Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Amato 2012

SR and meta-analysis of RCTs

Literature search up to 27 February 2012

n=16 included RCTs of which the following compared Shouldice to mesh techniques for hernia repair:

A: Barth 1998

B: Butters 2007

C: Danielsson 1999

D: Hetzer 1999

E: McGillicuddy 1998

F: Miedema 2004

G: Nordin 2002

H: Zieren 1998

The remaining RCTs compared Shouldice to non-mesh techniques; these are not addressed in this evidence table.

Inclusion criteria SR: randomised or quasi-randomised controlled trials (RCT) on the surgical treatment (non-laparoscopic techniques) of primary inguinal hernia in adults (>18 years)

Exclusion criteria SR: patients referred for femoral hernias

n=16 studies included, of which n=8 addressed the comparison mesh versus Shouldice.

N (patients); mean age (I/C)

A: 105; 53/51 yrs

B: 186; 56/51 yrs

C: 200; 58/56 yrs

D: 385; 58/53 yrs

E: 672; not reported

F: 101 hernias; 63/62

G: 300; age range 25-75

H: 160; 47/46

Sex:

A: 99% Male

B: 100% Male

C: 100% Male

D: 100% Male

E: 100% Male

F: 100% Male

G: 100% Male

H: 92% Male

Describe intervention: non-laparoscopic techniques, a.o. mesh:

A: Lichtenstein standard (n=54). Polypropylene mesh. Polypropylene 2/0

B: Lichtenstein standard (n=93). No details.

C: Lichtenstein standard (n=89). Polypropylene mesh

D: Lichtenstein (n=239 hernias) Polypropylene meshes. Polypropylene 2/0

E: Lichtenstein standard (n=371 hernias). Polypropylene mesh.

F: Lichtenstein standard (n=39). No details.

G: Lichtenstein (n=149 hernias). Marlex meshes. Polypropylene 2/0.

H: Plug and Patch (n=80). Polypropylene meshes and plugs.

Describe control: Shouldice

A: Shouldice standard (n=51) performing 4 layers. Polypropylene 2/0.

B: Shouldice (n=93). No details.

C: Shouldice standard (n=89) performing 4 layers. Non-absorbable monofilament

D: Shouldice Standard (n=171 hernias). 4 layers. PDS 2/0.

E: Shouldice standard (n=337 hernias) performing 4 layers. Non-absorbable monofilament.

F: Shouldice modified (n=41). 4 layers. Polypropylene.

G: Shouldice Standard (n=148 hernias). 4 layers. Polypropylene 2/0.

H: Shouldice modified (n=80). 4 layers. Unabsorbable monofilament.

End-point of follow-up:

A: 7 days

B: 52 months

C: 12 months

D: 3 months

E: 5 years

F: 7 years

G: 6 years

H: 25 months

For how many participants were no complete outcome data available?

A: none

B: 36

C: 8

D: none

E: 251

F: 29

G: 9

H: none

N.B. Results are presented for Shouldice versus mesh (control versus intervention)!

Recurrences

Defined as clinically manifest bulge or protrusion exacerbated by a Valsalva manoeuvre in the operated groin. The assessor could be a surgeon, a physician or the patient himself. Only studies with at least one year follow-up were considered.

Pooled effect (fixed effects model); 5 studies, n=1415 patients:

Peto OR 3.65 (95% CI 1.79 to 7.47) favoring mesh.

Heterogeneity (I²): 43.6%

Chronic pain

Defined as pain persisting for more than three months (only presence or absence)

Pooled effect (fixed effects model); 5 studies, n=1371 patients:

Peto OR 0.87 (95% CI 0.55 to 1.39).

Heterogeneity (I²): 59.6%

Postoperative complications

Wound infections

Pooled effect (fixed effects model); 7 studies, n=1938 patients

Peto OR 0.74 (95% CI 0.37 to 1.49).

Heterogeneity (I²): 0%

Seroma

Pooled effect (fixed effects model); 3 studies, n=1165 patients

Peto OR 0.96 (95% CI 0.37 to 2.50).

Heterogeneity (I²): 0%

Testicular atrophy

Pooled effect (fixed effects model); 3 studies, n=1155 patients

Peto OR 1.05 (95% CI 0.047 to 27.38).

Heterogeneity (I²): 56%

Haematoma

Pooled effect (fixed effects model); 3 studies, n=562 patients

Peto OR 0.64 (95% CI 0.25 to 1.64).

Heterogeneity (I²): 0%

Authors’ conclusion

Shouldice herniorrhaphy is a good technique for inguinal hernia repair compared to other non-mesh techniques, giving better results in terms of recurrence. When comparing Shouldice to mesh, the use of mesh is associated with a lower rate of recurrence. The quality of included studies, assessed with jaded scale, was low. Reasons for lost to follow-up were often not reported and the length of follow-up varied broadly among studies.

Bhatti 2015

Type of study: RCT

N=200

Setting: multicentre study

Country: Pakistan

Source of funding: not specified

Inclusion criteria:

Unilateral, primary, reducible inguinal hernia.

Exclusion criteria:

Patients with renal failure (serum creatinine more than 2mg/dl) and diabetes were excluded.

N total at baseline:

Intervention: 100

Control: 100

Important prognostic factors²:

Age ± SD:

53.25±6.768 (range 32-60)

Sex:

M 60.5%

F 39.5%

Groups comparable at baseline? Not specified.

Inguinal hernias were repaired by standard Lichtenstein (mesh) method (group L)

Inguinal hernias were repaired by Desarda’s (non-mesh) method (group D)

Length of follow-up:

30 days post-operation

Loss-to-follow-up:

I: N=0 (0%)

C: N=0 (0%)

Incomplete outcome data:

I: N=0 (0%)

C: N=0 (0%)

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic pain:

Not reported

Recurrences:

Not reported

Seroma formation:

I: 10/100 (10.0%)

C: 6/100 (6.0%)

p>0.05

Olasehinde 2016

Type of study:

quasi randomised

N=67

Setting: single center, outpatient clinic

Country: Nigeria

Source of funding: not specified

Inclusion criteria: inguinal or inguinoscrotal hernia. >18 years

Exclusion criteria:

recurrent, obstructed or strangulated hernias, patients who had bilateral hernias with dissimilar types on each side (inguinal/inguinoscrotal), immunosuppression (diabetes mellitus, malnutrition, patients on long- term steroids, known retroviral positive patients), chronic debilitating illnesses (chronic renal or liver failure, congestive cardiac failure) and urinary bladder outlet obstruction

N total at baseline:

Intervention: 33

Control: 34

Important prognostic factors²:

Age ± SD:

I: 48.1 ± 20.3 years

C: 48.8 ± 17.9 years

Sex:

I: M 90.9 %

C: M 79.4 %

Groups comparable at baseline? Unclear

Participants were separated into two blocks based on the type of hernia (inguinal hernia / inguinoscrotal hernia). This was aimed at achieving a balance in the type of hernias allocated to the two treatment groups.

Inguinal hernia repair with Lichtenstein (mesh) method

Inguinal hernia repair with darning (non-mesh) method

Length of follow-up:

Mean duration of follow-up was 7.5 months

Loss-to-follow-up:

Not reported

Incomplete outcome data:

Not reported

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic groin pain:

I: 1/33 (3.0%)

C: 1/34 (2.9%)

Post-operative pain (4h):

All patients had varying degrees of pain, but was found to be worse in those who had darning repair.

(p = 0.004).

Post-operative pain (10d):

I: 8/33 (24.2%)

C:15/34 (47.1 %)

Post-operative pain (1mo):

I: 0/33 (0%)

C: 3/34 (8.8 %)

Pain was assessed as present or absent and pain scores when present were categorized as ‘none’, ‘mild’,

‘moderate’, or ‘severe’. Zero score on the scale being regarded as ‘no pain’, 1–3 as ‘mild’ pain, 4–6 as ‘moderate’ and 7–10 as ‘severe’.

Recurrences (6mo):

There was no recurrence in any of the groups.

Complications:

Overall:

I:12/33 (36.4%)

C: 13/34 (38.2%)

Surgical site infection:
Overall rate was 4.5%.

I:2/33 (6.1%)

C: 1/34 (2.9%)

P > 0.05

Wound haematoma:

I: 5/33 (15.2%)

C: 9/34 (25.5%)

P>0.05

Scrotal oedema:

I: 6/33 (18.2%)

C:5/34 (14.7 %)

P>0.05

Acute urinary retention (4h):

I: 1/33 (3.0%)

C: 1/34 (2.9%)

Orchitis:

I:0/33 (0%)

C:1/34 (2.9%)

Palermo 2015

Type of study: RCT

N=263

Setting: single center

Country: Brasil

Source of funding: none

Inclusion criteria:

Nonsurgical risk or liable to compensation; primary inguinal hernia unilateral or bilateral; non complicated hernia; patients without coagulation disorder.

Exclusion criteria:

elevated surgical risk; high blood pressure that not respond to treatment; obstructed hernias; strangulated hernias; recurrent hernias; coagulation disorders.

N total at baseline:

Intervention: 135

Control: 128

Important prognostic factors²:

Age ± SD:

I: 55.5 ±16.1

C: 54.4±17.0

Sex:

I: M 88.2%

C: M 82.8%

Groups comparable at baseline? Yes

Lichtenstein (mesh) inguinal hernia repair

Bassini (non-mesh) inguinal hernia repair

Length of follow-up:

6 month

Loss-to-follow-up:

I: N=0 (0%)

C: N=0 (0%)

Incomplete outcome data:

I: N=0 (0%)

C: N=0 (0%)

Outcome measures and effect size (include 95%CI and p-value if available):

Postoperative pain:

(analgesia at day 7)

I: 74/135 =54.8%

C: 112/128 =87.5%

P<0.0001

Analgesia at day 15 and 1 month also reported.

Pain is reported as EVA scale outcome.

Recurrence:

I: 1/135 (0.7%)

C: 3/128 (2.3%)

p=0.36

Seroma (day7):

I: 10/135 =7.4%

C: 8/128 = 6.2%

p=0.71

Seroma at day 15 also reported.

Seroma (day30):

I: 1/135 =0.7%

C: 2/128 = 1.6%

p=0.61

Hematoma:

Amount not reported. No significant difference between groups

Infection:

I: 2/135 =1.5%

C: 9/128 =7.0%

p=0.031

Sample size needed was recalculated and reduced after finding benefit for mesh in an interim analysis.

Szopinski 2012

Type of study: RCT

N=208

Setting: multicenter

Country: Poland

Source of funding: grant from the

Nicolaus Copernicus University of Torun, Collegium Medicum in Bydgoszcz, Poland (grant 19/2006) and the Foundation for Clinical Surgery in Bydgoszcz, Poland.

Inclusion criteria:

adult male Caucasian patients with primary inguinal hernias, assessment of the condition of the external oblique aponeurosis.

Exclusion criteria:

with exclusion of patients with an aponeurosis that was divided, tiny, and/or weak. Patients with recurrent or strangulated hernias or mental disorders, those participating in other clinical trials, ASA> 3, history of a forced hernia reduction with subsequent hospitalization, a history of infection, or the presence of any scar in the inguinal area.

N total at baseline:

Intervention: 103

Control: 105

Important prognostic factors²:

Age ± SD:

I: 54.1 ±15.3

C: 50.2 ±17.5

Sex:

I: M 100%

C: M 100%

Groups comparable at baseline? Yes

Primary inguinal hernia repair with Lichtenstein (mesh) method

Primary inguinal hernia repair with Desarda method

(non-mesh)

Length of follow-up:

at 7, 30 days, and 6, 12, 24, 36 months

Loss-to-follow-up:

Given for 1yr, 2yr and 3yr

I: baseline N=103

assessed at 1yr N=92

assessed at 2yr N=91

assessed at 3yr N=87 (16%)

5 no reason given

3 refused

7 no data

1 death (heart infarct)

C: baseline N=105;

assessed at 1yr N=96

assessed at 2yr N=91

assessed at 3yr N=83 (21%)

6 no reason given

3 refused

12 no data

1 death (heart infarct)

Incomplete outcome data:

I: Analyzed after 3 y n = 87

Excluded from analysis n = 0

C: Analyzed after 3 y n = 83

Excluded from analysis n = 0

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic pain (>6mo after surgery):

I: 3/103 (2.9%)

C: 5/105 (4.8%)

p=0.464

Pain is measured using a visual analog scale (VAS)

Recurrence (>1 year):

I: 2/103 (1.9%)

C: 2/105 (1.9%)

p=1.00

Seroma (7d):

I: 6/103 (5.8%)

C: 4/105 (3.8%)

p=0.508

Seroma (30d):

I: 8/103 (7.8%)

C: 0/105

p=0.004

SSSI (infection):

I: 2/103 (1.9%)

C: 1/105 (0.9%)

p=1.0

Hematoma:

I: 7/ 103 (6.8%)

C: 8/ 105 (7.6%)

Testicular edema (30d):

I: 6/103 (5.8%)

C: 6/105 (5.7%)

Wamalwa 2015

Type of study:

RCT

N=50; 5 did not show up for surgery

Setting: single center

Country: Kenya

Source of funding: not reported

Inclusion criteria:

Primary unilateral inguinal hernia, men, age 18-80,

Exclusion criteria:

irreducibility, femoral hernia, coagulation abnormalities, anticoagulant treatment, and patients for whom anaesthesia was considered unsuitable

N total at baseline:

Intervention: 23

Control: 22

Important prognostic factors²:

Median age (range):

I: 60 (22-80)

C: 63.5 (25-76)

Sex:

I: M 100%

C: M 100%

Types of hernia:

I: direct n=5

indirect n=17

combined n=1

all types n=23

C: direct n=3

indirect n=18

combined n=1

all types n=22

Groups comparable at baseline? yes

Inguinal hernia repair with the tension-free Lichtenstein (mesh) method

Inguinal hernia repair with 4-layer tissue repair Shouldice (no-mesh) method

Length of follow-up:

24hrs, 2 weeks, 3 months and yearly for two years.

Loss-to-follow-up:

I: N=0 (0%)

C: N=0 (0%)

Incomplete outcome data:

I: N=0 (0%)

C: N=0 (0%)

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic pain

“One patient had persistent chronic pain after Lichtenstein repair. This was classified as neuralgia

on account of its radiation along the inguinal nerve.”

Post-operative pain (>3mo):

I: 1/23 (4.3%))

C: 0/22

Also reported after 24hours and after 2 weeks

Not reported how pain was assessed

Recurrences:

None in both groups

Urinary retention:

I: 1/23 (4.3%)

C: 0/22

Major complications:

There were no major complications related to the operation and no reoperations.

Postoperative complications (24hrs):

I: 4/23 (17.4%)

C: 3/22 (13.6%)

Postoperative complications (2wks):

I: 3/23 (13.0%)

C: 1/22 (4.5%)

Youssef 2015

Type of study:

RCT

N=168

Setting:

Single center

Country: Egypt

Source of funding: none

The study was registered at the clinical trials registry of the National Institute of Health (NCT02329938).

Inclusion criteria:

All patient with inguinal or inguinoscrotal hernia

Exclusion criteria:

patients under 18 years , patients with scar in inguinal region, patients with strangulated, recurrent or giant inguinoscrotal hernia, history of forced hernia reduction with subsequent hospitalization, patients with poorly controlled DM, chronic cough, those on TB treatment, severe hypertension, chronic obstructive pulmonary disease, obstructive uropathy, and ASA > 3, major psychological instability and drug abuse, patients found to have thin, weak or divided external oblique aponeurosis intraoperatively

N total at baseline:

Intervention: 83

Control: 85

Important prognostic factors²:

Age ± SD:

I: 44±10

C: 46±11

Sex:

I: M 96%

C: M 97%

Groups comparable at baseline? Yes

The age gender, BMI, and comorbidities were comparable in both groups. No significant difference in the characters of hernia was detected in both groups

Lichtenstein (mesh) hernia repair method for inguinal hernia

Desarda (no mesh) hernia repair method for inguinal hernia

Length of follow-up:

3, 6, 12, and 24 months

Hernia recurrence and postoperative complications were assessed by physical and instrumental examinations

Loss-to-follow-up:

I: 11/83 (13.3%)

C: 14/85 (16.5%)

Reasons not described

Incomplete outcome data:

The analyzed patients were 72 patients in Lichtenstein group and 71 patients in Desarda repair group.

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic pain (6mo):

I: 3/72 (4.2%)

C: 4/71 (5.6%)

P=0.68

pain was classified as moderate pain (VAS 3-5.5).

Post-operative pain at (7d): mean±SD (VAS score)

I: 1.4 ± 1.2

C: 1.5 ± 1.3

P=0.65

Also pain at 3hr, 24hr, 48hr, day 14 reported

Pain was assessed using visual analogue scale (VAS)

Recurrence (2yrs):

I: 1/72 (1.4%)

C: 1/71 (1.4%)

P=0.99

There were no early recurrences (<1 year).

Hematoma (at 14d):

I: 0/72

C: 1/71 (1.4)

SSSI (infection) (7d)

I: 2/72=2.7%

C: 1/71= 1.4%

P=0.55

Wound infection (14d)

I: 0/72

C: 0/71

Seroma (7d):

I: 1/72=1.4%

C: 0/71= 0%

P=0.32

Primary study

Mesh technique

Non-mesh technique

Barth 1998

Lichtenstein

Shouldice

Berndsen 2007

Tapp

Shouldice (modified)

Butters 2007

Tapp and Lichtenstein

Shouldice

Danielsson 1999

Lichtenstein

Shouldice

Hauters 1996

not specified

Shouldice

Hetzer 1999

Lichtenstein

Shouldice

Kux 1994

Lichtenstein

Shouldice

Leibl 1995

Tapp

Shouldice

Lermite 2012

mesh Plug

Shouldice

McGillicuddy 1998

Lichtenstein

Shouldice (slightly modified)

Miedema 2004

Lichtenstein

Shouldice

Nordin 2002

Lichtenstein

Shouldice

Pokorny 2008

Lichtenstein/Tapp/Tep

Bassini/Shouldice

Prieto-Diaz-Chavez 2009

mesh-plug tension-free (MPTF)

Shouldice

Schmitz 1997

Tension-free surgical treatment

Shouldice

Tschudi 2001

Tapp

Shouldice

Wamalwa 2015

Lichtenstein

Shouldice

Zieren 1998

mesh plug/Tapp

Shouldice

Abd El Maksoud 2014

Lichtenstein

Darning - Modified darn

Bhatti 2015

Lichtenstein

Desarda

Chakraborty 2007

not specified

Darning - Abraham's darn

Elsebae 2008

Lichtenstein

Bassini

Kaynak 2007

Lichtenstein

Darning - Moloney darn

Kucuk 2010

Lichtenstein

Darning - Moloney darn

Manyilirah 2012

Lichtenstein

Desarda

Nakagawa 2013

Prolene hernia system

Marcy

Naveen 2014

Lichtenstein

Bassini (modified)

Olasehinde

Lichtenstein

Darning technique

Palermo 2015

Lichtenstein

Bassini

Panda 2012

Lichtenstein

Bassini (modified)

Prior 1998

Lichtenstein

Bassini (modified)

Shi 2010

Tension-free

Bassini

Szopinski 2012

Lichtenstein

Desarda

van Veen 2007

Lichtenstein

According to surgeon's method of choice (applied were Bassini–McVay (in 51% of patients), Bassini (25%), Shouldice (20%), and McVay (4%))

Witkowski 2000

mesh plug

Bassini

Youssef 2015

Lichtenstein

Desarda

Study

First author, year

Appropriate and clearly focused question?¹

Yes/no/unclear

Comprehensive and systematic literature search?²

Yes/no/unclear

Description of included and excluded studies?³

Yes/no/unclear

Description of relevant characteristics of included studies?⁴

Yes/no/unclear

Appropriate adjustment for potential confounders in observational studies?⁵

Yes/no/unclear/notapplicable

Assessment of scientific quality of included studies?⁶

Yes/no/unclear

Enough similarities between studies to make combining them reasonable?⁷

Yes/no/unclear

Potential risk of publication bias taken into account?⁸

Yes/no/unclear

Potential conflicts of interest reported?⁹

Yes/no/unclear

Amato 2012

Yes

Yes

Yes

Yes

Not applicable

Yes

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)

Bhatti 2015

Random table

“Included patients were randomly divided into two groups using random table …”

Unclear

Insufficient information. “200 patients admitted for unilateral, primary, reducible inguinal hernia were evaluated by consultant, and the patients fulfilling the inclusion criteria were included in the study after taking the informed consent”

Likely

No blinding.

Likely

Blinding is not possible.

Unclear

Not specified who were outcome assessors and whether they were blinded.

Unlikely

Study aimed at looking at seroma and duration of operation and did not report key outcomes that would be expected for such a study (i.e. recurrence, wound infection, pain etc.)

Unlikely

No loss-to-follow-up reported.

Unlikely

No study flow presented, however patients seem to be analyzed as randomized.

Olasehinde 2016

“Patients were allocated into two treatment groups (mesh repair, Group A and Darning repair Group B). Participants were separated into two blocks based on the type of hernia.(…) Patients within each

block were alternated between the treatment groups as they

presented.”

Likely

“Patients within each

block were alternated between the treatment groups as they

presented”

Likely

No blinding.

Likely

Blinding is not possible.

Unclear

Pain was self-reported and patients were unlikely to be blinded. Unclear whether staff that performed the physical examination was aware of the operation technique.

Unclear

No protocol available, however, outcomes reported in methods were reported results for.

Unlikely

No participants lost to follow up.

Unlikely

No study flow presented, however patients seem to be analyzed in the group they were randomly allocated to.

Palermo 2015

Randomisation table, sealed envelopes

NB. Sample size needed was recalculated and reduced after finding benefit for mesh in an interim analysis.

Unlikely

“The surgical technique to be performed was contained in sequentially numbered opaque envelopes, using a table of random numbers to produce the series of interventions. The envelopes were in an inviolable dispenser that could only be extracted per unit in a sequentially way, and its extraction was performed before the surgery, when the patient was in the pre-anesthesia room”

Likely

No blinding.

Likely

Blinding is not possible.

Unclear

Pain was self-reported and patients were unlikely to be blinded. Not specified who were other outcome assessors and whether they were blinded.

Unclear

No protocol available, however, results were reported for all outcomes listed in the methods section.

Unlikely

No loss-to-follow-up reported.

Unlikely

All patients were treated as randomized

Szopinski 2012

“Computer-generated

allotments that were disclosed to the surgeon via sealed

envelope”

Unlikely

“The envelope was not opened until after the dissection

and assessment of the external oblique aponeurosis

had been performed because the condition of the aponeurosis

was the final inclusion criterion.”

Unlikely

“They (patients) each agreed to not be informed about

the technique used until 2 years following the date of surgery,”

Likely

Blinding is not possible.

Unlikely

“Both the patients and controlling investigators were blinded to the hernia surgery method used.”

Unlikely

Trial register record available: all outcomes listed were reported results for.

Unlikely

Loss-to-follow-up (including reasons) is reported and is balanced between intervention and control group.

Unlikely

All patients were treated as randomized according to flow chart.

Wamalwa 2015

Cards in sealed opaque envelopes

Unlikely

Cards in sealed opaque envelopes

Likely

No blinding.

Likely

Blinding is not possible.

Unclear

Pain was self-reported and patients were unlikely to be blinded. Outcomes were assessed by the 2 recruiting physicians who were unaware of the allocation.

Unclear

No protocol available, however, results were reported for all outcomes listed in the methods section.

Unlikely

No loss-to-follow-up reported

Unlikely

All patients were treated as randomized.

Youssef 2015

Sealed envelopes

“Randomization was simple and achieved using sealed envelopes”

Unlikely

All patients were subjected to preoperative evaluation including careful history taking, clinical examination and basic laboratory investigation. Patients were randomized into two study groups undergo one of two repairs: Desarda tissue-based repair (group I) or Lichtenstein mesh-based repair (group II). Randomization was simple and achieved using sealed envelopes.

Likely

No blinding.

Likely

Blinding is not possible.

Unclear

Patients were followed for 2 years postoperatively. The first follow-up assessments were performed at 1st and 2nd postoperative weeks by examining the patient in the outpatient clinic and data were collected during interview by an independent observer unaware of the surgical details. However, pain was self-reported and patients were unlikely to be blinded.

Unlikely

The study was registered at the clinical trials registry of the National Institute of Health (NCT02329938). Authors report additional outcomes compared to the registered outcomes, however, all were logical outcomes to report.

Unclear

Loss-to-follow-up is balanced between intervention and control group, but reasons for loss-in follow up are not reported.

Unlikely

All patients were treated as randomized according to flow chart.

Certainty assessment

№ of patients

Effect

Certainty

Importance

№ of studies

Study design

Risk of Bias

Inconsistency

Indirectness

Imprecision

Other considerations

mesh

Shouldice

Relative
(95% CI)

Absolute
(95% CI)

Recurrence (follow up: range 4 weeks to over 5 years; assessed with: various methods)

16

randomised trials

serious ^a

not serious

not serious

serious ^b

none

50/2225 (2.2%)

78/1798 (4.3%)

RR 0.43
(0.24 to 0.78)

25 fewer per 1.000
(from 10 fewer to 33 fewer)

⨁⨁◯◯
LOW

CRITICAL

Chronic pain (follow up: range 3 months to over 5 years; assessed with: various methods)

8

randomised trials

serious ^a

not serious

not serious ^c

serious ^d

none

No differences were found between mesh and Shouldice for any follow-up term, except for outcome measurement at 2 years, for which 2 studies (n=1097) reported significantly less pain for mesh compared to Shouldice.

⨁⨁◯◯
LOW

CRITICAL

Postoperative pain (assessed with: various methods)

14

randomised trials

serious ^a

serious ^e

serious ^f

serious ^b

none

In seven RCTs less postoperative pain was found for mesh compared to Shouldice. In the remaining RCTs no differences were found between mesh and Shouldice.

⨁◯◯◯
VERY LOW

IMPORTANT

Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison / control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Magnusson 2016

(PICO2)

Type of study:

RCT

Setting:

Hospital/free living

Country:

Sweden

Source of funding:

Swedish Research Counsil, The Erling-Persson Family Foundation Sweden. Authors declare that there are no conflicts of interest.

Inclusion criteria:

Male, age 18-75, scheduled for daycare surgery under local anesthesia

Exclusion criteria:

Impaired cognitive function, substance abuse, limited mobility, limited capacity to communicate in Swedish

N total at baseline:

L: 108

PHS: 99

UHS: 102

Important prognostic factors²:

Sex:

All male sample

Number in pain at baseline, n (%):

L: 67/102 (62%)

PHS: 65/99 (66%)

UHS: 65/102 (64%)

Median age (IQR):

L: 60 (49-64)

PHS: 58 (48-63)

UHS: 59 (46-66)

Median BMI (IQR):

L: 24.6 (23-26)

PHS: 24.8 (23.2-26.3)

UHS: 25 (23-26.5)

Heavy workload in occupation, n (%):

L: 28/109 (26%)

PHS: 37/99 (37%)

UHS: 25/102 (25%)

Groups comparable at baseline?

Yes, however larger proportion heavy work in occupation in the PHS group.

Describe intervention (treatment/procedure/test):

Prolene Hernia System (PHS) under local anesthesia

Or,

UltraPro Hernia System (UHS) under local anesthesia

Describe control (treatment/procedure/test):

Lichtenstein with a standard polypropylene mesh under local anesthesia.

Length of follow-up:

12/24/36 months

Loss-to-follow-up:

At 3 years: n=27 (8%)

Incomplete outcome data:

Lichtenstein:

At 3 months: N=8 (8%); At 6 months:

N=13 (13%); At 12 months: N=5 (6%); At 24 months: N=22 (21%); At 36 months:

N=15 (15%)

PHS:

At 3 months: N=12 (12%); At 6 months: N=11 (11%); At 12 months: N=4 (4%); At 24 months: N=12 (12%); At 36 months:

N=7 (8%)

UHS:

At 3 months: N=12 (12%); At 6 months: N=12 (12%); At 12 months: N=7 (7%); At 24 months: N=21 (21%); At 36 months:

N=4 (4%)

No reasons for incomplete data provided.

Outcome measures and effect size (include 95%CI and p-value if available):

Participants in pain at 12/24/36 months, n(%):

L: 15 (15%) / 7 (8%) / 6 (6%)

PHS: 12 (12%) / 12 (14%) / 6 (7%)

UHS: 13 (13%) / 12 (15%) / 9 (9%)

No sign. differences between groups (no p-value reported)

Median pain at 12 months in rest, VAS-score (IQR):

L: 0.2 (0-0.6)

PHS: 0.2 (0-1.9)

UHS: 1 (0.2-2.5)

No significant difference between groups reported, No p-value reported

Median pain at 24 months in rest, VAS-score (IQR):

L: 0.3 (0.1-0.7)

PHS: 0.3 (0.1-0.8)

UHS: 0.8 (0.4-1.3)

No significant difference between groups reported, No p-value reported/

Median pain at 36 months in rest, VAS-score (IQR):

L: 0.4 (0.2-1.7)

PHS: 0.2 (0.1-2.3)

UHS: 1.6 (0.7-4.6)

No significant difference between groups reported, No p-value reported/

Median pain at 12 months in activity, VAS-score (IQR):

L: 0.5 (0.2-3.1)

PHS: 0.5 (0.3-3)

UHS: 2.3 (0.7-2.9)

No significant difference between groups reported, No p-value reported/

Median pain at 24 months in activity, VAS-score (IQR):

L: 0.7 (0.4-1.5)

PHS: 0.9 (0.4-2.1)

UHS: 1.2 (0.4-2.6)

No significant difference between groups reported, No p-value reported/

Median pain at 36 months in activity, VAS-score (IQR):

L: 0.6 (0.2-1.7)

PHS: 0.4 (0.2-2.3)

UHS: 2 (1.4-3)

No significant difference between groups reported, No p-value reported

L= Lichtenstein

PHS= Prolene Hernia System

UHS= UltraPro Hernia System

Sample characteristics from: Magnusson J, Nygren J, Thorell A (2012) Lichtenstein, prolene

hernia system, and UltraPro Hernia System for primary inguinal

hernia repair: one-year outcome of a prospective randomized

controlled trial. Hernia 16(3):277–285

Čadanová 2016

(PICO2)

Type of study:

RCT

Setting:

Hospital/outpatient clinic/free living

Country:

Netherlands

Source of funding:

Not reported in the manuscript.

Inclusion criteria:

ASA status I-III, adults, unilateral primary hernia, signed informed consent

Exclusion criteria:

Incarcerated hernia, recurrent hernia, local inguinal inflammation, ASA≥4, previous inguinal or preperitoneal surgery, inadequacy for follow-up

N total at baseline:

Intervention: 116

Control: 122

Important prognostic factors²:

age ± range:

I: 57.2 (54.5-59.9)

C: 59 (56.5-61.4)

Sex:

I: 110/116 (94.8% M)

C: 114/122 (93.4% M)

Mean BMI (range):

I: 25.1 (24.5-25.7)

C: 25.7 (25.1-26.3)

ASA status:

I: I: 55.2%; II: 38.8%; III: 6%

C: I: 53.3%; II: 41%; III: 5.7%

Groups comparable at baseline?

Yes

Describe intervention (treatment/procedure/test):

Inguinal hernia repair using the ProGrip technique

Describe control (treatment/procedure/test):

Inguinal hernia repair using the transinguinal preperitoneal technique with a PolySoft mesh

Length of follow-up:

14 days/ 3 months / 1 year

Loss-to-follow-up:

N=20 were excluded for analyses: declined surgery, did not show up for first follow-up visit, did not receive the assigned treatment. Not mentioned to what groups the excluded cases were allocated. Follow-up participation rate declined from 99.6% (2 weeks) to 62.6% (1 year). Loss to follow-up of 38.4% over the course of a year.

Incomplete outcome data:

Recurrence at 3mo:

I: 24/116 (20.6%)

C: 26/122 (21.3%)

No reasons given

Recurrence at 1y:

I: 43/116 (37%)

C: 46/122 (37.7%)

No reasons given

Complications unclear.

VAS pain 3mo:

I: 25/116 (21.5%)

C: 28/122 (22.9%)

No reason given

VAS pain 1y:

I: 44/116 (37.9%)

C: 47/122 (38.5%)

No reasons given

LANSS at 3mo:

I: 25/116 (21.5%)

C: 26/122 (21.3%)

No reasons given

LANSS at 1y:

I: 45/116 (38.7%)

C: 46/122 (37.7%)

No reasons given

Outcome measures and effect size (include 95%CI and p-value if available):

Recurrence at 2 weeks:

I: 0/115 (0%)

C: 1/121 (0.8%)

No sign. differences between groups (p=0.331)

Recurrence at 3 months:

I: 1/92 (1.1%)

C: 1/96 (1%)

No sign. differences between groups (p=0.976)

Recurrence at 1 year:

I: 2/73 (2.7%)

C: 1/76 (1.3%)

No sign. differences between groups (p=0.536)

Complications according to patients at 2 weeks, n (%):

I: 57/113 (50.4%)

C: 39/120 (32.5%)

Sign. differences between groups (p=0.005)

Complications according to patients at 3 months, n (%):

I: 19/91 (20.9%)

C: 12/94 (12.8%)

No sign. differences between groups (p=0.140)

Complications according to patients at 1 year, n (%):

I: 10/72 (13.9%)

C: 11/75 (14.7%)

No sign. differences between groups (p=0.893)

Pain in rest at 3 months (>VAS3), n(%):

I: 12/91 (13.2%)

C: 4/94 (4.3%)

Sign. difference (p=0.031)

Pain in rest at 1 year (>VAS3), n(%):

I: 3/72 (4.3%)

C: 5/75 (6.7%)

No sign. difference (p=0.504)

Pain in ADL at 3 months (>VAS3), n(%):

I: 15/91 (16.5%)

C: 4/94 (4.3%)

Sign. difference (p=0.006)

Pain in ADL at 1 year (>VAS3), n(%):

I: 3/72 (4.2%)

C: 6/75 (8%)

No sign. difference (p=0.332)

Pain in sport/heavy lifting at 3 months (>VAS3), n(%):

I: 19/91 (20.9%)

C: 3/94 (3.2%)

Sign. difference (p<0.001)

Pain in sport/heavy lifting at 1 year (>VAS3), n(%):

I: 3/72 (4.2%)

C: 6/75 (8%)

No sign. difference (p=0.332)

Number of participants in pain at rest at 1 year, n (%):

I: no pain (VAS 0): 57/72 (79%); mild pain (VAS 1-3): 13/72 (18.1%), moderate pain (VAS 4-6): 2/72 (2.8%), severe pain (VAS >7): 0 (0%)

C: no pain (VAS 0): 60 (80%); mild pain (VAS 1-3): 12/75 (16%), moderate pain (VAS 4-6): 3/75 (4%), severe pain (VAS >7): 0 (0%)

No sign. differences between groups (p=0.624)

Number of participants in pain in ADL at 1 year, n (%):

I: no pain (VAS 0): 62 (83.3%); mild pain (VAS 1-3): 10/72 (13.9%), moderate pain (VAS 4-6): 2/72 (2.8%), severe pain (VAS >7): 0 (0%)

C: no pain (VAS 0): 60 (80%%); mild pain (VAS 1-3): 13/75 (17.3%), moderate pain (VAS 4-6): 1/75 (1.3%), severe pain (VAS >7): 1 (1.3%)

No sign. differences between groups (p=0.494)

Number of participants in pain in sport/ heavy lifting at 1 year, n (%):

I: no pain (VAS 0): 55/75 (73.3%); mild pain (VAS 1-3): 14/75 (18.6%), moderate pain (VAS 4-6): 3/75 (4%), severe pain (VAS >7): 0 (0%)

C: no pain (VAS 0): 59 (78.7%); mild pain (VAS 1-3): 12/75 (16%), moderate pain (VAS 4-6): 3/75 (4%), severe pain (VAS >7): 1 (1.3%)

No sign. differences between groups (p=0.494)

LANSS neuropathic pain at 3 months, n (%):

I: not neuropathic (<12): 72/91 (79.1%); neuropathic (>12): 19/91 (20.9%)

C: not neuropathic (<12): 94/96 (97.9%); neuropathic (>12): 2/96 (2.1%)

Sign. differences between groups (p<0.001)

VDS pain at 3 months:

I: <moderate pain: 83/91 (91.2%);

≥ moderate pain: 8/91 (8.8%)

C: <moderate pain: 89/94 (94.7%);

≥ moderate pain: 5/94 (5.3%)

No sign. differences between groups (p=0.356)

VDS pain at 1 year:

I: <moderate pain: 68/72 (94.4%);

≥ moderate pain: 4/72 (5.6%)

C: <moderate pain: 71/75 (94.6%);

≥ moderate pain: 4/75 (5.4%)

No sign. differences (p=0.953)

LANSS neuropathic pain at 1 year, n (%):

I: not neuropathic (<12): 60/71 (84.5%); neuropathic (>12): 11/71 (15.5%)

C: not neuropathic (<12): 68/76 (89.5%); neuropathic (>12): 8/76 (10.5%)

No sign. differences between groups (p=0.37)

Karaca 2013

(PICO2)

Type of study:

RCT

Setting:

Hospital/free living

Country:

Turkey

Source of funding:

Not stated in manuscript

Inclusion criteria:

Exclusion criteria:

Systemic disease, Gilbert type 7 and 8 hernia

N total at baseline:

L: 50

RR: 50

PHS: 50

Important prognostic factors²:

age ± SD:

L: 53.06 (13.03)

RR: 51.69 (14.66)

PHS: 48.06 (17.27)

Sex:

L: 86% M

RR: 92% M

PHS: 86% M

Mean BMI (SD):

L: 26.45 (4.94)

RR: 24.22 (2.71)

PHS: 25.06 (3.71)

Groups comparable at baseline?

Differences in age.

Describe intervention (treatment/procedure/test):

Rutkow-Robbins onlay method using a premade Rutkow plug hernia sac.

Or,

Gilbert double layer graft (Prolene Hernia System, PHS)

Describe control (treatment/procedure/test):

Lichtenstein procedure with polypropylene mesh (6x11 cm)

Length of follow-up:

Patients were followed 4-9 years

Loss-to-follow-up:

Not described

Incomplete outcome data:

Late complications, PHS group: n=2 (4%), no reasons given.

Outcome measures and effect size (include 95%CI and p-value if available):

Early complications, n (%)

L: 1/50 (2%)

RR: 5/50 (10%)

PHS: 3/50 (6%)

Sign. difference between groups (p=0.033)

Late complications, n (%):

L: 5/50 (10%)

RR: 4/50 (8%)

PHS: 5/48 (10.4%)

No sign. difference between groups (p=0.924)

1y recurrence

L: 0

RR: 0

PHS: 0

No difference between groups (p-value not reported)

Randomization in 3 groups by throwing heads/tail.

Definition of time in early and late complications is unclear.

Pain scores not extraced because the follow-up for pain was shorter than 3 months (i.e. 1/7/30 days).

L= Lichtenstein

RR= Rutkow-Robbins

PHS = Prolene hernia system

Mahmoudvand 2017

(PICO3)

Type of study: parallel RCT

Setting: Lorestan University of Medical Sciences, Khorramabad, Iran

Country: Iran

Source of funding: not reported

Inclusion criteria: having direct hernia with

defects in the posterior wall, being a candidate for classic

herniorrhaphy, being a candidate for preperitoneal herniorrhaphy,

and satisfaction to enter the study

Exclusion criteria: diabetes, bleeding disorders, and aspirin and corticosteroid

consumption

N total at baseline:

Intervention: 75

Control: 75

Important prognostic factors²:

Sex:

I: 64% M

C: 54.7% M

Groups comparable at baseline? Unclear, only sex was described

Describe intervention (treatment/procedure/test):

Lichtenstein

Patients underwent a

surgical repair in inguinal hernia with classic method (Lichtenstein) under spinal anesthesia.

The surgeon incised the skin and subcutaneous tissue of the lower part of the abdomen and then the fascia of Scarpa and the

roof of the inguinal canal. After reinforcement of the posterior wall of the inguinal canal, the Mersilene mesh (7.5 × 10 cm) was placed and fixed using Round nylon stitch 3/0 to the edges of the defect or weakness in the posterior wall.

Describe control (treatment/procedure/test):

Preperitoneal method

Patients underwent a

surgical repair in inguinal hernia with preperitoneal

method under spinal anesthesia.

The surgeon incised the skin and subcutaneous tissue of the lower part of the abdomen and then the fascia of Scarpa and the

roof of the inguinal canal. after acquiring the posterior wall of the inguinal canal, the Mersilene mesh (7.5 × 10 cm) was placed and fixed using Round nylon stitch 3/0 under the posteriorwall and then was rehabilitated based on modified Bassini repair method.

Length of follow-up: 6-12 months

Loss-to-follow-up:

Intervention: 0

Control: 0

Incomplete outcome data:

Intervention: 0

Control: 0

Outcome measures and effect size (include 95%CI and p-value if available):

Morbidity

Pain after surgery

I: 21

C: 9

P=0.014

Postoperative hematoma

I: 7

C: 9

P=0.597

Postoperative seroma

I: 8

C: 1

P=0.034

Mortality

Not reported.

Recurrence

I: 10

C: 2

P=0.016

Author’s conclusion

It seems that the preperitoneal method is a more suitable method for inguinal herniorrhaphy than the classic one because of fewer complications, according to the findings

of this study. It should be noted that the determination of the type of operation needs a lot of benchmarks, and medical staffs should perform the most appropriate procedure according to all aspects to treat the patients.

Andresen 2015

(PICO3)

Type of study: parallel RCT

Setting: five surgical departments in Denmark. The participating centers had to have one or two

designated surgeons that were experienced with the

Lichtenstein technique ((40 procedures) and had performed

a minimum of 10 Onstep procedures following a

standard training program with hands-on proctoring.

Country: Denmark

Source of funding: not reported

Inclusion criteria: Participants had to be adult, male patients,>18 years, diagnosed with a primary inguinal

hernia that required surgical intervention and with no

contraindications against general anesthesia.

Exclusion criteria:

not being able to understand Danish, requiring

an emergency procedure, having recurrent inguinal hernia, having an ASA score of more than 3, suffering from incarcerated or irreducible hernia, suffering from local (site

of surgery) or systemic infection, having a contralateral

hernia that had to be operated at the same time or planned operated during follow-up (1 year), suffering from other abdominal hernias being operated at the same time or

planned operated during the follow-up period, having had previous surgery that had impaired the sensation in the

groin area, having a body mass index (BMI) more than

40 kg/m2 or lower than 20 kg/m2, having a daily intake of

alcohol larger than five units (1 unit = 12 g pure alcohol),

having a disease that impaired central or peripheral nerve

function, suffering from concurrent malignant disease,

being diagnosed with cognitive functional impairment

(e.g., dementia), suffering from chronic pain that required daily pain medication, or suffering from mental disorder

that required medication.

N total at baseline:

Intervention: 144

Control: 146

Important prognostic factors²:

age ± SD:

I: 53.8 (14.6)

C:54.5 (15.2)

BMI ± SD:

I: 24.8 (2.6)

C: 24.7 (2.9)

Current smoker:

I: 38 (26.4%)

C: 32 (21.9%)

Groups comparable at baseline? Yes

Describe intervention (treatment/procedure/test):

Lichtenstein

Lichtenstein repair of inguinal hernia where a Softmesh (Bard, Davol Inc., Warwick RI) was used. The Lichtenstein repair had to be performed according to the guidelines published by the Danish Hernia Database.

Describe control (treatment/procedure/test):

Onstep

This procedure is an open procedure where access to the inguinal canal is made through a 4-cm horizontal incision lateral to the rectus muscle and cranial to the pubic bone. A Polysoft

mesh (Bard, Davol Inc., Warwick RI) is then inserted with the medial part of the mesh in the preperitoneal space and the lateral part surrounding the spermatic cord and placed between the oblique muscle layers. Three sutures are used

to close the mesh around the cord, but no sutures are placed in the tissue to fixate the mesh.

Length of follow-up: 10 and 30 days

Loss-to-follow-up:

Intervention: 0

Control: 0

Incomplete outcome data:

Intervention: 3 2 (%)

Reasons: 1 recurrence, 1 did not respond to phone call, 1 because of pain

Control: 7 (4.7%)

Reasons: 3 recurrence, 1 did not respond to phone call, 1 withdrew due to personal reasons, 1 did not want to continue due to back pain, 2 did not want to continue in study after conversion to Lichtenstein.

Analysed at 30 days:

I:141

C: 139

Outcome measures and effect size (include 95%CI and p-value if available):

Morbidity

Complications classified according to the Clavien-Dindo classification (I/ II/ IIIa/ IIIb)

I: 5/ 7/ 1/ 1 (out of the 148)

C: 2/ 8/ 0/ 1 (out of the 142)

Pain at rest at 1^st postoperative day, median (range)

I: 18 (0–100)

C: 16 (0–100)

p = 0.778

Pain at rest at day 30

I: 0 (0-4)

C: 0 (0-6)

P=0.38

Mortality

Not reported.

Recurrence within 30 days

I: 1

C: 3

P=0.30

Author’s conclusion

The Onstep technique for inguinal hernia

repair was safe and with comparable results to the Lichtenstein

repair regarding short-term pain and postoperative

complications. Further follow-up of the patients will report

levels of chronic pain and sexual dysfunction.

Andresen 2017

(PICO3)

See Andresen 2015 (above)

See Andresen 2015 (above)

See Andresen 2015 (above)

See Andresen 2015 (above)

Length of follow-up: 6 and 12 months

Loss-to-follow-up:

Intervention: 0

Control: 1 (0.7%)

Reason coordinating center did not receive info.

Incomplete outcome data:

Intervention: 27 (18.8%)

Reasons: recurrence before day 30 (n=1), 6 months questionnaire not received (n=13), excluded few days after surgery due to pain (n=1), recurrence (n=4), 12 months questionnaire not received (n=6), disabling chronic pain at 6 months (n=2)

Control:

26 (17.8%)

Reasons (describe)

withdrew after surgery due to back pain (n=1), withdrew after converted Lichtenstein (n=2), withdrew, personal reasons (n=1), - Recurrence (n=2), 12 months questionnaire not received (n=9)

Analysed at 6 months:

I:129

C: 130

Analysed at 12 months:

I:117

C: 119

Outcome measures and effect size (include 95%CI and p-value if available):

Morbidity

Pain at 6 months at rest, median (IQR)

I: 0 (0-2)

C: 0 (0-2)

P=0.67

Pain at 12 months at rest, median (IQR)

I: 0 (0-1)

C: 0 (0-2)

P=0.10

Mortality

Not reported.

Recurrence within 12 months

I: 5/124

C: 6/125

P=0.78

Author’s conclusion

This study found no difference between the ONSTEP

and the Lichtenstein repairs of primary inguinal hernia

in men regarding chronic pain. Both the number of patients

with impairment of daily function measured by

AAS and comparison of VAS for pain during rest and

movement showed no differences between groups. Two patients from the Lichtenstein group and none in the ONSTEP group were diagnosed with disabling chronic pain.

Azeem 2015

(PICO3)

Type of study: RCT

Setting: Hospital

Country: Pakistan

Source of funding: not reported

Inclusion criteria:

Inguinal hernia

Exclusion criteria:

Recurrent hernia, BMI >40, type IV hernia

N total at baseline:

Intervention: 87

Control: 89

Important prognostic factors²:

age ± SD:

I: 50.01 (13.22)

C: 49.4 (12.06)

Sex (M/F):

I: 85/2

C: 88/1

Mean BMI (SD):

I: 25.01 (2.9)

C: 24.9 (2.02)

ASA classification, n:

I: I: 57, II: 23, III: 7

C: I: 61, II: 19, III: 9

Groups comparable at baseline?

Yes

Describe intervention (treatment/procedure/test):

Lichtenstein’s repair for inguinal hernia

Describe control (treatment/procedure/test):

Modified Kugel repair for inguinal hernia

Length of follow-up:

2 and 4 months (pain), 2 year (recurrence rate)

Loss-to-follow-up:

Intervention:

N=1 (1.1%)

Reason: not provided

Control:

N=1 (1.1%)

Reason: not provided

N=2 who were lost to follow up were excluded from analyses.

Incomplete outcome data:

Not described other than loss-to follow-up.

Outcome measures and effect size (include 95%CI and p-value if available):

Chronic pain at 4 months: not clearly deducible from description or figures in manuscript.

Complications at unclear follow-up, n:

I: 23

C: 5

Significant difference between groups (p<0.005)

Recurrence at 2 years, n

I: 3

C: 0

Significant difference between groups (p<0.05)

Bansal 2016

(PICO4)

Type of study: RCT

Setting: hospital

Country: India

Source of funding: not reported in manuscript

Inclusion criteria:

Age 18-60, primary uncomplicated uni/bilateral inguinal hernia

Exclusion criteria:

Female, previous surgery in inguino-scrotal region, recurrent or complicated inguinal hernia, orchiectomy, co-morbidities: coronary artery disease / uncontrolled hypertension and DM / chronic bronchitis / renal and hepatic failure, unfir for general anaesthesia, uncontrolled coagulopathy, not giving consent

N total at baseline:

Intervention: 80

Control: 80

Important prognostic factors²:

age ± SD:

I: 40.9 (12.3)

C: 40 (12.5)

Sex:

All male sample

BMI, kg/m² (SD):

I: 24.1 (2)

C: 24 (2)

Unilateral hernia, n (%):

I: 58 (72.5%)

C: 46 (57.5%)

Groups comparable at baseline?

Describe intervention (treatment/procedure/test):

Trans-abdominal preperitoneal repair for inguinal hernia (TAPP)

Describe control (treatment/procedure/test):

Totally extraperitoneal repair for inguinal hernia (TEP)

Length of follow-up:

3 and 6 months

Loss-to-follow-up:

11.3% of the participants did not complete the 6-months follow-up and were lost.

Incomplete outcome data:

None reported other than known to low-to-folow-up

Outcome measures and effect size (include 95%CI and p-value if available):

Mean chronic pain at 6 months, VAS (SD):

I: 0.87 (0.6)

C: 0.86 (0.7)

No sign. difference between groups (p=0.86). Measure of dispersion not described.

Salma 2015

(PICO5)

Type of study: RCT

Setting: Hospital

Country: Pakistan

Source of funding: Not stated in manuscript

Inclusion criteria:

Male, age >30,direct hernia, ASA I or II

Exclusion criteria:

Contraindications to pelvic laparoscopy, history of repair with mesh, recurrent hernia, previous pelvic surgery, transvesical prostatectomy, hepatitis B or C

N total at baseline:

Intervention: 30

Control: 30

Important prognostic factors²:

No baseline characteristics per group were presented.

age ± SD:

n=60: 61.46 (7)

Sex:

All male sample

Groups comparable at baseline?

unclear

Describe intervention (treatment/procedure/test):

Lichtenstein’s repair (classic)

Describe control (treatment/procedure/test):

Trans-abdominal preperitoneal repair (TAPP), 3-port technique

Length of follow-up:

Total follow-up is unclear. Pain measured 4 hours postoperatively.

Loss-to-follow-up:

None described

Outcome measures and effect size (include 95%CI and p-value if available):

Pain at 4h PO, VAS (SD):

I: 6.23 (1.87)

C: 4.43 (1.59)

Sign. difference (p=0.005).

Patients received an analgesic (diclofenic sodium, 75mg I/M) immediately after surgery and 6h after surgery. No preoperative or further postoperative analgesics were provided.

Westin 2015

(PICO5)

Type of study: RCT

Setting: Hospital

Country: Sweden

Source of funding: Uppsala-Örebro Regional Research council, Stockholm County council, Swedish Society of Medicine, Olle Engqvist Research Foundation.

Inclusion criteria:

Age 20-80, primary unilateral hernia

Exclusion criteria:

Female, <20 years, >80 years, ASA ≥3, bilateral / scrotal / recurrent hernia, previous surgery in lower abdomen (except appendectomy)

N total at baseline:

Intervention: 195

Control: 194

Important prognostic factors²:

Mean age (range):

I: 53.2 (27-77)

C: 52.9 (23-79)

Sex:

All male sample

BMI, kg/m²:

I: 24.9

C: 26.5

ASA classification, n (%):

I: I: 134 (71.7%), II: 51 (27.3%), III: 2 (1.1%)

C: I: 124 (66%), II: 62 (33%), III: 2 (1.1%)

Anatomy, n (%):

I: indirect: 109 (58.2%), direct 57 (30.5%), femoral: 0, combined: 20 (10.7%), other: 1 (0.5%)

C: indirect: 124 (66%), direct 56 (29.8%), femoral: 1 (0.5%), combined: 7 (3.8%), other: 1 (0.5%)

Groups comparable at baseline?

Yes

Describe intervention (treatment/procedure/test):

Lichtenstein’s procedure for inguinal hernia repair

Describe control (treatment/procedure/test):

Totally extraperitoneal procedure for inguinal hernia repair (TEP)

Length of follow-up:

1 year

Loss-to-follow-up:

Intervention:

N=4

Reasons: no response (n=2), declined further participation (n=2)

Control:

N=5

Reasons: deceased (n=1), no response (n=3), declined further participation (n=1)

Incomplete outcome data:

Intervention:

N=4

Reasons: excluded from analysis, protocol violation before surgery

Control:

N=1

Reasons: excluded from analyses, protocol violation before surgery

Outcome measures and effect size (include 95%CI and p-value if available):

Pain after 1 year (% of participants):

Persistent pain, n (%):

I: 62 (33.2%)

C: 39 (20.7%)

Sign. difference (p=0.007)

IPQ no pain (% estimated from figure)

I: 65%

C: 77%

Sign difference (p=0.007)

IPQ ignorable pain (% estimated from figure)

I: 22%

C: 14%

Sign difference (p=0.011)

IPQ non-ignorable pain (% estimated from figure)

I: 7%

C: 5%

No sign difference (p-value not reported

IPQ pain affecting concentration and activities (% estimated from figure)

I: 2%

C: 1%

No sign difference (p-value not reported

IPQ pain affecting most activities (% estimated from figure)

I: <1%

C: <1%

No sign difference (p-value not reported

Recurrence, n:

I: 4

C: 2

No sign. differences (p-value not reported)

Complications not reported, although this was an outcome measure reported in the methods.

Waris 2016

(PICO5)

Type of study:RCT

Setting: Hospital

Country: Pakistan

Source of funding: Not reported in manuscript

Inclusion criteria:

Inguinal hernia

Exclusion criteria:

Obstructed hernia

N total at baseline:

Intervention: 200

Control: 200

Important prognostic factors²:

Not reported for groups separately

Mean age: 33.89 (SD: 13.45)

Sex (M/F): 304/96

Groups comparable at baseline?

Unclear

Describe intervention (treatment/procedure/test):

Lichtenstein’s procedure

Describe control (treatment/procedure/test):

Totally Extraperitoneal repair (TEP)

Length of follow-up:

Unclear, pain measurement on 1^st and 2^nd day postoperatively

Loss-to-follow-up:

None described

Incomplete outcome data:

None described

Outcome measures and effect size (include 95%CI and p-value if available):

Postoperative pain on day 1 and 2, n(%):

I: 20 (10%)

C: 8 (4%)

Sign difference (p=0.031)

Gürbulak 2015

(PICO5)

Type of study: RCT

Setting: Hospital

Country: Turkey

Source of funding: not reported in the manuscript

Inclusion criteria:

Age 18-70, unilateral hernia Nyhus type 1/2/3a/3b, ASA I/II

Exclusion criteria:

Age <18, age >70, female, femoral / recurrent /bilateral hernia, previous pelvic trauma or surgery, benign prostate hypertrophy, COPD, ASA III or IV, previous incarcerated hernia

N total at baseline:

Intervention: 70

Control: 64

Important prognostic factors²:

age (range):

I: 51.4 (18-70)

C: 48.2 (18-70)

Sex:

All male sample

Groups comparable at baseline?

Unclear

Describe intervention (treatment/procedure/test):

Lichtenstein’s procedure

Describe control (treatment/procedure/test):

Totally extraperitoneal repair (TEP)

Length of follow-up:

Follow-up at 10 days, 3 months and 6 months after discharge (mean hospital stay 1.1 (TEP) and 1.2 (Lichtenstein) days

Loss-to-follow-up:

N=14 lost: did not show up for follow-up after surgery. Excluded from all analyses. Unclear what study arm they were allocated to.

Incomplete outcome data:

None other than reported as loss-to-follow-up

Outcome measures and effect size (include 95%CI and p-value if available):

Intraoperative complications, n:

I: 1

C: 2

No sign. differences (p=0.92)

Postoperative complications, n:

I: 10

C: 3

No sign. difference (p=0.24)

Recurrence, n:

I: 1

C: 1

No sign. difference between groups (p=0.89)

Akgül 2016

(PICO6)

Type of study: RCT

Setting: Hospital

Country: Turkey

Source of funding: sponsored by Akdeniz University (authors declare no financial interest)

Inclusion criteria:

Age 18-65, presented between March 2012 and February 2013

Exclusion criteria:

Bilateral hernia, recurrent hernia, contralateral hernia, comorbidities (heart failure/ rheumatoid arthritis /severe hypertension, systemic or neurologic diseases altering lower extremity function)

N total at baseline:

Intervention: 25

Control:25

Important prognostic factors²:

age (range):

I: 45.1 (22-64)

C: 44.1 (19-64)

Sex:

I: 25 M

C: 22 M / 3 F

Mean operation time (no measure of dispersion):

I: 55 min

C: 70 min

Groups comparable at baseline?

Unclear, only operation time, age and gender are reported.

Describe intervention (treatment/procedure/test):

Stoppa repair for inguinal hernia by using a 6cm Pfannenstiel incision (modifiel Stoppa)

Describe control (treatment/procedure/test):

Totally extraperitoneal repair for inguinal hernia (TEP)

Length of follow-up:

Range: 4-15 months

Loss-to-follow-up:

Not described

Incomplete outcome data:

Not described

Outcome measures and effect size (include 95%CI and p-value if available):

Morbidity (complications), n (%):

I:6/25

C: 3/25

Not reported/tested if this was a significant difference between groups

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)

Magnusson 2016 (PICO2)

Computer generated information for group allocation was produced. When the patient was under local anaesthesia a sealed envelope was opened containing the group allocation.

Unlikely

Reason: Sealed envelopes were used, however there was no description of who performed the randomization procedure (i.e. who generated the list and sealed the envelopes) and that the envelopes were opaque.

Unlikely

Reason: It was stated that the patients did not receive information about which technique was used until after the study.

Likely

Reason: Surgeons performing the hernia repair can not be blinded for the procedure.

Unlikely

Reason: It was stated that a designated nurse collected all data and was blinded for the treatment allocation. However, it is unclear if the nurse could tell what procedure was used if they observed the wound/scar tissue from surgery.

Unlikely

Reason: Outcomes of interest were reported both in the methods and in the results

Unlikely

Reason: Loss-to-follow-up was low (8%), however it was unclear how the loss-to-follow-up was distributed among the groups.

Unlikely

Reason: No ITT was performed, however missing data and the loss-to-follow-up were considered small which was deemed unlikely to introduce bias.

Cadanova 2016 (PICO2)

Patients were randomized by computer generated allocation sequence and stratified by gender, age, BMI, and surgeon.

Unclear

Reason: no information provided on how the allocation was concealed.

Unlikely

Reasons: It was stated that the patients remained unaware of the used procedure, although it is not stated whether the procedure was disclosed at a later time in follow-up.

Likely

Reasons: Surgeons were not blinded for the procedure they conducted.

Unlikely

Reason: Participants were assessed in the outpatient department by a resident who did not have access to the surgery reports. However, it is unclear if the residents could tell what procedure was used if they observed the wound/scar tissue from surgery.

Unlikely

Reasons: Outcomes in interest were reported in the methods and results.

Likely

Reasons: Loss of participation rate (unclear if missing data or loss to follow-up) of 38.4% over the course of 1 year is high.

Likely

Reason: Loss to follow-up and missing data is large while ITT was not followed.

Karaca 2013 (PICO2)

Randomization in three groups was performed by throwing heads or tail.

Likely

Reason: It is not described who performed the randomization. When throwing heads or tail (when done with a coin) the allocation can be predicted while randomizing if there is a consistent method, introducing selection bias.

Unclear

Reasons: No information provided on blinding

Likely

Reasons: Surgeons performing a procedure cannot be blinded.

Unclear

Reason: It is not described who assessed the participants.

Unlikely

Reasons: Outcomes in interest were reported in the methods and results.

Unlikely

Reason: no loss to follow-up described

Unlikely

Reason: No mention of ITT, however no loss-to-follow-up was described and missing data (4%) was seen in only 1 outcome.

Mahmoudvand 2017

(PICO3)

Not described

Unclear

Unclear

Unclear

Unclear

Unclear*

Unlikely

Unlikely

Andresen 2015

(PICO3)

Randomization was done on participant level in blocks

of six. Each center received a unique batch of CRFs to

stratify on center level. Randomization sequence was generated by an online sequence generator

Unlikely

Unlikely

Likely

Unlikely

Unlikely

Unlikely

Likely

Andresen 2017

(PICO3)

Randomization was done on participant level in blocks

of six. Each center received a unique batch of CRFs to

stratify on center level. Randomization sequence was generated by an online sequence generator

Unlikely

Unlikely

Likely

Unlikely

Unlikely

Unlikely

Likely

Azeem 2017

(PICO3)

Prospectively randomized; unclear randomization procedure

Unclear

Reason: no allocation concealment procedure reported.

Likely

Reason: blinding is not mentioned in the manuscript. It is likely that patients cannot be blinded for the surgical procedure

Likely

Reason: blinding is not mentioned in the manuscript. Surgeons cannot be blinded for the surgical procedure

Likely

Reason: blinding is not mentioned in the manuscript. It is likely that the surgeons were the outcome assessors and were not blinded for the surgical procedure

Unlikely:

Reason: No mention of a trial protocol, however outcomes of interest were mentioned in methods and reported in the results. Complications were reported in the results only.

Unlikely

Reason: loss of n=2 (1 in each study arm)

Unlikely

Reason: Although ITT was not followed, no indication of protocol deviations and loss-to-follow-up was n=2.

Bansal 2017

(PICO4)

Block randomisation was done using computer generated random numbers (size of the blocks were not reported). Allocation was concealed by using sealed opaque envelopes.

Unlikely

Reason: sealed opaque envelopes were used

Likely

Reason: although both procedures are endoscopic, it is likely that the patients were aware which procedure was/would be used.

Likely

Reason: no mention of blinding in the manuscript. It is likely the surgeon could not be blinded for the procedure.

Likely

Reason: no mention of blinding in the manuscript. It is likely that the surgeon and/or healthcare team were outcome assessors. The surgeon could not be blinded for the procedure.

Unlikely

Reason: No mention of a trial protocol, however outcomes of interest were mentioned in methods and reported in the results.

Unclear

Reason: although 11.3% loss is relatively low, it was not described in which arm these participants dropped out or whether the drop-out was the same among the both arms.

Unlikely

Reason: stated that 160 (80/80) participants were analysed

Salma 2015

(PICO5)

Simple computer generated randomization

Unclear

Reason: no mention of allocation concealment or its procedure.

Likely

Reason: Authors state that blinding was not possible

Likely

Reason: Authors state that blinding was not possible

Likely

Reason: Authors state that blinding was not possible

Unlikely

Reason: No mention of a trial protocol, however outcomes of interest were mentioned in methods and reported in the results.

Unlikely (pain post-op)

Reason: stated that all 60 patients were analysed for pain (4h after operation).

Unlikely (pain post-op)

Reason: stated that all 60 patients were analysed for pain (4h after operation).

Westin 2016

(PICO5)

Block randomization (size 20) in a 1:1 ratio. Consecutively numbered sealed envelopes were used. Method of sequence generation is unclear.

Likely

Reason: Reported that allocation was not concealed.

Likely

Reason: No mention of blinding in the manuscript. Patients were likely to know which procedure was/would be conducted

Likely

Reason: Surgeons are aware of the procedure they are conducting

Likely

Reasons: Outcome assessors were most likely the surgeons (at 1y follow-up).

Unclear

Reason: postoperative complications are mentioned in the methods as a secondary outcome measure, but were not reported in the results. (could be described in the short-term paper)

Unlikely

Reason: loss to follow up is small and similar in both groups

Unlikely

Reason: although ITT was not followed, protocol deviations were excluded from analyses. This accounted for a small number participants (n=4 in intervention group, n=1 in control group)

Waris 2016

(PICO5)

Randomly allocated in two groups after matching for confounders

Unclear

Reason: allocation concealment procedure not described

Likely

Reason: No mention of blinding in the manuscript. Patients were likely to know which procedure was/would be conducted

Likely

Reason: Surgeons are aware of the procedure they are conducting

Likely

Reasons: Outcome assessors were most likely the surgeons

Likely

Reason: no predefined outcomes were given in the methods. No mention of a research protocol.

Unlikely

Reason: no loss to follow-up reported

Unlikely

Reason: no protocol deviations and loss-to-follow-up reported.

Gürbulak 2015

(PICO5)

Randomly divided in groups using a random number table.

Unclear

Reason: allocation concealment procedure not described

Likely

Reason: No mention of blinding in the manuscript. Patients were likely to know which procedure was/would be conducted

Likely

Reason: Surgeons are aware of the procedure they are conducting

Likely

Reasons: Outcome assessors were most likely the surgeons

Unlikely

Reason: No mention of a trial protocol, however outcomes of interest were mentioned in methods and reported in the results.

Unclear

Reason: although n=14 is relatively small, the drop-out distribution among groups was not described.

Unclear

Reason: ITT was not followed and it was not described to which groups the n=14 were allocated. No conversions from laparoscopic to open procedure were made.

Akgül 2016 (PICO6)

Randomization was performed at the presentation of the inguinal hernia. Patients were randomized 1:1 by the primary researcher using a box with opaque sealed envelopes.

Unlikely

Reason: Opaque sealed envelopes were used

Likely

Reason: surgery was either Stoppa or laparoscopic.

Likely

Reason: The primary researcher was the operating surgeon (which also performed the randomization). It was stated that the entire health care team was informed about the procedure.

Likely

Reason: it was stated that the patients were followed up by the primary researcher / operating surgeon to assess complications and recurrence

Unlikely

Reason: variables of interest were stated in the protocol and reported in the manuscript.

Unclear

Reason: there was a variable follow-up time (4-12 month). No information on loss-to-follow-up was provided.

Unlikely

Reason: Only one potential protocol deviation, which is unlikely to bias the results. When describing the complications, probably all participants were considered.

Study reference

Study characteristics

Patient characteristics

Intervention (I)

Comparison / control (C)

Follow-up

Outcome measures and effect size

Comments

Wei 2015

Study characteristics and results are extracted from the SR (unless stated otherwise)

(PICO4)

SR and meta-analysis of RCTs

Literature search up to March 2014

A: Schrenk 1996

B: Dedemadi 2006

C: Butler 2007

D: Pokomy 2008

E: Zhu 2009

F: Hamza 2010

G: Gong 2011

H: Krishna 2012

I: Wang 2013

J: Bansal 2013

Study design: All RCTs

Setting and Country:

A: Austria

B: Greece

C: USA

D: Austria

E: China

F: Egypt

G: China

H: India

I: China

J: India

Source of funding:

Not reported

Inclusion criteria SR: uni- or bilateral inguinal hernia, age<71 years, TAPP and TEP performed, reported outcomes: recurrence / pain / complications / operation time / time to return to activities / hospital stay

Exclusion criteria SR: patients with gastric surgery / previous chemo/radiationtherapy, DVT, cardiovascular / respiratory / hepatic / renal disease

10 studies included

Important patient characteristics at baseline:

N, mean age

Age not reported

A: 28/24

B: 24/26

C: 22/22

D: 93/36

E: 20/20

F: 25/25

G: 50/50

H: 47/53

I: 84/84

J: 154/160

Sex:

Not reported

Hernia type:

A: unilateral indirect, direct, recurrent

B: uni- and bilateral recurrent

C: unilateral

D: unilateral

E: unilateral nonrecurrent

F: unclear indirect

G: uni- and bilateral indirect, direct, mixed

H: uni- and bilateral, indirect, direct

I: uni- and bilateral, indirect, direct, femoral

J: uni- and bilateral nonrecurrent

Groups comparable at baseline?

Heterogeneity in type of inguinal hernias between studies

Describe intervention:

Transabdominal preperitoneal repair (TAPP)

Describe control:

Totally extraperitoneal repair (TEP)

End-point of follow-up:

A: Unclear

B: Mean: 1087 days ±588

C: Range: 7-14 days

D: Range: 2-3 weeks; 3 months; Range: 1-3 years

E: Unclear

F: Range: 2, 12, 24 weeks

G: Mean: 15.6 months ±9.9 / 15.6 months ±7.7

H: Range: 1, 6 weeks; 3, 6, 12, 18, 24, 38 months

I: Mean: 16 months, range: 3-32 months

J: Mean: 30.1 months ±14.3 / 30.9 months ±13.2

For how many participants were no complete outcome data available?

All studies were reported to have more than 80% follow-up data

Postoperative pain

Pain at 1-h PO:

Effect measure: Standardized mean difference (95% CI):

H: 1.95 (1.47-2.43)

J: 0.14 (-0.09-0.36)

Pooled effect (random effects model):

1.03 (95% CI -0.75 to 2.81) favoring TEP

Heterogeneity (I²): 97.8%

Pain at 6-h PO:

Effect measure: Standardized mean difference (95% CI):

H: -1.36 (-1.79- -0.92)

J: 0.18 (-0.04-0.40)

Pooled effect (random effects model):

-0.58 (95% CI -2.08 to 0.92) favoring TAPP

Heterogeneity (I²): 97.4%

Pain at 1-d PO:

Effect measure: Standardized mean difference (95% CI):

F: 0.50 (-0.06-1.07)

G: -0.14 (-0.53-0.25)

H: 2.02 (1.54-2.50)

J: 0.18 (-0.04-1.45)

Pooled effect (random effects model):

0.63 (95% CI -0.20 to 1.45) favoring TEP

Heterogeneity (I²): 94.5%

Pain at 1-w PO:

Effect measure: Standardized mean difference (95% CI):

G: 0.00 (-0.39-0.39)

H: 1.22 (0.80-1.65)

Pooled effect (random effects model):

0.61 (95% CI -0.59 to1.81) favoring TEP

Heterogeneity (I²): 94.2%

Pain at unclear follow-up:

Effect measure: Standardized mean difference (95% CI):

E: 0.21 (-41, 0,83)

I: 1.15 (0.82-1.48)

Pooled effect (random effects model):

0.72 (95% CI -0.21 to 1.48) favoring TEP

Heterogeneity (I²): 85.6%

Recurrence

No included study for pooling the recurrence had a mean follow- up of >3 years.

Time to return to activities

Effect measure: Standardized mean difference (95% CI):

B: 0.12 (-44-0.67)

F: 0.20 (-0.36-0.75)

G: 0.00 (-0.39-0.39)

J: -0.29 (-0.51- -0.07)

Pooled effect (random effects model):

-0.08 (95% CI -0.33 to 0.16) favoring TAPP

Heterogeneity (I²): 33.8%

Costs:

G, H, I, J compared the costs of TAPP vs TEP, however found no significant differences (no p-values reported).

Facultative:

Brief description of author’s conclusion

Personal remarks on study quality, conclusions, and other issues (potentially) relevant to the research question

Level of evidence: GRADE (per comparison and outcome measure) including reasons for down/upgrading

Sensitivity analyses (excluding small studies; excluding studies with short follow-up; excluding low quality studies; relevant subgroup-analyses); mention only analyses which are of potential importance to the research question

Heterogeneity: clinical and statistical heterogeneity; explained versus unexplained (subgroupanalysis)

Sajid 2015

Study characteristics and results are extracted from the SR (unless stated otherwise)

(PICO6)

SR and meta-analysis of RCTs

Literature search up to June 2014

A: Aitola 1998

B: Bostanci 1998

C: Günal 2007

D: Hamza 2010

E: Johanson 1999

F: Kawji 1999

G: Ozmen 2004

H: Simmermacher 2000

I: Sinha 2006

J: Vatansev 2002

Study design: All RCTs

Setting and Country:

A: Finland

B: Turkey

C: Turkey

D: Egypt

E: Sweden

F: Austria

G: Turkey

H: Netherlands

I: India

J: Turkey

Source of funding:

No funding

Inclusion criteria SR: Placement of a mesh in the preperitoneal space through trans-inguinal incision or any other open approach

Exclusion criteria SR: Not reported

10 studies included

Important patient characteristics at baseline:

N (n male), mean age

Age not reported

A: 49 (48), 55/53

B: 64 (63), 31/25

C: 118 (118), 23/25/22

D: 75 (75), 35/36/34

E: 400 (400), 56/55

F: 56 (unclear), 65/48

G: 80 (75), 41/51

H: 162 (unclear), 54

I: 241 (241), unclear

J: 41 (36), 50.7/54.6

Hernia type:

A: bilateral, primary, recurrent

B: primary

C: primary

D: primary

E: primary, recurrent

F: primary

G: primary, recurrent

H: primary

I: recurrent, unilateral, bilateral

J: primary, femoral

Groups comparable at baseline?

Heterogeneity in type of inguinal hernias between studies

Describe intervention:

Open preperitoneal hernia repair

A: Trans-inguinal preperitoneal + Marlex mesh + suture mesh fixation

B: TIPP + Prolene mesh + suture mesh fixation

C: Nyhus + Prolene mesh + fixation not reported

D: TIPP + mesh and fixation not reported

E: TIPP + Prolene mesh +suture fixation

F: Wantz TIPP + mesh and fixation not reported

G: Nyhus + prolene mesh + stitch fixation

H: Ugahary (gridiron) + Prolene mesh + no fixation

I: TIPP + Prolene mesh + no fixation

J: Nyhus + mesh and fixation not reported

Describe control:

Laparoscopic preperitoneal hernia repair

A: TAPP + Marlex mesh + staple mesh fixation

B: TEP + Prolene esh + no fixation reported

C: TAPP and TEP + Prolene mesh + staple fixation

D: TAPP and TEP + mesh and fixation not reported

E: 3-port TAPP + Prolene mesh + staple fixation

F: TAPP + mesh and fixation not reported

G: TEP + Prolene mesh + tacker fixation

H: TEP + Prolene mesh + no fixation

I: TEP + Prolene mesh + tacker fixation

J: TEP + mesh and fixation not reported

End-point of follow-up:

A: 18 months

B: 24 months

C: 96 months

D: 12 months

E: 12 months

F: 18 months

G: NR

H: NR

I: 12 months

J: 1 week

For how many participants were no complete outcome data available?

Not reported

Recurrence

Effect measure: Odds ratio (95% CI):

A: 0.61 (0.09-4.01)

C: 2.05 (0.12-33.71)

D: 0.38 (0.02-8.23)

E: 2.85 (0.89-9.11)

I: 0.49 (0.09-2.71)

Pooled effect (random effects model):

1.21 (95% CI 0.52 to 2.83) favoring laparoscopic

Heterogeneity (I²): 8%

N=1003 (Events in I: 16, events in C: 14)

Morbidity (complications)

Effect measure: Odds ratio (95% CI):

A: 0.33 (0.06-1.90)

B: 2.14 (0.36-12.63)

C: 2.07 (0.76-5.61)

D: 2.09 (0.28-15.76)

E: 0.66 (0.43-1.02)

F: 5.67 (0.55-58.48)

G: 1.79 (0.61-5.22)

H: 0.56 (0.22-1.43)

I: 0.39 (0.07-2.03)

Pooled effect (random effects model):

1.00 (95% CI 0.60 to 2.03) favoring none

Heterogeneity (I²): 40%

N=1245 (Events in I: 90, events in C: 111)

Facultative:

Brief description of author’s conclusion

Personal remarks on study quality, conclusions, and other issues (potentially) relevant to the research question

Level of evidence: GRADE (per comparison and outcome measure) including reasons for down/upgrading

Sensitivity analyses (excluding small studies; excluding studies with short follow-up; excluding low quality studies; relevant subgroup-analyses); mention only analyses which are of potential importance to the research question

Heterogeneity: clinical and statistical heterogeneity; explained versus unexplained (subgroupanalysis)

Study

First author, year

Appropriate and clearly focused question?¹

Yes/no/unclear

Comprehensive and systematic literature search?²

Yes/no/unclear

Description of included and excluded studies?³

Yes/no/unclear

Description of relevant characteristics of included studies?⁴

Yes/no/unclear

Appropriate adjustment for potential confounders in observational studies?⁵

Yes/no/unclear/notapplicable

Assessment of scientific quality of included studies?⁶

Yes/no/unclear

Enough similarities between studies to make combining them reasonable?⁷

Yes/no/unclear

Potential risk of publication bias taken into account?⁸

Yes/no/unclear

Potential conflicts of interest reported?⁹

Yes/no/unclear

Wei 2015 (PICO4)

Yes, however outcomes could be specified more precisely

Yes, however no complete search strategy is provided. Multiple sources were searched.Date of last search was stated.

Yes, not referenced but exact number and reasons were provided.

No, studies were minimally described and no sample description was included or which group received TEP or TAPP.

NA

No, there is a “quality assessment table” includeed but the authors seem to indicate only whether information on the criteria was reported and not whether there was a (un)likely Risk of Bias.

Unclear, because no information was provided on the sample characteristics. Heterogeneity (I²) is large in most analyses.

Yes, Egger’s test was performed and reported for recurrence, operation time, and total complication.However, no Funnel plots were presented.

No, no mention of conflicts of interest for the manuscript or its included studies.

No, Egger’s test and funnelplots were not presented for pain and time to return to activities.

Sajid 2015 (PICO6)

Yes, however complications is not named as an outcome in the research aim.

Yes, however no complete search strategy is provided. Multiple sources were searched.Date of last search was stated.Unclear how many hits from each database. Number of hits is limited.

Yes, not referenced but exact number and reasons were provided (however no exclusion criteria were given in the text)

Yes, minimally described

NA

Yes, a quality table (Jadad) and GRADE table are presented.

Unclear, minimal information on the sample characteristics were given and in both groups several different techniques were combined for meta-analysis (for laparoscopic some subgroup analyses were performed, but not fort he open technique group)

No, publication bias was not considered

Unclear, the authors declase that they have no conflict of interest, however this was not checked for the included studies.