Research

JAMA Surgery | Original Investigation | ASSOCIATION OF VA SURGEONS

Factors Associated With Long-term Outcomes

of Umbilical Hernia Repair
Divya A. Shankar, BA; Kamal M. F. Itani, MD; William J. O’Brien, MS; Vivian M. Sanchez, MD

Invited Commentary
IMPORTANCE Umbilical hernia repair is one of the most commonly performed general surgical Supplemental content
procedures. However, there is little consensus about the factors that lead to umbilical hernia
recurrence.

OBJECTIVE To better understand the factors associated with long-term umbilical hernia
recurrence.

DESIGN, SETTING, AND PARTICIPANTS A retrospective cohort of 332 military veteran patients
who underwent umbilical hernia repair was studied between January 1, 1998, and December
31, 2008, at the VA Boston Healthcare System. Recurrence and mortality outcomes were
tracked from that period until June 1, 2014. Data were collected on patient characteristics,
operative, and postoperative factors and univariate and multivariable analyses were used to
assess which factors were significantly associated with umbilical hernia recurrence and
mortality. All patients with primary umbilical hernia repair, with or without a concurrent
unrelated procedure, were included in the study. Patients excluded were those who
underwent umbilical hernia repair as a part of another major planned procedure with
abdominal incisions. Data were collected from June 1, 2014, to November 1, 2015. Statistical
analysis was performed from November 2, 2015, to April 1, 2016.

MAIN OUTCOMES AND MEASURES The primary study outcomes were umbilical hernia
recurrence and death.

RESULTS Of the 332 patients in this study, 321 (96.7%) were male, mean age was 58.4 years,
and mean (SD) time of follow-up was 8.5 (4.1) years. The hernia recurrence rate was 6.0%
(n = 20) at a mean 3.1 years after index repair (median, 1.0-year; range, 0.33-13 years). The
primary suture repair recurrence rate was 9.8% (16 of 163 patients), and the mesh repair
recurrence rate was 2.4% (4 of 169 patients). On univariate analysis, ascites (P = .02), liver
disease (P = .02), diabetes (P = .04), and primary suture (nonmesh) repairs (P = .04) were
significantly associated with increased recurrence rates. Patients who had a history of hernias
(125 [39%]) were less likely to have umbilical hernia recurrences (χ 12 = 4.65, P = .03). On
multivariable regression analysis, obesity and ascites were associated with significantly
increased odds ratios of recurrence of 3.3 (95% CI, 1.0-10.1) and 8.0 (95% CI, 1.8-34.4),
respectively. Mesh repair was seen to decrease recurrence with odds of 0.28 (95% CI,
0.08-0.95). There was no significant difference in complication rates between mesh repair
and primary suture repair. The survival rate was 73% (n = 242) at the end of the study. Factors
associated with mortality were older age, smoking, liver disease, ascites, emergency or
semiurgent repair, and need for intraoperative bowel resection.

Author Affiliations: Department of

CONCLUSIONS AND RELEVANCE Ascites, liver disease, diabetes, obesity, and primary suture
Surgery, Boston University School of
repair without mesh are associated with increased umbilical hernia recurrence rates. Elective Medicine, Boston, Massachusetts
umbilical hernia repair with mesh should be considered in patients with multiple (Shankar, Itani, Sanchez);
comorbidities given that the use of mesh offers protection from recurrence without major Department of Surgery, VA Boston
Healthcare System, Boston,
morbidity.
Massachusetts (Itani, O’Brien,
Sanchez).
Corresponding Author: Vivian M.
Sanchez, MD, Department of Surgery,
Boston University School of
Medicine, 1400 VFW Pkwy, SS112,
JAMA Surg. doi:10.1001/jamasurg.2016.5052 West Roxbury, MA 02132
Published online January 25, 2017. (vivian.sanchez@va.gov).

(Reprinted) E1

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017

 Research Original Investigation
U
mbilical hernias account for approximately 6% to 14%
of all abdominal wall hernias in adults, and almost 90%
of adult umbilical hernias are acquired.1,2 Despite um-
bilical hernia being common, there are no set surgical guide-
lines for its repair and there is no consensus on the best type
of repair.2,3 Recurrence rates range from 1% to 43%,3-6 but the
literature offers little consensus on factors that affect recur-
rence. Asolati et al1 found that type 2 diabetes, hyperlipid-
emia, and human immunodeficiency virus–positive status may
be a factor in increased recurrence rates. Obesity (body mass
index [BMI; calculated as weight in kilograms divided by height
in meters squared]>30) has also been associated with higher
rates of recurrence.7,8 A recent study9 found that patients
with another known hernia at the time of repair were at in-
creased risk of recurrence. In addition, studies have shown
that hernia defects larger than 2 cm have increased rates of
recurrence and that mesh repairs may decrease rates of
recurrence.7,10 Still, few studies to date have assessed a wide
range of patient and operative factors to determine predic-
tors of umbilical hernia recurrence. In addition, to our knowl-
edge, no study to date has had mean follow-up of patients with
umbilical hernia repairs beyond 7 years. The primary aim of
this study was to assess the long-term risk of umbilical hernia
recurrences and the role that various patient characteristics and
surgical techniques play.
To better understand the factors associated with umbili-
cal hernia recurrence, we reviewed a 10-year cohort of vet-
eran patients who underwent umbilical hernia repair at the
VA (Veterans Affairs) Boston Healthcare System, Boston,
Massachusetts.
Methods
We studied a retrospective cohort of 332 patients who
underwent umbilical hernia repair at the VA Boston Health-
care System between January 1, 1998, and December 31,
2008, and we tracked recurrence and mortality outcomes
from that period until June 1, 2014. The study was approved
by the institutional review board of the VA Boston Health-
care System and met the criteria for patient consent exclu-
sion. Data were collected from June 1, 2014, to November 1,
2015. Statistical analysis was performed from November 2,
2015, to April 1, 2016.
Using VistA Computerized Patient Record System (CPRS),
version 1.0.30.75 (US Department of Veterans Affairs), we iden-
tified patients on the basis of Current Procedural Terminology
codes for various umbilical hernia repairs. Patients included
were those who underwent primary umbilical hernia repair (ie,
open, laparoscopic, recurrent, emergent, or elective), 2 con-
current hernia repairs (ie, for inguinal hernia and umbilical her-
nia), and umbilical hernia repair with a concurrent unrelated
procedure (eg, umbilical hernia repair and breast biopsy). Pa-
tients excluded were those who underwent umbilical hernia
repair as a part of another major planned procedure with ab-
dominal incisions (eg, umbilical hernia repair during a colec-
tomy). The primary endpoints were recurrence (reoperation
or clinical) and death. Patients were noted to have a hernia re-
E2 JAMA Surgery Published online January 25, 2017 (Reprinted)
Copyright 2017 American Medical Association. All rights reserved.
Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017
Factors Associated With Long-term Outcomes of Umbilical Hernia Repair
Key Points
Question What factors lead to umbilical hernia recurrence?
Findings This cohort study of 332 military veteran patients who
underwent umbilical hernia repair found that ascites, liver disease,
diabetes, obesity, and primary suture repair were significantly
associated with increased rates of umbilical hernia recurrence.
There was no significant difference in complication rates between
mesh repair and primary suture repair.
Meaning Mesh repair should be considered for all patients
undergoing umbilical hernia repair, given its lower recurrence rates
and complication rates that are similar to that of primary suture
repair.
currence if it was reported during a physical examination and
documented in the CPRS, if they had an operation for recur-
rence, or if it was indicated in imaging findings. The fol-
low-up period was from the time of index repair to the last avail-
able record as of June 1, 2014, or death.
We collected patient data through medical record review
of operative notes, general surgery notes, primary care notes,
radiology reports, and other medical-surgical specialty notes.
We studied 51 patient comorbidities, operative factors, and
postoperative complications thought to be clinically signifi-
cant in umbilical hernia recurrence. The variables and crite-
ria used for each patient, operative, and postoperative fac-
tors are shown in the eTable in the Supplement.
We used logistic regression models to estimate the odds
of recurrence and mortality as a function of demographic and
clinical factors. The models included a time trend to control
for improvements in medical knowledge and technology dur-
ing the study period. We initially examined univariate rela-
tionships of all available variables and the 2 outcomes. Vari-
ables with a χ2 or Fisher exact test of P < .05 were eligible for
inclusion in the multivariable models. We further narrowed the
candidate variables to avoid multicollinearity (eg, liver dis-
ease and alcohol abuse). The final covariates in the recur-
rence model included age older than 60 years, chronic ob-
structive pulmonary disease, BMI greater than 30, ascites, mesh
repair, and emergency or semiurgent operation. The mortal-
ity model included age as a continuous variable, smoking, liver
disease, BMI greater than 35, large defect (≥2 cm), and emer-
gency or semiurgent operation. We measured model discrimi-
nation using the C statistic. We also ran a sensitivity analysis
using the same model specification, but with the redefined out-
comes of 7-year mortality and 7-year recurrence. We per-
formed all analyses in SAS, version 9.3 (SAS Institute Inc).
Results
Baseline Patient Characteristics
There were 332 consecutive patients who underwent umbili-
cal hernia repair between January 1, 1998, and December 31,
2008, and who fit our inclusion criteria. In our patient
cohort, most were male (321 [96.7%]); the mean age was
58.4 years; the mean (SD) time of follow-up was 8.5 (4.1)
jamasurgery.com
 Factors Associated With Long-term Outcomes of Umbilical Hernia Repair Original Investigation Research

years; and the mean BMI was 30.9 with 178 (54%) patients
identified as obese, as defined by a BMI of 30 or greater.
Among these patients, 242 (73%) had a smoking history and
70 (21%) had liver disease, as defined by cirrhosis, hepatitis,
or ascites. In addition, 163 (49%) underwent primary suture
repairs and 169 (51%) underwent mesh repairs. There was no
significant difference in the underlying medical conditions
(eg, diabetes, chronic obstructive pulmonary disease, asci-
tes) between patients who underwent primary suture repair
and those who underwent mesh repair. Of the hernias
repaired, 292 (88%) were elective, 30 (9%) were an emer-
gency, and 13 (4%) were semiurgent; 4 repairs (1.2%) were
laparoscopic, and the remainder were open repairs. Eighty
(24%) patients had concomitant operations at the time of
their umbilical hernia repair, and 60 (74%) of these proce-
dures were other hernia operations.
Figure 1. Time From Index Repair to Umbilical Hernia Recurrence
100
90
80
70
% Recurrence
Recurrence
The recurrence rate was 6.0% (n = 20) at a mean 3.1 years after
index repair (median, 1.0-year; range, 0.33-13 years). The primary
suture repair recurrence rate was 9.8% (16 of 163 patients), and
the mesh repair recurrence rate was 2.4% (4 of 169 patients). Post-
operatively, 10 (50%) of the recurrences occurred within the first
year, 15 (75%) occurred within the first 5 years, and 5 (25%) oc-
curred after 5 years (Figure 1). Sixteen patients (80%) who had
a recurrence underwent primary suture repairs, and 4 (20%) un-
derwent mesh repairs. Within the primary suture repairs, 4 of the
fascial defects were repaired with absorbable sutures, 11 were re-
paired with non–absorbable sutures, and 1 was repaired with un-
known sutures. Within the mesh repairs, 3 were overlays and 1
was a combination underlay and overlay. Mesh location and type
of suture could not be analyzed because of our small sample size.
On univariate analysis, ascites (Fisher exact test, P = .02),
liver disease (Fisher exact test, P = .02), diabetes (χ 21 = 4.17,
P = .04), and primary suture (nonmesh) repairs (χ 21 = 8.13,
P = .04) were significantly associated with increased rates of
recurrence (Table 1). Patients who had a history of hernias (125
[39%]) were less likely to have umbilical hernia recurrences
(P = .03). Seventy-six (61%) patients with prior hernias re-
ceived mesh as opposed to 87 (44%) patients without a his-

60 tory of hernia (P = .003). No postoperative complications sig-

50 nificantly affected the rates of recurrence. Defect size did not
40 affect the rates of recurrence; however, defect size was only
30 recorded in 174 (54%) patients. On multivariable regression
20
analysis, obesity and ascites were associated with signifi-
cantly increased odds ratios (ORs) of recurrence—3.3 (95% CI,
10
1.0-10.1) and 8.0 (95% CI, 1.8-34.4), respectively (Table 2). Mesh
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 repair was seen to decrease recurrence by 3.6 times with odds
Postoperative Month of recurrence of 0.28 (95% CI, 0.08-0.95).
No. at risk 20 15 8 7 7 7 5 5 5 5 5 3 3 3

Morbidity
This figure shows 50% of recurrences occurred with the first year of operation, Sixty-one (18%) patients experienced at least 1 complication
75% within the first 5 years, and 25% after 5 years.
within 30 days of repair. The most common complication was

Table 1. Baseline Patient Characteristics Stratified by Recurrence and Death

No. (%)
Recurrence Death
Characteristic No Yes P Value No Yes P Value
No. 312 20 242 90
Age, mean (SD), y 58.4 (11.7) 53.9 (9.2) .04 56.1 (11.3) 64.3 (10.4) <.001
Male 301 (96.5) 20 (100) >.99 234 (96.7) 87 (96.7) >.99
Diabetes 102 (32.7) 11 (55) .04 88 (36.4) 25 (27.8) .14
Obesity 164 (52.6) 14 (70) .13 136 (56.2) 42 (46.7) .12
Morbid obesity 18 (5.8) 3 (15) .09 17 (7.0) 4 (4.4) .39
Smoking 229 (73.4) 13 (65) .41 168 (69.4) 74 (82.2) .02
Liver disease 43 (13.8) 7 (35) .02 26 (10.7) 24 (26.7) <.001
Ascites 16 (5.1) 4 (20) .02 1 (0.4) 19 (21.1) <.001
Chronic obstructive pulmonary disease 92 (29.5) 2 (10) .06 63 (26) 31 (34.4) .13
History of hernia 122 (39.1) 3 (15) .03 96 (39.7) 29 (32.2) .21
Primary suture repair 147 (47.1) 16 (80) .04 111 (45.9) 52 (57.8) .05
Defect ≥2 cm 92 (29.5) 4 (20) .36 79 (32.6) 17 (18.9) .01
Emergency or semiurgent operation 39 (12.5) 4 (20) .31 22 (9.1) 21 (23.3) <.001
Bowel resection 3 (1) 1 (5) .22 0 4 (4.4) .005

jamasurgery.com (Reprinted) JAMA Surgery Published online January 25, 2017 E3

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017

 Research Original Investigation Factors Associated With Long-term Outcomes of Umbilical Hernia Repair

Table 2. Multivariable Analysis of Recurrencea Table 3. Multivariable Analysis of Deatha

Characteristic Recurrence, Odds Ratio (95% CI) Characteristic Death, Odds Ratio (95% CI)
Time trend 0.92 (0.78-1.08) Time trend 0.86 (0.78-0.94)
Age >60 y 0.47 (0.16-1.41) Age 1.09 (1.06-1.13)
Chronic obstructive pulmonary disease 0.27 (0.06-1.30) Smoking 2.27 (1.12-4.56)
Obesity 3.3 (1.0-10.1) Liver disease 5.60 (2.69-12.10)
Ascites 8.0 (1.8-34.4) Morbid obesity 0.86 (0.23-3.20)
Mesh repair 0.28 (0.08-0.95) Large defect 0.44 (0.23-0.86)
Emergent or semiurgent operation 1.10 (0.38-3.92) Emergent or semiurgent operation 2.18 (1.02-4.67)
a a
C statistic, 0.805. C statistic, 0.807.

ascites (χ 21 = 49.64, P < .001), emergency or semiurgent repair

Figure 2. Percentage of Patient Survival After Index Repair
(χ 21 = 11.80, P < .001), and need for intraoperative bowel resec-
100 tion (Fisher exact test, P = .005) were significantly associ-
90 ated with increased long-term mortality rates. Twenty deaths
80 occurred more than 30 days after emergency or semiurgent re-
70
pair. These all occurred more than 30 days postoperatively.
Patient Survival, %

Twelve (60%) of these deaths were secondary to unknown

60
causes and 25 (5%) were secondary to liver disease. On mul-
50
tivariable regression analysis, age (OR, 1.09; 95% CI, 1.06-
40
1.13), history of smoking (OR, 2.27; 95% CI, 1.12-4.56), liver dis-
30
ease (OR, 5.60; 95% CI, 2.69-12.10), and emergency or
20
semiurgent repair (OR, 2.18; 95% CI, 1.02-4.67) significantly
10 increased the odds of long-term mortality (Table 3). Defects
0 of 2 cm or greater had a protective effect on mortality with a
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Postoperative Years 0.44 OR of death (95% CI, 0.23-0.86).

No. at risk 332 323 308 298 290 282 275 266 258 254 249 248 248 245 243

This figure shows a 27% mortality rate at a mean of 5.1 years after index repair.
seroma (31 [9.6%]), followed by surgical-site infection (22
[6.9%]) and hematomas (8 [2.4%]). Of note, 3 of the 4 pa-
tients (75%) who needed bowel resection at the time of index
repair developed a surgical-site infection. There were 2 cases
of mesh infection and 3 cases of ascites leaks. Thirty-four pa-
tients (20%) who had a mesh repair had 1 or more complica-
tions, which was slightly higher, but not significantly, than the
27 patients (16%) with primary suture repair who had a com-
plication after the procedure (P = .35).
Mortality
The 30-day mortality rate was 0.3% (n = 1). This 1 death oc-
curred on postoperative day 21 for unclear reasons.
Survival
The total survival rate was 73% (n = 242) during the period of
follow-up, as assessed by our medical record review from June
1, 2014, to November 1, 2015 (Figure 2). Ninety patients (27%)
died at a mean of 5.1 years postoperatively (median, 5 years),
despite the mean age of 58.4 years of our patient cohort. Among
the deaths, 43 (48%) were of unknown causes, 18 (20%) were
cancer related, 12 (13%) were secondary to liver disease, 6 (6.7%)
were secondary to heart disease, 4 (4.4%) were secondary to
renal disease, and 3 (3.3%) were secondary to sepsis. On uni-
variate analysis (Table 1), older age (t = 6.19, P < .001), smok-
ing (χ 21 = 5.44, P = .02), liver disease (χ 21 = 13.00, P < .001),
Discussion
Umbilical hernias are a common problem encountered by gen-
eral surgeons. Despite umbilical hernias being common, little
has changed in their indications and methods of repair for many
years. Several studies have attempted to analyze the factors
associated with recurrence with short-term follow-up. In con-
trast, our study focused on surgical procedures during a 10-
year period, and our mean follow-up time of 8.5 years is lon-
ger than that of any comparable study. Because of the small
patient migration outside of the VA system and the ability to
access the CPRS and Medicare mortality data, we have been
able to capture outcomes that are difficult to study in other
health care settings. Thus, our study sought to seek out the fac-
tors that might be associated with patients who are more likely
to develop hernia recurrences.
Our finding of a 6.0% umbilical hernia recurrence rate is
comparable to that of other studies. The primary suture re-
pair recurrence rate was 9.8% and the mesh recurrence rate
was 2.4%, which are also comparable to findings in the
literature.3,11 Both univariate and multivariable analyses sug-
gested that the use of mesh decreases the risk of developing a
recurrence by 3.6 times. This information suggests that um-
bilical hernias should be repaired using mesh, especially if a
patient has multiple comorbidities that are significantly asso-
ciated with recurrence, such as obesity, diabetes, liver dis-
ease, and ascites. Surgeons commonly consider the size of the
hernia defect when deciding whether to use mesh in a repair.

E4 JAMA Surgery Published online January 25, 2017 (Reprinted) jamasurgery.com

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017

 Factors Associated With Long-term Outcomes of Umbilical Hernia Repair Original Investigation Research

Of the 167 patients who underwent repair with mesh, 64 (38%)
did not have their defect size reported, 55 (33%) had a defect
of 2 cm or greater, and 48 (29%) had a defect size of less than
2 cm. Because there was no significant difference between
these groups, we are unable to conclude whether the size of
defects should play a role in a surgeon’s decision to use mesh.
Most studies evaluating umbilical hernia recurrences have
looked at short-term follow-up. Our median follow-up time was
8.5 years. In our study, 50% of recurrences occurred within the
first year, 75% within the first 5 years, and 25% after 5 years of
repair. This information implies that most studies reporting re-
currence rates are not capturing up to 25% of recurrences that
are occurring in the long term.
Interestingly, 193 (58%) patients who underwent an umbili-
cal hernia repair had other hernias that either were repaired be-
fore the index repair or developed postoperatively. Therefore,
we propose that umbilical hernias may be a type of “field defect,”
and we support the idea that abnormal collagen metabolism
could play a role in hernia development.9,10 We found that on uni-
variate analysis, patients who had a history of other hernias (n
= 125) were less likely to have umbilical hernia recurrences (P
= .03). We speculate that surgeons might be more inclined to use
mesh in patients with a history of other hernias. Seventy-six (61%)
patients with other hernias received mesh, while only 87 (44%)
patients without prior hernias received mesh (P = .003). The use
of mesh in patients who are prone to develop hernia may pro-
tect them from recurrence.
Our study had a low complication rate of 18%, despite the
50% of patients who underwent repair with mesh. If seromas
are excluded, the total complication rate was only 11.7%, which
is comparable to the rate in the literature.12,13 In 332 patients,
we saw a low surgical-site infection rate of 6.6% (n = 22) as well
as a low mesh infection rate of 0.9% (n = 2). There were no cases
of mesh causing fistulas in our long-term follow-up. There was
no significant difference in the complication rate (of which
most are seromas) between those who had mesh repair and
those who did not. In view of its low risks and high benefits,
mesh repair should be considered in most cases of umbilical
hernia repair, regardless of defect size.
Our study had a 27% mortality rate, at a mean of 5.1 years
after index repair, for reasons that are unclear. Because the
mean age of our patients was 58.4 years, this rate suggests that
27% of patients were dying by age 63 years. Expectedly, age,
smoking, and liver disease increased the odds of mortality in
this patient cohort. There was only 1 death within 30 days, for
a 30-day mortality of 0.3%.
A surprising finding was that patients who underwent
emergency or semi-urgent repair had a 2.2 times increased odds
of death. Twenty deaths occurred more than 30 days after
emergency or semi-urgent repair, and 12 (60%) of these deaths
were secondary to unknown causes and 5 (25%) were second-
ary to liver disease. Although there is a trend toward higher
mortality rates in patients who underwent emergency repair,
it is difficult to interpret whether the deaths were related to
the emergency or to underlying medical conditions given that
the etiology of the majority of these deaths is unknown.
Limitations
A major limitation of this study was that patients had vari-
able follow-up periods. For example, patients studied in 1998
had 16 years of follow-up, while those who had operations in
2008 had only 6 years of follow-up. To assess the severity of
this bias, we ran sensitivity analyses that included only pa-
tients who strictly met a 7-year follow-up period. We found that
there was minimal bias in our multivariable mortality model,
with no changes in the sign or significance of covariates. In our
recurrence model, however, the important covariates lost sig-
nificance because of the smaller sample size used in the sen-
sitivity analysis, and we could not retain the major findings of
our study while applying the model.
There were several other limitations to the study. First, we
were able to note recurrences only if they were described in
the CPRS. Second, we were unable to accurately account for
the size of the hernias because of inconsistent recording in the
operative notes. Third, we studied a cohort composed of mostly
male veterans, so these results may not be applicable to other
populations. Fourth, the causes of death of the majority of the
deceased patients in our cohort were unknown. Therefore, it
was difficult to know if their deaths were related to operative
complications.
Conclusions
Umbilical hernia repairs in our cohort study of 332 mostly
male military veterans reveal that ascites, liver disease, dia-
betes, obesity, and primary suture repair without mesh are
significantly associated with an increased rate of recurrence.
Mesh repairs should be considered in these patients, if not in
all patients, given that there was no significant difference in
the complication rates between mesh repair and primary
suture repair. In this cohort, there was a 27% long-term
death rate, at a median of 5 years, from index repair opera-
tion for reasons that are unclear. Future research looking
into the mortality rates in a matched cohort comparing
patients with umbilical hernia may help elucidate whether
the death rates in our cohort are different from the rates in
the general population.

ARTICLE INFORMATION
Accepted for Publication: October 28, 2016.
Published Online: January 25, 2017.
doi:10.1001/jamasurg.2016.5052
Author Contributions: Dr Sanchez and Ms Shankar
had full access to all the data in the study and take
responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Shankar, Itani, Sanchez.
Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Shankar, Sanchez.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: O'Brien, Sanchez.
Administrative, technical, or material support: Itani,
O'Brien, Sanchez.
Conflict of Interest Disclosures: None reported.
Funding/Support: This work was funded by the
VA Healthcare System.
Role of the Funder/Sponsor: The funding source
had no role in the design and conduct of the study;
collection, management, analysis, and
interpretation of the data; preparation, review, or

jamasurgery.com (Reprinted) JAMA Surgery Published online January 25, 2017 E5

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017

 Research Original Investigation Factors Associated With Long-term Outcomes of Umbilical Hernia Repair

approval of the manuscript; and decision to submit
the manuscript for publication.
Additional Contributions: The authors
acknowledge the VA Healthcare System for funding
the design and conduct of this study and the
interpretation of the data.
REFERENCES
1. Asolati M, Huerta S, Sarosi G, Harmon R, Bell C,
Anthony T. Predictors of recurrence in veteran
patients with umbilical hernia: single center
experience. Am J Surg. 2006;192(5):627-630.
2. Muschaweck U. Umbilical and epigastric hernia
repair. Surg Clin North Am. 2003;83(5):1207-1221.
3. Aslani N, Brown CJ. Does mesh offer an
advantage over tissue in the open repair of
umbilical hernias? a systematic review and
meta-analysis. Hernia. 2010;14(5):455-462.
4. Deysine M. Ventral herniorrhaphy: treatment
evolution in a hernia service. Hernia. 1998;2(1):
15-18. doi:10.1007/BF01207768
5. Lau H, Patil NG. Umbilical hernia in adults. Surg
Endosc. 2003;17(12):2016-2020.
6. Bingener J, Buck L, Richards M, Michalek J,
Schwesinger W, Sirinek K. Long-term outcomes in
laparoscopic vs open ventral hernia repair. Arch Surg.
2007;142(6):562-567.
7. Schumacher OP, Peiper C, Lörken M,
Schumpelick V. Long-term results after Spitzy’s
umbilical hernia repair [in German]. Chirurg. 2003;
74(1):50-54.
8. Venclauskas L, Silanskaite J, Kiudelis M.
Umbilical hernia: factors indicative of recurrence.
Medicina (Kaunas). 2008;44(11):855-859.
9. Winsnes A, Haapamäki MM, Gunnarsson U,
Strigård K. Surgical outcome of mesh and suture
repair in primary umbilical hernia: postoperative
complications and recurrence [published online
February 15, 2016]. Hernia. 2016;20(4):509-516.
10. Junge K, Klinge U, Rosch R, et al. Decreased
collagen type I/III ratio in patients with recurring
hernia after implantation of alloplastic prostheses.
Langenbecks Arch Surg. 2004;389(1):17-22.
11. Arroyo A, García P, Pérez F, Andreu J, Candela F,
Calpena R. Randomized clinical trial comparing
suture and mesh repair of umbilical hernia in adults.
Br J Surg. 2001;88(10):1321-1323.
12. Dalenbäck J, Andersson C, Ribokas D,
Rimbäck G. Long-term follow-up after elective adult
umbilical hernia repair: low recurrence rates also
after non-mesh repairs. Hernia. 2013;17(4):
493-497.
13. Iversen E, Lykke A, Hensler M, Jorgensen LN.
Abdominal wall hernia repair with a composite
ePTFE/polypropylene mesh: clinical outcome and
quality of life in 152 patients. Hernia. 2010;14(6):
555-560.

E6 JAMA Surgery Published online January 25, 2017 (Reprinted) jamasurgery.com

Copyright 2017 American Medical Association. All rights reserved.

Downloaded From: http://archsurg.jamanetwork.com/pdfaccess.ashx?url=/data/journals/surg/0/ by a University of California - San Diego User on 01/25/2017