months

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Postoperative pain at 48 hours:

I: 3,1 (2-7)

C: 4,0 (2-6)

P=0,01

Postoperative pain at 96 hours:

I: 2,0 (1-3)

C: 2,9 (1-5)

NS

HHS at six months

I: 67,9 (26-92)

C: 66,1 (38-90)

Wound infection

I: 1/24

C: 1/24

Sikorski 1981

Anterolateral - Posterior

Type of study:

RCT

Setting:

Single centre

Country:

UK

Source of funding:

Not mentioned

Inclusion criteria:

- displaced femoral neck fracture;

- age ≥70 years.

Exclusion criteria:

- excluded if the consultant in charge felt that a particular form of treatment was strongly indicated in that patient. (most common grounds for exclusion were a long-standing fracture, rheumatoid arthritis, the suspicion of a malignant deposit in the hip and where the distinction between a Stage II and Stage III fracture could not be made confidently)

N total at baseline:

Intervention: 57

Control:57

Important prognostic factors²:

age ± SD:

Not mentioned

Sex:

I: 16% M

C: 9 % M

Not possible to assess.

Describe intervention:

Posterior approach

Thompson hemiarthroplasty through a Moore posterior approach (posterior Thompson).

Describe control:

Anterolateral approach

Thompson hemiarthroplasty through a McKee anterolateral approach (anteriorThompson).

Length of follow-up:

24 months

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Wound infection

I: 5/57

C: 9/57

Dislocation

I: 1/57

C: 1/57

Parker 2015

Lateral - Posterior

Type of study:

RCT

Setting:

Single Centre

Country:

UK

Source of funding:

This study received internal funding from the Peterborough

Hospitals Hip Fracture Fund.

Inclusion criteria:

- patients with a displaced intracapsular fracture that was to be treated with cemented hemiarthroplasty.

Exclusion criteria:

- patients who declined to participate;

- patients without the capacity to give informed consent and the next of kin or patient’s representative was unavailable to give assent or declines assent;

- patients admitted when the lead researcher (MJP) was not available to supervise treatment;

- pathological fracture requiring a specific prosthesis or procedure;

- patients treated conservatively due to poor health or late presentation.

- Patients in which an uncemented arthroplasty was used to avoid bone cement;

- patients in whom a total hip replacement was chosen as the primary treatment in preference to a hemiarthroplasty;

-patients who were entered in one of two concurrently running randomised studies or hemiarthroplasty versus internal fixation for male patients or hemiarthroplasty versus total hip replacement.

N total at baseline:

Intervention: 108

Control:108

Important prognostic factors²:

age (range):

I:84,3 (61-100)

C: 83,6 (65-99)

P=-,5

Sex:

I: 7% M

C: 9% M

P=0,8

ASA class 1 or 2

I:41%

C: 40%

P=1,0

Groups comparable at baseline?

Describe intervention:

Lateral approach

For their surgery all patients were positioned on their side and a minimal surgical exposure used, in which whenever possible incisions were less than 10cms in length

The lateral approach involved splitting the tendon to the gluteus medius muscle with two thirds being left intact and a third of the muscle retracted anteriorly to expose the anterior hip joint capsule. The capsule was opened with a T shaped cut and later repaired after insertion of the prosthesis.

Describe control:

Posterior approach

For their surgery all patients were positioned on their side and a minimal surgical exposure used, in which whenever possible incisions were less than 10cms in length

For the posterior approach the piriformis tendon was preserved and the other short rotators cut and retracted posteriorly. An L shaped cut to the capsule was made along the line of the femur and then posteriorly immediately distal to the piriformis tendon. After insertion of the prosthesis both the capsule and tendons to the short rotators were repaired.

Length of follow-up:

12 months

Loss-to-follow-up:

Intervention:

N=0 0(%)

Control:

N=0 (0%)

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Deep Wound Infection

I:0/108

C: 2/108

P=0,5

Dislocation

I: 2/108

C: 1/108

P=1,0

Leonardsson 2012

Anterolateral - posterior

Type of study:

Prospective cohort

Setting:

Registry

Country:

Sweden

Source of funding:

This work was supported by grants from the Swedish Research Council, the Greta and Johan Kock Foundation, the H. Järnhardt Foundation, the Malmö University Hospital Research Foundation, and the Research and Development Council of Region Skåne, Sweden.

Inclusion criteria:

- a fracture-related hemiarthroplasty (acute fracture or secondary procedure after failed internal fixation);

- use of a modular unipolar or bipolar implant

Exclusion criteria:

- monoblock-type implants (i.e. Austin-Moore, Thompson, and ETS Endo)

N total at baseline:

Intervention: 12086

Control: 11423

Important prognostic factors²:

No further specification for different groups.

Not possible to assess whether groups are comparable.

Describe intervention:

anterolateral approach (Hardinge + Gammer)

the anterolateral transgluteal approach with the patient in a lateral position (Gammer 1985), the anterolateral transgluteal approach with the patient in a supine position (Hardinge 1982),

Describe control:

Posterior approach (Moore)

the posterior approach with the patient in a lateral position (Moore 1957)

Length of follow-up:

Median 18 months

Loss-to-follow-up:

Intervention:

Not mentioned

Incomplete outcome data:

Completeness varied between 89% and 96%, no further specification.

Outcome measures and effect size:

Dislocation

HR 0,72 (0,58-0,89)

P=0,002

In favour of anterolateral approach.

Infection

HR 0,94 0,73 – 1,20)

p=0,6

Biber 2012

Lateral - Posterior

Type of study:

Retrospective cohort

Setting:

Single Centre

Country:

Germany

Source of funding:

Inclusion criteria:

- patients with femoral neck fracture;

Exclusion criteria:

- patients operated for hip

Osteoarthritis;

- revision surgery;

- rare individual

indications such as acetabular fracture or osteolysis.

N total at baseline:

Intervention: 217

Control: 487

Important prognostic factors²:

age ± SD:

I:80,3 ± 1,3

C:80,4 ± 0,9

NS

Sex:

I: 31% M

C: 30% M

Groups comparable at baseline.

Describe intervention:

Lateral approach (transgluteal)

The transgluteal approach was started by a straight lateral approach, dividing the M. gluteus medius longitudinally at a line dividing the width of the greater trochanter ventrodorsally by a ratio of 1:5. The vastogluteal continuity was preserved whenever possible. Joint capsule reconstruction was also part of the standard procedure.

Irrespective of the surgical approach the patients were mobilised as soon as possible, allowing for full weightbearing.

Describe control:

Posterior approach (dorsal)

The dorsal approach was performed by a lateral incision over the greater trochanter with a slight dorsal curve cranially. The piriformis tendon was divided close to its attachment and reinserted later. The joint capsule was opened by a T-shaped incision and reconstructed by default

Irrespective of the surgical approach the patients were mobilised as soon as possible, allowing for full weightbearing. After a dorsal approach a polystyrene foam wedge was placed between and elastically attached to the patient’s legs while supine, preventing the affected hip from internal rotation.

Length of follow-up:

Not mentioned

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Wound infection

I: 7/217

C:12/487

P=0,57

Dislocation

I: 1/217

C: 19/487

P=0,01

Yang 2010

Anterolateral - posterolateral

Type of study:

RCT

Setting:

Single Centre

Country:

China

Source of funding:

Not mentioned

Inclusion criteria:

- patients admitted for unilateral THA.

Exclusion criteria:

- history of previous surgery on the affected hip;

- inflammatory polyarthritis with severity to compromise postoperative mobility;

- pulmonary and heart insufficiency intolerant of surgery;

- cerebrovascular diseases accompanied by physical sequelae;

- body mass index (BMI)[30, and Crowe III-IV of DDH.

N total at baseline:

Intervention: 55

Control: 55

Important prognostic factors²:

age ± SD:

I: 59,47 ± 13,24

C: 55,82 ± 13,91

P=0,16

Sex:

I: 47% M

C: 55% M

P=0,45

Groups comparable at baseline.

Describe intervention:

MIS Anterolateral approach (OCM approach)

The skin incision was made on a line beginning at the anterior tubercle of the greater trochanter and extending along the femoral axis approximately 7 cm in length. Onefourth of the incision was over the trochanter and the rest was proximal. Tensor fasciae latae muscle and gluteus medius were revealed in the gap, followed by the anterior capsule. When a Z-shaped capsulotomy was made, the retractors on the inferior and superior aspects of the neck were moved from an extracapsular to an intracapsular position to expose the femoral neck.

Describe control:

Posterolateral approach

Not described

Length of follow-up:

3 years

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Postoperative pain at day 1

I: 30,76 ± 21,77

C:50,34 ± 13,73

P< 0,01

HHS at 3 months

I:83,80 ± 5,64

C:74,96 ± 7,47

P<0,01

Barthel index at 3 months

I: 93,45 ± 6,37

C:88,64 ± 9,64

P<0,01

Dislocation

I: 0/55

C: 0/55

Most patients did not have a femoral neck fracture.

Skoldenberg 2010

Anterolateral - posterolateral

Type of study:

Cross sectional

Setting:

Single Centre

Country:

Sweden

Source of funding:

Not mentioned

Inclusion criteria:

- patients with a femoral

neck fracture;

- non-pathological

displaced fracture (Garden III or IV);

- failed internal fixation after a femoral neck fracture

Exclusion criteria:

None mentioned

N total at baseline:

Intervention: 180

Control:192

Important prognostic factors²:

No further specification for different groups.

Not possible to assess whether groups are comparable.

Describe intervention:

Anterolateral approach (Gammer)

No further specification

Describe control:

Posterolateral approach (Moore)

No further specification

Length of follow-up:

3 months

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Dislocation

I:2/180

C: 17/192

Enocson 2009

Anterolateral - posterolateral

Type of study:

Prospective cohort

Setting:

Single Centre

Country:

Sweden

Source of funding:

Not mentioned

Inclusion criteria:

- patients with a non-pathological displaced femoral neck fracture (Garden III or IV) or a secondary THR.

Exclusion criteria:

Not mentioned

N total at baseline:

Intervention: 463

Control: 250

Important prognostic factors²:

age ± SD:

I: 77,2 ± 8,4

C: 76,1 ± 10,1

P=0,1

Sex:

I: 19% M

C: 20% M

Groups comparable at baseline.

Describe intervention:

Anterolateral approach

No further specification

Describe control:

Posterolateral approach

No further specification.

A posterior repair with re-attachment of the short external rotators and/or the posterior joint capsule was performed in 110 of the 250 posterolateral approaches

Length of follow-up:

Median 4,3 years

Loss-to-follow-up:

Not mentioned

Incomplete outcome data:

Not mentioned

Outcome measures and effect size:

Dislocation

I: 9/463

C: 32/250

Yang et al. 2014

Yes

Yes

Searched in Medline, Embase, Pubmed Central & Cochrane Central Register

Yes

But no references of excluded studies.

Yes

Not applicable

Yes

Risk of bias grap and Risk of bias summary included

unclear

yes

No:

Yes for SR.

No for the included studies (not all studies provided report on potential conflict of interest)

Kannan, 2012

No

Unclear

No

Yes

Not Applicable

No

Yes

no

Yes        

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)

Parker, 2015

opening of numbered sealed opaque envelopes containing details of the operation to be used. The envelopes were prepared by a person independent to the study

unlikely

unlikely

unlikely

unclear

unlikely

unlikely

unlikely

Renken, 2012

A randomization list was created with an external generator (Random.org, Dublin, Ireland) using atmospheric noise. The randomization was via dark blue envelopes opened shortly before surgery.

unlikely

unlikely

unlikely

unlikely

unlikely

unlikely

unclear

Auffarth, 2011

On the day of admission, patients were randomly assigned to each approach group by the first author using a randomization list (Seed: 53262004) generated online at www.randomisation.eu. Only the first author had access to this list.

unclear

unlikely

unlikely

unlikely

unlikely

unlikely

unclear

Yang, 2010

By simple randomization method with table of random digit, they were randomly assigned to the postlateral approach group (PL group) or to OCM approach group (OCM group) (n = 55).

unlikely

unlikely

unclear

unlikely

unlikely

unclear

unclear

Sikorski, 1981

As soon as the patient was admitted to the trial, a card was drawn from a box in the operating theatre. These cards randomly allocated patients to one of three groups: internal fixation, Thompson hemiarthroplasty through a McKee anterolateral approach (anterior Thompson). or Thompson hemiarthroplasty through a Moore internal fixation but the fracture was irreducible by manipulation and a hemiarthroplasty was therefore performed:

likely

unclear

likely

unclear

unlikely

likely

unclear

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)

Leonardsson 2012

unlikely

unlikely

unlikely

unclear

Biber 2012

unlikely

unclear

unlikely

unclear

Tsukada 2010

likely

unlikely

unlikely

likely

Skoldenberg 2010

unlikely

unclear

unlikely

unlikely

Enocson 2009

unlikely

unlikely

unlikely

unlikely

Chan 1975

likely

likely

unlikely

likely