t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4

Mesh fixation technique in totally extraperitoneal

inguinal hernia repair e A network meta-analysis

Suphakarn Techapongsatorn a,b, Amarit Tansawet b,

Wisit Kasetsermwiriya b, Mark McEvoy c, John Attia d,
Chumpon Wilasrusmee e, Ammarin Thakkinstian a,*
a
Section for Clinical Epidemiology and Biostatistics, Faculty of Medicine Ramathibodi Hospital, Mahidol University,
Bangkok, Thailand
b
Department of Surgery, Faculty of Medicine Vajira Hospital, Navamindrahiraj University, Bangkok, Thailand
c
School of Medicine and Public Health, University of Newcastle, New South Wales, Australia
d
School of Medicine and Public Health, University of Newcastle; Hunter Medical Research Institute, New South
Wales, Australia
e
Department of Surgery, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand

article info abstract

Article history: Laparoscopic totally extra-peritoneal inguinal hernia repair is the standard option for
Received 20 May 2018 inguinal hernia treatment. However, there are various types of mesh fixation and their
Received in revised form relative uses are still controversial. This network meta-analysis was conducted to
1 August 2018 compare and rank the different fixations available for TEP. Medline and Scopus data-
Accepted 27 September 2018 bases were search until February 1, 2017 and using randomized controlled trials
Available online 27 October 2018 comparing outcomes between different mesh fixation techniques were included. The
results demonstrated that fifteen RCTs (n ¼ 1783) were eligible for pooling. Five types of
Keywords: mesh fixation were used; metallic tack, no-fixation, absorbable tack, suture, and glue.
Inguinal hernia Network meta-analysis that use metallic tack as the reference, indicated that suture and
Mesh fixations glue both carried a lower risk of recurrence with pooled risk ratios (RR) of 0.29 (95% CI
Network meta-analysis 0.00, 18.81) and 0.29 (0.07, 1.30), respectively. For overall complications, absorbable tack
had lower risk (0.63, 95% CI: 0.02, 16.13). However, none of these estimates reached
statistical significance. So, this network meta-analysis suggests that glue and absorbable
tack might be best in lowering recurrence risk and complications. However, a large scale
RCT is still needed to confirm these results.
© 2018 The Authors. Published by Elsevier Ltd on behalf of Royal College of Surgeons of
Edinburgh (Scottish charity number SC005317) and Royal College of Surgeons in Ireland.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.
org/licenses/by-nc-nd/4.0/).

various forms, such as, groin pain, mass, or even complicated

Introduction cases (e.g., intestinal obstruction and strangulation), which
can affect a patient's quality of life, morbidity, and mortality.
Inguinal hernia is a common problem in general surgery with Operative repair is the standard of care for symptomatic
a prevalence of 1.7% for all ages and 4% for those aged over 45 patients, in which mesh is used to reinforce the inguinal floor
years.1 It is more common in men than women, with lifetime as European Hernia Society recommendations and Interna-
risks of 27% and 3%, respectively.2 It can clinically present in tional guideline for groin hernia managements.3e5

* Corresponding author. Section for Clinical Epidemiology and Biostatistics, Faculty of Medicine Ramathibodi Hospital, 270 Rama VI
Road, Rachatevi, Bangkok 10400, Thailand. Fax. þ66 2 201 1284.
E-mail addresses: suphakarn@nmu.ac.th (S. Techapongsatorn), godjiru@hotmail.com (A. Tansawet), wisit@vajira.edu.ac.th,
numnim168@gmail.com (W. Kasetsermwiriya), Mark.McEvoy@newcastle.edu.au (M. McEvoy), john.attia@newcastle.edu.au (J. Attia),
chumpon.wil@mahidol.ac.th (C. Wilasrusmee), ammarin.tha@mahidol.ac.th (A. Thakkinstian).
https://doi.org/10.1016/j.surge.2018.09.002
1479-666X/© 2018 The Authors. Published by Elsevier Ltd on behalf of Royal College of Surgeons of Edinburgh (Scottish charity number SC005317) and
Royal College of Surgeons in Ireland. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
216 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4

Laparoscopic inguinal hernia repair has become increasingly and selected studies; any disagreements were resolved by
widespread since the 1990s. The hernia recurrence rate is not consensus.
much different compared to open repair, albeit laparoscopic Inclusion criteria were as follows: randomized controlled
repair offers less pain and numbness, and has a faster re- trials (RCT) in adults with inguinal hernia who underwent
covery period.6 Totally extra-peritoneal repair (TEP) and trans- laparoscopic repair with TEP; compared at least one pair of
abdominal pre-peritoneal repair (TAPP) are two common mesh fixation techniques as described below; and had at least
laparoscopic approaches and they differ in the way they ac- one of the outcomes of interest as described below. The
cess the pre-peritoneal space. Although the two techniques studies that reported open or TAPP repairs, case reports, ani-
are similar in efficacy and complications, TEP may be more mal studies, meta-analyses, and non-English language reports
technically difficult than TAPP in the first step of reaching the were excluded.
abdominal wall layer but has the advantage that it does not
require additional intervention for peritoneal closure.7e9 As a Interventions
result, this study focused on the TEP technique.
The current practice for TEP varies from using mesh with Our interventions of interest were mesh fixation techniques in
no-fixation through to fixation with tack (e.g., metallic tack, TEP including tack (i.e., metallic and absorbable tack), no-
absorbable tack), suture, glue, and using self-gripping. Previ- fixation, suture, glue, and self-gripping mesh.
ous meta-analyses have not been optimal in helping to choose
between these options for a number of reasons: One meta- Outcome of interest
analysis10 directly compared different techniques among
TAPP not TEP; three direct meta-analysis11e13 performed The outcomes of interest were as follows: hernia recurrence,
pairwise comparisons of fixations with glue in TEP; four meta- quality of life, post-operative pain score, chronic groin pain,
analyses14e17 mixed TEP and TAPP inguinal repairs; and two operative time, return to work, return to daily life, and length
meta-analysis11,16 directly compared fixation versus no- of hospital stay. These outcomes were defined according to
fixation. Results to date indicate that risk of recurrence, the original studies. In addition, adverse events were consid-
complication and post-operative pain were not different be- ered as a composite event because individual adverse events
tween tack fixation and no-fixation or glue. However, evi- (seroma, hematoma, SSI, urinary retention, or intra-operative
dence for suture and self-gripping mesh is still lacking. In injury) were rare.
addition, none of these studies indicated which type of fixa-
tion is the best for TEP. This network meta-analysis (NMA) Risk of bias assessment
was therefore conducted to compare the efficacy and com-
plications of mesh fixation techniques in TEP and rank them. The methodological quality of the included studies was
assessed by ST and AT1 as recommended by the Cochrane
Collaboration using the “Risk of Bias Assessment Tool”.19 This
Material and methods included random sequence generation, allocation conceal-
ment, blinding participants and personnel, blinding outcome
A systematic review and network meta-analysis was con- assessment, incomplete outcome data, and selective outcome
ducted following the PRISMA extension for network meta- reporting. Disagreement was resolved by a third reviewer
analysis.18 The review protocol has been registered at PROS- (AT2).
PERO (number CRD42016044018).
Data extraction
Search strategies and study selection
Data extraction was performed by ST and AT1 independently.
Medline (via PubMed) and Scopus databases were used to Characteristics of studies and patients were extracted
identify relevant studies through to February 1, 2018. The including type of presentation (primary or recurrent hernia),
search terms were constructed based on patients, in- mean age, percent male, body mass index (BMI), underlying
terventions/comparators, and outcomes as follows: disease, hernia type, inguinal hernia classification, size of
hernia, and mean follow-up time. Cross-tabulated data be-
 patients: inguinal hernia and laparoscopy; tween interventions and outcome were extracted. For com-
 intervention/comparator: tack, clip, staple, suture, glue, plications, the highest number of these events (i.e., seroma,
cyanoacrylate, no-fixation, non-fixation, self-gripping, and hematoma, SSI, urinary retention, or intra-operative injury)
Progrip™; was used for representing a composite event. Disagreement
 outcome: hernia recurrence, quality of life, Carolina com- was resolved by a third reviewer (AT2).
fort scale, post-operative complications (e.g., seroma, he-
matoma, surgical site infection (SSI), urinary retention, or Statistical analysis
intra-operative injury), return to work, return to daily life,
length of hospital stay, operative time, chronic groin pain, Direct meta-analysis (DMA)
post-operative pain. The efficacy of each pair of mesh fixations was directly
compared and pooled for each outcome of interest if there
Search-term details and results are described in Appendix were at least two studies for each comparison. The risk ratio
1. Three reviewers (ST, AT1 and WK) independently searched (RR) was estimated and pooled across studies using a fixed-
 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4 217

effect model if heterogeneity was not present (p-value < 0.10 The probability of being the best mesh fixation technique
or I2  25%), otherwise a random-effect model was applied. was estimated using surface, under the cumulative ranking
Heterogeneity was assessed using Cochrane Q test and I2 curve (SUCRA) method. Finally, the inconsistency assumption
statistic. Source of heterogeneity was explored by fitting co- (i.e., difference between direct and indirect estimates) was
variables (i.e., age, sex, BMI, hernia size, mean follow-up, checked using a design-treatment interaction model. A pre-
underlying disease) one-by-one in a meta-regression model, dictive interval was estimated, taking into account uncer-
subgroup analysis was performed accordingly. Publication tainty from heterogeneity and inconsistency of the network.
bias was assessed using funnel plots and Egger test. If one of All analyses were performed using STATA software version
these showed asymmetry, a contour enhanced funnel plot 14.0; p-values <0.05 were considered statistically significant,
was constructed to identify the cause of asymmetry. except for heterogeneity where p < 0.10 was used. All STATA
commands are provided in Appendix 2.
Network meta-analysis (NMA)
A NMA was applied using a two-stage meta-analysis
approach. Interventions were respectively coded as 1, 2, 3, 4, 5, Results
and 6 for metallic tack, no-fixation, absorbable tack, suture
fixation, glue, and self-gripping mesh, with metallic fixation Characteristics of selected studies
as the reference group. A network map was constructed,
which consisted of nodes and edges where nodes were A total of 2254 studies were identified from Scopus and Med-
weighted by the numbers of studies. A contribution plot was line databases. Among them, 15 RCTs (n ¼ 1783), published
also constructed to show the number of included studies between 1999 and February 2018, were eligible for inclusion;
pooled in direct and indirect comparisons. reasons for ineligibility are described in Fig. 1. The charac-
Relative treatment effects (i.e., coefficients or ln(RRs)) and teristics of studies are described in Table 1. These studies were
variance-covariance were estimated for each study. A multi- from AsiaeAustralia (N ¼ 9), Europe (N ¼ 3), USA (N ¼ 2), and
variate meta-analysis with a consistency model was applied to South America (N ¼ 1). Among them, 7, 1, and 7 studies
pool ln(RRs) across studies. The multiple-treatment contrasts considered unilateral, bilateral, and both unilateral and
were then estimated among all possible fixation techniques. bilateral inguinal hernias, respectively. Mean age and percent

Fig. 1 e Flow diagram of included studies.

 218
Table 1 e Describe characteristics of RCTs comparing the mesh fixation techniques and outcomes in TEP.
Authors (Year) Country Study Inguinal Hernia Outcome Type of mesh No. of No. of % Male Age BMI
type type fixation subjects hernia (mean ± SD)
Ferzli GS. (1999) [20] USA RCT Unilateral recur, optime, work, daily seoma, injur Metallic tack 43 50 100 55 NR
Bilateral No fixation 49 50 100 53 NR
Moreno-Egea A. Spain RCT Unilateral recur, optime, pain, cgp, injur, ssi Metallic tack 85 118 91.8 53.8 þ 15.6 NR
(2004) [21] Bilateral No fixation 85 111 92.9 56.9 þ 16.3 NR
Parshad R. (2005) [28] Hong Kong RCT Unilateral recur, pain, daily, cgp, seroma, ssi Metallic tack 25 34 NA 46.4 þ 15.19 NR
Bilateral Glue fixation 25 29 NA 47.16 þ 16.4 NR
Lau H. (2005) [29] India RCT Bilateral recur, optime, pain, hos, daily, cgp, seroma Metallic tack 47 57 100 65.75 þ 6.06 NR

t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4
Glue fixation 46 61 97.82 63.78 þ 4.48 NR
Koch CA. (2006) [22] USA RCT Unilateral recur, optime, pain, hos, injur, reten Metallic tack 20 26 100 56.3 þ 11.5 27 ± 3.6
No fixation 20 27 100 54.6 þ 16.1 27.2 ± 3.1
Taylor C. (2007) [23] Australia RCT Unilateral recur, optime, cgp Metallic tack 180 250 NA 56.9 þ 21.12 NR
Bilateral No fixation 180 250 NA 57.05 þ 21.12 NR
Garg P. (2011) [25] India RCT Unilateral recur, optime, pain, hos, daily, seroma, reten Metallic tack 52 98 98.1 47.2 þ 12.9 NR
Bilateral No fixation 52 96 94.2 51.9 þ 16.8 NR
Subwongcharoen S. Thailand RCT Unilateral recur, optime, pain, cgp, seroma Metallic tack 30 30 96.67 48.27 þ 17.33 NR
(2013) [24] Glue fixation 30 30 96.67 52.4 þ 14.95 NR
Chan MS. (2014) [30] Hong Kong RCT Unilateral recur, optime, cgp, seroma, reten, ssi Metallic tack 65 65 NA 53.31 þ 11.78 NR
Glue fixation 64 64 NA 75.84 þ 19.01 NR
Ayyaz M. (2015) [26] Ireland RCT Unilateral recur, optime, pain, reten Metallic tack 32 32 87.5 44.6 þ 16.3 NR
No fixation 31 31 90.32 31.3 þ 12.5 NR
Chandra P. (2015) [31] India RCT Unilateral recur, optime, pain, hos, cgp, seroma, Metallic tack 50 50 70 41.7 þ 8.51 28.72 ± 4.52
hem, reten, ssi Glue fixation 50 50 76 40.64 þ 8.39 28.92 ± 4.66
Claus CM. (2016) [34] Brazil RCT Unilateral recur, optime, hos, seroma, hem, injur No fixation 50 50 88 51.1 þ 15.7 NR
Absorbable tack 10 10 100 49 þ 14 NR
Jani K. (2016) [32] India RCT Unilateral recur, optime, pain, hos, cgp, seroma, Suture 127 173 93.7 45.4 þ 15.02 24.2 ± 3.5
Bilateral hem, reten, ssi Glue fixation 124 171 92.7 46.6 þ 15.02 23.7 ± 3.8
Buyukasik K. (2017) [27] Turkey RCT Unilateral recur, optime, pain, hos, cgp, Metallic tack 50 50 100 27.3 ± 7.0 28.1 ± 4.7
Bilateral No fixation 50 50 100 31.1 ± 12.8 28.2 ± 4.1
Shen YM. (2017) [33] China RCT Unilateral recur, optime, pain, hos, cgp, seroma, hem, ssi No fixation 80 80 81.2 60.0 þ 13.5 22.1 ± 2.3
Glue fixation 80 80 70 55.9 þ 14.6 22.3 ± 3.1

NR ¼ Not reported.
 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4 219

male respectively ranged from 29.2 to 64.8 years and 70 to 100, concealment (selection bias), respectively. Less than 50% of
and mean follow-up was 8e41 months. included studies had low risk of bias for blinding of partici-
A network map of interventions was constructed pants, blinding of outcome assessment (performance bias),
(Supplement file e Fig. 1). All 15 RCTs had two intervention data management (attribution bias), incomplete outcome data
arms, i.e., metallic tack versus no-fixation20e27 (N ¼ 8, n ¼ 979), (detection bias) and selective outcome reporting (reporting
metallic tack versus glue28e31 (N ¼ 4, n ¼ 382), suture fixation bias) (Supplement file e Table 1).
versus glue32 (N ¼ 1, n ¼ 251), glue33 and absorbable tack34
versus no-fixation (n ¼ 160 & 60). None of them reported Hernia recurrence
self-gripping mesh as an intervention.
The most common reported outcome was hernia recur- Interventions of 15 RCTs20e34 addressing recurrence were
rence (N ¼ 15) followed by mean operative time (N ¼ 12) and mapped (Fig. 2a); the most common comparison was no-
chronic groin pain (N ¼ 11); only a few reported return to fixation versus metallic tack (N ¼ 8, n ¼ 976) followed by
work20,29 and quality of life.30 Therefore, this outcome was glue versus metallic tack (N ¼ 4, n ¼ 382), absorbable tack
insufficient for pooling. For complications, some studies re- versus no-fixation (N ¼ 1, n ¼ 60), suture versus glue (N ¼ 1,
ported seroma (N ¼ 10), urinary retention (N ¼ 8), SSI (N ¼ 6), n ¼ 251), and no-fixation versus glue (N ¼ 1, n ¼ 160). All
but only a few reported hematoma (N ¼ 3) and intra-operative recurrence data are provided in Appendix 3-Table 1.
injury (N ¼ 2). These were combined as a composite compli- DMA was performed for no-fixation and glue versus
cation due to a small number of each event. metallic tack, yielding a pooled RRs of 1.43 (95% CI: 0.41, 4.98)
and 0.35 (95% CI: 0.06, 1.89), respectively (Appendix 3-Fig. 1).
Risk of bias in included studies The RRs were homogenous (I2 ¼ 0%) for both pooling. Funnel
plots did not show asymmetry (Appendix 3-Fig. 2).
Among 15 studies, 80% and 40% were considered at low risk of A NMA with consistency model was applied to pool RRs
bias for random sequence generation and allocation across studies using metallic tack as the reference. The

Fig. 2 e Network plot of comparisons of outcomes among different types of fixation in TEP inguinal hernia repair: a) hernia
recurrence b) chronic groin pain c) composite complication d) operative time. Size of nodes are weighted by number of
studies, size of edges are weighted by numbers of included subjects. Numbers on the edges refer to numbers of studies and
subjects, respectively.
220 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4

number of included studies and percent contribution of each suture and no-fixation, respectively. However, none of these
direct comparison are shown in Appendix 3-Fig. 3. Mixed reached statistical significance (Appendix 4-Fig. 4). Rankings
treatment comparisons indicated that absorbable tack and favored glue, followed by no-fixation, metallic tack, and su-
no-fixation had 7.3 and 1.6 times higher risk, whereas suture ture (Appendix 4-Fig. 5).
and glue had approximately 71% lower risk of recurrence than
metallic tack, but none of these was significant (Table 2). Composite complication
Likewise, predictive interval of these effects were not signifi-
cant (Appendix 3-Fig. 4). Glue was ranked best for lowering Thirteen RCTs20,21,24e34 reporting complications were mapped
recurrence, followed by suture fixation, whereas absorbable (Fig. 2c and Appendix 5-Table 1). The DMAs showed moderate
tack was ranked worst (Appendix 3-Fig. 5). to high heterogeneity (I2 ¼ 43.3% and 76.9%) for no-fixation
versus metallic tack and glue versus metallic tack (Appendix
Chronic groin pain 5-Fig. 1). The corresponding pooled RRs were 0.83 (95% CI:
0.52e1.31) and 1.04 (95% CI: 0.62e1.73), without evidence of
Eleven RCTs21e24,27e33 reported chronic groin pain (Fig. 2b), publication bias (Appendix 5-Fig. 2).
and their data are provided in Appendix 4-Table 1. DMAs of The number of RCTs and contributions to the NMA are
no-fixation and glue versus metallic tack yielded pooled RRs of shown in Appendix 5-Fig. 3. Suture had about 38% higher
0.64 (95% CI: 0.47e0.88) and 0.56 (95% CI: 0.306e1.02), respec- risk of having complications and absorbable tack had 37%
tively and no to low heterogeneity (I2 ¼ 0 and 26.5%, respec- lower risk, whereas glue and no-fixation were similar to
tively) (Appendix 4-Fig. 1). Funnel plots did not show evidence metallic mesh, see Table 4. Suture had about 2.2 and 1.5
of publication bias (Appendix 4-Fig. 2). times higher risk of developing complications than absorb-
The number of included studies and their contributions to able tack and no-fixation, respectively; glue had about 1.64
the network are shown in Appendix 4-Fig. 3. The NMA indi- times higher risk than absorbable tack. However, none of
cated that suture had about 2.58 times higher risk but glue and them was statistically significant (Appendix 5-Fig. 4).
no-fixation had 47% and 25% lower risk of groin pain than Rankings indicated that absorbable tack had the lowest
metallic tack (see Table 3). In addition, glue tended to be bet- complication rate, followed by no-fixation and glue
ter, i.e., 80% and 20% lower risk of chronic groin pain than (Appendix 5-Fig. 5).

Table 2 e Mixed relative treatment comparisons of hernia recurrence between mesh fixation techniques.
RR (95% CI) Mesh fixation technique
[SUCRA, %prob best]
No fixation Absorbable tack Suture fixation Glue fixation
No fixation 1.57 (0.49,5.07) 4.64 (0.10,221.24) 0.18 (0.00,12.94) 0.19 (0.03,1.02)
[31.5, 0.8]
Absorbable tack 7.30 (0.13,414.04) 0.04 (0.00,12.40) 0.04 (0.00,2.74)
[14.4, 4.7]
Suture fixation 0.29 (0.00,18.81) 1.02 (0.02,51.21)
[71.8, 49.6]
Glue fixation 0.29 (0.07,1.30)
[83.8, 43.3]

In diagonal lines were values of each mesh fixation technique compared with metallic tack (reference); values are risk ratios (RRs) with 95%
confident interval in parenthesis; values in the square parenthesis are SUCRA and probability of being the best technique in lowering recur-
rence. Values off the diagonal line are RRs along with 95% CI of the column intervention compared with the row intervention.

Table 3 e Mixed relative treatment comparisons of chronic groin pain between mesh fixation techniques.
RR (95% CI) [SUCRA, %prob best] Mesh fixation technique
No fixation Suture Glue fixation
No fixation 0.75 (0.33,1.68) 3.45 (0.12,95.33) 0.71 (0.21,2.41)
[61.3, 25.3]
Suture 2.58 (0.11,61.71) 0.20 (0.01,4.47)
[22.6, 14.9]
Glue fixation 0.53 (0.25,1.12)
[83.0, 59.2]

In diagonal lines were values of each mesh fixation technique compared with metallic tack (reference); values are risk ratios (RRs) with 95%
confident interval in parenthesis; values in the square parenthesis are SUCRA and probability of being the best technique in lowering chronic
groin pain. Values off the diagonal line are RRs along with 95% CI of the column intervention compared with the row intervention.
 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4 221

Table 4 e Mixed relative treatment comparisons of post-operative complication between mesh fixation techniques.
RR (95% CI) Mesh fixation technique
[SUCRA, %prob best]
No fixation Absorbable tack Suture Glue fixation
No fixation 0.95 (0.47,1.89) 0.66 (0.03,15.79) 1.46 (0.24,9.01) 1.09 (0.44,2.72)
[54.1, 13.3]
Absorbable tack 0.63 (0.02,16.13) 2.21 (0.06,85.36) 1.64 (0.06,44.58)
[62.5, 53.2]
Suture 1.38 (0.24,8.00) 0.74 (0.15,3.59)
[35.6, 15.1]
Glue fixation 1.03 (0.47,2.25)
[47.4, 8.3]

In diagonal lines were values of each mesh fixation technique compared with metallic tack (reference); values are risk ratios (RRs) with 95%
confident interval in parenthesis; values in the square parenthesis are SUCRA and probability of being the best technique in lowering
complication. Values off the diagonal line are RRs along with 95% CI of the column intervention compared with the row intervention.

Operative time Consistency assumption and publication bias

Data from 12 RCTs20e24,27,29e34 were used to compare opera- The consistency assumption was checked for individual out-
tive time, see Fig. 2d and Appendix 6-Table 1. DMA was per- comes (see Appendix 10 e Table 1). The global test yielded
formed which yielded the pooled mean differences of 3.18 p-value of 0.68, 0.85, 0.99 and 0.73 for hernia recurrence,
(95% CI -6.6, 0.24; I2 ¼ 0%) and 3.55 (95% CI -5.33, 1.76; chronic groin pain, composite complication and operative
I2 ¼ 66.1%) minutes for no-fixation and glue versus metallic time, respectively. This indicated that the consistency
tack, respectively (Appendix 6-Fig. 1). There was no evidence assumption was not violated for all outcomes.
of publication bias (Appendix 6-Fig. 2). Adjusted funnel plots were constructed for checking pub-
NMA indicated that suture fixation had about 4 min longer lication bias for all NMAs (see Appendix 10 e Fig. 1), which
operation time than metallic tack whereas the rest of the in- indicated no evidence of asymmetry.
terventions were not much different, see Appendix 6-Table 2.
No fixation was ranked best for this outcome, followed by glue
(Appendix 6-Fig. 3). Discussion

Post-operative pain We performed a NMA using data from 15 RCTs to assess

intervention effects on recurrence, chronic groin pain, com-
Ten RCTs21,22,24e29,32,33 were used to compared post-operative plications, and other relevant outcomes in TEP. Although our
pain score (Appendix 7-Fig. 1 and Appendix 7-Table 1). DMA results indicated non-significant differences among no fixa-
indicated a pooled mean difference in post-operative pain tion and various types of mesh fixations on surgical outcomes,
score between no fixation versus metallic tack of 0.09 (95% CI glue and suture seem to be highest-ranked for lowering
-0.23, 0.05; I2 ¼ 70.2%), see Appendix 7-Fig. 2. recurrence; whereas absorbable and no fixation were the
The NMA indicated that no fixation, suture, and glue had worst compared with metallic tack. In addition, glue was also
about 0.26, 0.28 and 0.3 pain scores units lower than metallic the best for lowering chronic groin pain compared to no fix-
tack, respectively, and only no-fixation was significant ation and suture. Furthermore, non-penetrating fixation such
(Appendix 7-Table 2). Rankogram indicated that no fixation as glue and no fixation had similar risk of complications;
had the lowest pain, followed by glue and suture, see whereas absorbable tack was much lower, but suture was
Appendix 7-Fig. 3. much higher when compared to metallic tack. Other out-
comes such as operative time, return to work and length of
Day of return to daily life hospital stay, in days, were not much different among these
techniques, except that suture seemed to require longer
Four RCTs20,25,28,29 reported return to daily life, i.e., no fixation operative time (minutes) than the others.
(N ¼ 3, n ¼ 246) and glue (N ¼ 1, n ¼ 93) versus metallic tack, Use of mesh, also known as tension free, hernia repair is
see Appendix 8-Fig. 1, Table 1. There were no significant dif- important for reinforcing the inguinal floor and thus pre-
ferences seen (Appendix 8-Fig. 2 and Appendix 8-Table 2). venting unfavorable outcomes such as recurrence and chronic
groin pain. Previous evidences of TEP by three meta-analyses
Length of hospital stay directly compared no-fixation11,12 and glue13 with metallic
tack, which did not demonstrate benefits in prevention of
Six RCTs21,25,27,29,31,33 reported length of hospital stay in day recurrence from using no-fixation and glue when compared to
between no fixations versus metallic tack groups (Appendix 9). metallic tack, see Supplement file, Table 2. The two former
Applying DMA yielded a mean difference of 0.06 (95% CI meta-analyses11,12 included 5 to 6 RCTs, but they mainly
-0.37, 0.21) day with heterogeneity (I2 ¼ 79.6%). pooled effect sizes using data from only 2 RCTs with similar
222 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4

pooled odds ratios of 2.01 (0.37, 11.02) because the other RCTs infection.40,41 Only adjacent organs (e.g., bladder, bowel or
were excluded due to zero recurrence in both intervention and vessels) are directly related to the tack, which causes injury
comparator. The later meta-analysis17 pooled data based on during the process of keeping mesh in place.42e44
only 3 observational studies with the pooled odds ratio of In the early post-operative period, pain is a common
metallic tack versus glue of 1.61 (0.40, 6.42). All these pooling concern. Severity of pain may depend on tissue injury from
results were not significant; as a result, the most recent in- dissection, or the use of penetrating mesh fixation that en-
ternational guidelines with moderate level of evidence stated traps nerve or muscle.45e47 This fits with our result of about 0.3
that any type of mesh fixation is unnecessary but recommend units lower the 24-h pain scores with non-penetrating mesh
with weak evidence that mesh fixation might be required for fixation (i.e., no fixation and glue) when compared to metallic
only a large direct hernia.5 However, these pointed estima- mesh.
tions of pooled effect sizes seemed to favor metallic tack and In this network meta-analysis, we found little or no dif-
glue for comparisons to no-fixation versus metallic tack and ference between mesh fixation technique on operative time,
metallic tack versus glue given uncertainty of estimations hospital stay, and day return to daily life. Therefore,
because of very wide 95% CIs. A number of included studies metallic tack may be unnecessary in TEP, and alternative
may play a role for the uncertainty of these direct meta- mesh fixation techniques should be considered because
analyses. Our study was conducted attempting to increase they do not appear to increase recurrence or complications.
numbers of RCTs (N ¼ 15) and increased numbers of subjects However, there is only one study reporting suture technique
in each intervention arm by applying a NMA method to and another one for absorbable tack; neither of these is
combine direct comparative data with indirect data using common in general practice. Suture fixation in totally extra-
common comparators such as metallic tack, no-fixation, and peritoneal hernia repair is technically difficult due to limited
even glue to compare benefits and complications across in- space and angle of use. Absorbable tack is expensive, more
terventions. This method leads to increased power to detect than double the cost when compared with metallic tack;
intervention effects compared to conventional DMA. there was no data about self-gripping mesh in this network
Although our estimated intervention effects were still non- meta-analysis.
significant, they were more precise estimations, which could Our study had some strengths. We applied NMA to
be seen from narrowing 95% CIs compared with previous compare the most relevant clinical outcomes considering
meta-analyses. These findings however need to be confirmed both efficacy and complications of type of mesh fixation
by a large-scale RCT. techniques available in clinical practice. Probability of being
The main concern for management of groin hernia is the best technique was estimated and ranked. However, the
hernia recurrence, however, chronic groin pain and compli- number of included RCTs for each outcome was still small and
cations can also threaten patients' quality-of-life and should this NMA should be updated when additional RCTs are pub-
influence decision making. Our findings indicate that glue is lished. In addition, none of the included RCTs applied self-
the best technique in lowering recurrence and chronic groin gripping mesh which is cheaper than the standard metallic
pain and has a similar risk of complications when compared mesh.
to the standard metallic tack. However, the cost of applying
glue is about $170 more expensive than applying metallic tack.
Therefore, cost-effectiveness should be further assessed for Conclusion
this option in routine practice. Although the suture technique
ranked closely with glue for prevention of recurrence, it No-mesh fixation or mesh fixation technique with glue might
appeared to cause more chronic groin pain and complications. be alternatives for TEP groin hernia repair as they could lower
Other techniques with good efficacy but low complications recurrence rate, chronic groin pain, and complications when
should also be explored as alternative options in low/middle compared to metallic tack. However, our findings need to be
income countries. For instance, self-gripping mesh e a new confirmed by a large-scale RCT. In addition, economic evalu-
type of mesh that is used for both open35 and laparoscopic ation should also be further assessed.
hernia repair,36 was comparable in both perioperative and
long term outcomes with suture mesh in a meta-analysis of
open hernia surgery.37 However, evidence about efficacy of Disclosure of funding
self-gripping mesh in TEP was not available. Given that the
cost of self-gripping mesh is cheaper than the standard Suphakarn Techapongsatorn is a Ph.D. student of Clinical
metallic tack, further RCTs are needed to compare both clin- Epidemiology, Faculty of Medicine Ramathibodi Hospital and
ical effectiveness and cost-utility. Faculty of Graduate Studies, Mahidol University. This study is
Complications that are usually reported in post laparo- a part of his dissertation which will be applied for granting
scopic inguinal hernia repair include SSI, seroma, hematoma, graduation. All authors have no conflict of interest and all
urinary retention, and adjacent structural injury. The cause of author approval for submission to The Surgeon.
these complications were multifactorial, such as seroma, he-
matoma and urinary retention usually related with dissection
area, time of surgery or hemostasis.38,39 Surgical site infection Appendix A. Supplementary data
(either superficial or deep) may be caused by suture material,
mesh or foreign body material from mesh fixation device; Supplementary data to this article can be found online at
these may cause a nidus for bacterial contamination and https://doi.org/10.1016/j.surge.2018.09.002.
 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4 223

references reporting of systematic reviews incorporating network meta-

analyses of health care interventions: checklist and
explanations. Ann Intern Med 2015;162(11):777e84.
19. Armijo-Olivo S, Stiles CR, Hagen NA, Biondo PD,
1. Jenkins JT, O'Dwyer PJ. Inguinal hernias. BMJ (Clinical research
Cummings GG. Assessment of study quality for systematic
ed) 2008;336(7638):269e72.
reviews: a comparison of the cochrane collaboration risk of
2. Kingsnorth A, LeBlanc K. Hernias: inguinal and incisional.
bias tool and the effective public health practice project
Lancet 2003;362(9395):1561e71.
quality assessment Tool: methodological research. J Eval Clin
3. Simons MP, Aufenacker T, Bay-Nielsen M, Bouillot JL,
Pract 2012;18(1):12e8.
Campanelli G, Conze J, et al. European Hernia Society
20. Ferzli GS, Frezza EE, Pecoraro Jr AM, Ahern KD. Prospective
guidelines on the treatment of inguinal hernia in adult
randomized study of stapled versus unstapled mesh in a
patients. Hernia 2009;13(4):343e403.
laparoscopic preperitoneal inguinal hernia repair. J Am Coll
4. Miserez M, Peeters E, Aufenacker T, Bouillot JL, Campanelli G,
Surg 1999;188(5):461e5.
Conze J, et al. Update with level 1 studies of the European
21. Moreno-Egea A, Torralba Martinez JA, Morales Cuenca G,
Hernia Society guidelines on the treatment of inguinal hernia
Aguayo Albasini JL. Randomized clinical trial of fixation vs
in adult patients. Hernia 2014;18(2):151e63.
nonfixation of mesh in total extraperitoneal inguinal
5. International guidelines for groin hernia management. Hernia
hernioplasty. Arch Surg 2004;139(12):1376e9.
2018;22(1):1e165.
22. Koch CA, Greenlee SM, Larson DR, Harrington JR, Farley DR.
6. McCormack K, Scott NW, Go PM, Ross S, Grant AM.
Randomized prospective study of totally extraperitoneal
Laparoscopic techniques versus open techniques for inguinal
inguinal hernia repair: fixation versus no fixation of mesh.
hernia repair. Cochrane Database Syst Rev 2003;(1), Cd001785.
JSLS 2006;10(4):457e60.
7. Kockerling F, Bittner R, Jacob DA, Seidelmann L, Keller T,
23. Taylor C, Layani L, Liew V, Ghusn M, Crampton N, White S.
Adolf D, et al. TEP versus TAPP: comparison of the
Laparoscopic inguinal hernia repair without mesh fixation,
perioperative outcome in 17,587 patients with a primary
early results of a large randomised clinical trial. Surg Endosc
unilateral inguinal hernia. Surg Endosc 2015;29(12):3750e60.
2008;22(3):757e62.
8. Wake BL, McCormack K, Fraser C, Vale L, Perez J, Grant AM.
24. Subwongcharoen S, Ruksakul K. A randomized controlled
Transabdominal pre-peritoneal (TAPP) vs totally
trial of staple fixation versus N-butyl-2-cyanoacrylate fixation
extraperitoneal (TEP) laparoscopic techniques for inguinal
in laparoscopic inguinal hernia repair. J Med Assoc Thai
hernia repair. Cochrane Database Syst Rev 2005;(1), Cd004703.
2013;96(Suppl 3):S8e13.
9. Gass M, Banz VM, Rosella L, Adamina M, Candinas D,
25. Garg P, Nair S, Shereef M, Thakur JD, Nain N, Menon GR, et al.
Guller U. TAPP or TEP? Population-based analysis of
Mesh fixation compared to nonfixation in total
prospective data on 4,552 patients undergoing endoscopic
extraperitoneal inguinal hernia repair: a randomized
inguinal hernia repair. World J Surg 2012;36(12):2782e6.
controlled trial in a rural center in India. Surg Endosc
10. Shi Z, Fan X, Zhai S, Zhong X, Huang D. Fibrin glue versus
2011;25(10):3300e6.
staple for mesh fixation in laparoscopic transabdominal
26. Ayyaz M, Farooka MW, Malik AA, Khan A, Mansoor R, Toor AA,
preperitoneal repair of inguinal hernia: a meta-analysis and
et al. Mesh fixation vs. non-fixation in total extra peritoneal
systematic review. Surg Endosc 2017;31(2):527e37.
mesh hernioplasty. J Pakistan Med Assoc 2015;65(3):270e2.
11. Tam KW, Liang HH, Chai CY. Outcomes of staple fixation of
27. Buyukasik K, Ari A, Akce B, Tatar C, Segmen O, Bektas H.
mesh versus nonfixation in laparoscopic total extraperitoneal
Comparison of mesh fixation and non-fixation in
inguinal repair: a meta-analysis of randomized controlled
laparoscopic totally extraperitoneal inguinal hernia repair.
trials. World J Surg 2010;34(12):3065e74.
Hernia 2017;21(4):543e8.
12. Teng YJ, Pan SM, Liu YL, Yang KH, Zhang YC, Tian JH, et al. A
28. Parshad R, Kumar R, Hazrah P, Bal S. A randomized
meta-analysis of randomized controlled trials of fixation
comparison of the early outcome of stapled and unstapled
versus nonfixation of mesh in laparoscopic total
techniques of laparoscopic total extraperitoneal inguinal
extraperitoneal inguinal hernia repair. Surg Endosc
hernia repair. JSLS 2005;9(4):403e7.
2011;25(9):2849e58.
29. Lau H. Fibrin sealant versus mechanical stapling for mesh
13. Kaul A, Hutfless S, Le H, Hamed SA, Tymitz K, Nguyen H, et al.
fixation during endoscopic extraperitoneal inguinal
Staple versus fibrin glue fixation in laparoscopic total
hernioplasty: a randomized prospective trial. Ann Surg
extraperitoneal repair of inguinal hernia: a systematic review
2005;242(5):670e5.
and meta-analysis. Surg Endosc 2012;26(5):1269e78.
30. Chan MS, Teoh AY, Chan KW, Tang YC, Ng EK, Leong HT.
14. Antoniou SA, Kohler G, Antoniou GA, Muysoms FE,
Randomized double-blinded prospective trial of fibrin sealant
Pointner R, Granderath FA. Meta-analysis of randomized
spray versus mechanical stapling in laparoscopic total
trials comparing nonpenetrating vs mechanical mesh fixation
extraperitoneal hernioplasty. Ann Surg 2014;259(3):432e7.
in laparoscopic inguinal hernia repair. Am J Surg 2016;211(1).
31. Chandra P, Phalgune D, Shah S. Comparison of the clinical
239-49.e2.
outcome and complications in laparoscopic hernia repair of
15. Lederhuber H, Stiede F, Axer S, Dahlstrand U. Mesh fixation in
inguinal hernia with mesh fixation using fibrin glue vs tacker.
endoscopic inguinal hernia repair: evaluation of methodology
Indian J Surg 2016;78(6):464e70.
based on a systematic review of randomised clinical trials.
32. Jani K. Randomised controlled trial of n-butyl cyanoacrylate
Surg Endosc 2017;31(11):4370e81.
glue fixation versus suture fixation of mesh in laparoscopic
16. Sajid MS, Ladwa N, Kalra L, Hutson K, Sains P, Baig MK. A
totally extraperitoneal hernia repair. J Minimal Access Surg
meta-analysis examining the use of tacker fixation versus no-
2016;12(2):118e23.
fixation of mesh in laparoscopic inguinal hernia repair. Int J
33. Shen YM, Liu YT, Chen J, Sun L. Efficacy and safety of NBCA
Surg 2012;10(5):224e31.
(n-butyl-2-cyanoacrylate) medical adhesive for patch fixation
17. Shah NS, Fullwood C, Siriwardena AK, Sheen AJ. Mesh
in totally extraperitoneal prosthesis (TEP): a prospective,
fixation at laparoscopic inguinal hernia repair: a meta-
randomized, controlled trial. Eur Rev Med Pharmacol Sci
analysis comparing tissue glue and tack fixation. World J Surg
2017;21(4):680e6.
2014;38(10):2558e70.
34. Claus CM, Rocha GM, Campos AC, Bonin EA, Dimbarre D,
18. Hutton B, Salanti G, Caldwell DM, Chaimani A, Schmid CH,
Loureiro MP, et al. Prospective, randomized and controlled
Cameron C, et al. The PRISMA extension statement for
224 t h e s u r g e o n 1 7 ( 2 0 1 9 ) 2 1 5 e2 2 4

study of mesh displacement after laparoscopic inguinal 41. Salamone G, Licari L, Augello G, Campanella S, Falco N,
repair: fixation versus no fixation of mesh. Surg Endosc Tutino R, et al. Deep SSI after mesh-mediated groin hernia
2016;30(3):1134e40. repair: management and outcome in an Emergency Surgery
35. Sanders DL, Nienhuijs S, Ziprin P, Miserez M, Gingell- Department. G Chir 2017;38(1):41e5.
Littlejohn M, Smeds S. Randomized clinical trial comparing 42. Ates M, Kinaci E, Kose E, Soyer V, Sarici B, Cuglan S, et al.
self-gripping mesh with suture fixation of lightweight Corona mortis: in vivo anatomical knowledge and the risk of
polypropylene mesh in open inguinal hernia repair. Br J Surg injury in totally extraperitoneal inguinal hernia repair. Hernia
2014;101(11):1373e82. 2016;20(5):659e65.
36. Ferrarese A, Bindi M, Rivelli M, Solej M, Enrico S, Martino V. 43. Kojima S, Sakamoto T, Honda M, Nishiguchi R, Ogawa F.
Self-gripping mesh versus fibrin glue fixation in laparoscopic Rare complication after totally extraperitoneal endoscopic
inguinal hernia repair: a randomized prospective clinical trial inguinal hernia repair: small bowel perforation
in young and elderly patients. Open Med (Wars) without peritoneal disruption. Asian J Endosc Surg
2016;11(1):497e508. 2016;9(4):311e3.
37. Pandanaboyana S, Mittapalli D, Rao A, Prasad R, Ahmad N. 44. Xue TM, Tao LD, Zhang J, Zhang PJ. Mesh erosion causes small
Meta-analysis of self-gripping mesh (Progrip) versus sutured bowel obstruction: a rare complication of laparoscopic
mesh in open inguinal hernia repair. Surgeon inguinal hernia repair: case description and review of
2014;12(2):87e93. literature. Hepato-Gastroenterology 2015;62(137):55e8.
38. Kargar S, Shiryazdi SM, Zare M, Mirshamsi MH, Ahmadi S, 45. Hedberg HM, Hall T, Gitelis M, Lapin B, Butt Z, Linn JG, et al.
Neamatzadeh H. Comparison of postoperative short-term Quality of life after laparoscopic totally extraperitoneal repair
complications after laparoscopic transabdominal of an asymptomatic inguinal hernia. Surg Endosc
preperitoneal (TAPP) versus Lichtenstein tension free 2018;32(2):813e9.
inguinal hernia repair: a randomized trial study. Minerva Chir 46. Liew W, Wai YY, Kosai NR, Gendeh HS. Tackers versus glue
2015;70(2):83e9. mesh fixation: an objective assessment of postoperative
39. Liu H, Zheng X, Gu Y, Guo S. A meta-analysis examining the acute and chronic pain using inflammatory markers. Hernia
use of fibrin glue mesh fixation versus suture mesh fixation in 2017;21(4):549e54.
open inguinal hernia repair. Dig Surg 2014;31(6):444e51. 47. Shazi B, Koto MZ. Comparing outcomes after laparoscopic
40. Boonchan T, Wilasrusmee C, McEvoy M, Attia J, totally extraperitoneal repair versus open (Lichtenstein)
Thakkinstian A. Network meta-analysis of antibiotic repair of inguinoscrotal hernia at dr george mukhari
prophylaxis for prevention of surgical-site infection after academic hospital. S Afr J Surg 2017;55(2):75e6.
groin hernia surgery. Br J Surg 2017;104(2):e106e17.