118
Table 2. Outcome measures of the two operations
OR time
(min)
EBL
(ml)
Hosp
(days)
Mesh
(sq cm)
Days to PO
Laparoscopic
Open
124 ± 64a
78 ± 41a
68 ± 42
168 ± 145
5 (1–33)
5.5 (2–30)
495 ± 343b
97 ± 69b
1 (1–22)
2.5 (1–22)
OR, operating room; EBL, estimated blood loss; Hosp, hospitalization time; PO, oral intake
b p ס
0.0030, Mann-Whitney rank sum test
mesh was routinely sequestered from the bowel with omentum if the peri-
toneum could not be closed. Drains were used selectively.
therefore that a laparoscopic repair would simplify the op-
eration and result in a shorter recovery. Our results, con-
trolled by unbiased case matching, shows that this is not the
case. We did not, however, look at other outcome measures
such as total and itemized cost or time off work.
The laparoscopic repair was also performed under general anesthetic.
Pneumoperitoneum was induced via Veress needle puncture. Four to six
ports were placed as far laterally as possible from the defect. Angled (30°
and 45°) laparoscopes were used routinely. Intraabdominal lysis of adhe-
sions was done as needed. No attempt was made to resect the hernia sac.
After reduction of the hernial contents and identification of the hernial
defect, an ePTFE (Dualmesh; W. L. Gore, Flagstaff, AZ) sized for a 4-cm
overlap of the defect was inserted through a 10- or 12-mm port and secured
to the anterior abdominal wall using titanium screws (Origin, Menlo Park,
CA, USA). Three concentric circles of screws, placed 3–4 cm apart, were
placed. Anchoring sutures, now routine in many techniques, were not used
in those early cases. No drains were used.
We have identified two factors that may contribute to
this unexpected conclusion. The first problem is an inad-
equate sample size, which, as in many surgical trials, is
often a major reason for detecting no difference when a true
difference exists (the beta error). Our series, which con-
sisted of operations performed in a community hospital,
clearly suffers from this disadvantage. Thus, sampling error
may explain the negative results. A larger comparison study
[4], using retrospective controls and combining 56 patients
from two centers, arrived at a totally different conclusion.
The second factor is the lack of patient selection. Com-
paring the two operations in a consecutive series without
selection may yield heterogeneous results influenced by
chance. When all patients are considered for the laparoscop-
ic operation, the hazards of the technique, particularly in
dissection of complex adhesions, can lead to major compli-
cations. It only takes one of two major complications to
wipe out the potential benefits for the entire series. This
explanation is consistent with the results reported by Holz-
man et al. [2], who found a 25% incidence of prolonged
hospitalization in their series of 21 laparoscopic incisional
hernia repairs.
Results
Table 2 lists the outcome measures compared. The operat-
ing time was 40% longer for the laparoscopic operation (p
ס
0.0387). Neither operative blood loss nor hospitalization
time were significantly different. The mesh size was larger
for the laparoscopic operation—not so much a reflection of
the size of the hernia as a reflection of differences in tech-
nique. There was no difference in the time to resumption of
oral intake following both procedure.
Major complications occurred in both groups. In the
open group, one patient had an enterotomy, which was re-
paired without sequelae; another patient had a prolonged
ileus with atelectasis and pneumonitis, requiring ventilatory
management.
Based on the observations in this study, we believe that,
just as in all laparoscopic operations, there are circum-
stances that favor the laparoscopic approach over the open,
and vice versa. Therefore, patient selection is critical for
good results.
In the laparoscopic group, two patients had enteroto-
mies. One of these patients ultimately required mesh re-
moval due to infection, and another had a prolonged post-
operative course with respiratory failure and sepsis. No
deaths occurred in either group.
Since complex adhesions were the main cause of non-
optical results in this series, we recommend that the lapa-
roscopic approach be abandoned and the procedure con-
verted to an open operation as soon as this problem is dis-
covered on initial examination. Additionally, based on this
experience, we conclude that mesh placement in a patient
with enterotomy is ill-advised. Patients who developed in-
cisional hernias after an uncomplicated abdominal opera-
tion, or who had no history of peritonitis, usually had few
adhesions and were thus good candidates for the laparo-
scopic technique. By the same token, patients who have
small defects and are not obese should be repaired with the
open method since this method will result in as fast a re-
covery as the laparoscopic operation, without the added
expenses associated with the high-tech procedure.
Discussion
To adequately repair any incisional hernia, certain basic
principles have long been recognized [1]. The defect must
be completely defined, the adhesions must be separated, and
the repair must be done without tension. If a prosthetic patch
is to be used, it should be placed beneath the plane of the
fascial defect [1, 7], and the size of the patch should be
larger than the hernial orifice.
Experience with the laparoscopic repair has shown that
the anatomic defect of the incisional hernia can be visual-
ized readily from inside the abdomen, and the adhesions are
often—but not always—few and easy to divide. With ap-
propriate instrumentation, a large patch can be secured onto
the abdominal wall for a tension-free repair. It would appear
Acknowledgment. This work was supported by departmental research
funds.