Page 43 - World Journal of Laparoscopic Surgery
P. 43
Use of Laparoscopic vs Open Repair for Perforated Peptic Ulcers
hospital network. IRB approval was obtained before proceeding 63; p = 0.394), NSAID use (30.8 vs 28.7%; p = 0.916), PPI use (21.2
with the study. vs 12.6%; p = 0.192), previous surgery (44.9 vs 33.7%; p = 0.210).
Adult patients who presented with PPU and underwent Patient comorbidities such as COPD (13.5 vs 12.6%; p = 1.000), CHF
surgery were divided into two groups—laparoscopic or open (5.77 vs 8.38%; p = 0.768), CKD (3.85 vs 11%; p = 0.738) were also
repair. Patients who underwent laparoscopic converted to open not significant in determining the surgical approach (Reference
repair were considered laparoscopic on an intention-to-treat Table 1 for complete patient demographic data). In terms of the
basis, but were included in the open group for outcome analysis. clinical presentation of the patient (Table 2), 36.1% of patients
Patients with iatrogenic bowel injury or those who developed a (n = 79) were deemed septic upon presentation based on SIRS
perforated peptic ulcer during an unrelated hospital admission criteria; however, this was not a statistically significant factor
were excluded. in determining the operative approach (28.8% laparoscopic vs
We assessed several surgeon specific factors that could 38.3% open; p = 0.281). Very few patients (n = 9) presented with
influence the surgical approach chosen. This included the surgeon’s hypotension and only 16 patients presented with an abnormal
level of training (residency only vs fellowship training), surgeon’s INR above 2—factors not found influential in choosing the type
graduation year, and hospital level of care (i.e., tertiary referral of repair (5.7 vs 10.2% open; p = 0.531).
center vs community hospital). The level of training was determined A subset of factors in Table 2 were circumstantial factors
based on the information provided in the hospital credentialing relating to the case that may have an influence on the surgeon’s
information system. MIS fellowship trained surgeons were analyzed operative choice. These factors did not relate specifically to
as a subgroup of the fellowship training category. the characteristics of the patient or surgeon and included the
Circumstantial factors evaluated include the time of time of day, location of the ulcer, and preoperative imaging
presentation, ability to localize the perforation on imaging, and time localization. Intraoperative ulcer sites were found 42.9% of the
of diagnosis. Patient factors included vital signs in the emergency time in the stomach (28.8% laparoscopic, 47.3% open) and 57.1%
department, medical comorbidities, and preoperative labs. Patients in the duodenum (71.2% laparoscopic, 52.7% open). Of the 52
who met the criteria for sepsis or systemic inflammatory response laparoscopic repairs of PPU, 37 were found to be in the duodenum
syndrome on presentation were categorized as being septic in which was statistically significant (71.2% laparoscopic; p = 0.029).
the statistical analysis. Outcome variables analyzed included LOS, PPU was localized on preoperative imaging (duodenal vs stomach)
complications, readmissions, and discharge disposition. in 59.4% of total cases (61.5% laparoscopic, 58.7% open); however,
Laparoscopic and open surgical approaches were compared its relation to operative planning was not found statistically
based on demographics, clinical characteristics, and lab variables significant. The time of diagnosis was 55.3% in the daytime defined
by using Chi-square, Fisher’s exact, and Wilcoxon rank-sum tests. as 7 am–7 pm (61.5% laparoscopic, 53.3% open; p = 0.838), but
Multiple logistic regressions were used to establish relevant this association was not statistically significant in determining the
associations and to calculate adjusted odds ratios, expressed as surgical approach.
odds ratios, and 95% confidence intervals. The patient presentation Surgeon specific characteristics were also analyzed in the
and surgeon specific variables were analyzed on an intention to laparoscopic and open groups on an intention to treat basis
treat basis. With regard to the outcome variables, the laparoscopic (Table 3). A total of 56 surgeries (25.6%) were performed by
surgeries that were converted to open were analyzed within the surgeons with MIS fellowship training and these surgeons were
open repair group. An additional analysis was performed that found to perform laparoscopic repair of PPU more frequently
excluded patients with missing laboratory values (n = 49). Variables (46.2% open vs 63.8% laparoscopic; p ≤0.0001). The median
included in this regression model were selected by forward stepwise year of residency graduation was 2006; however, the length
regression. All tests were two-tailed and statistical significance was of time the surgeon has been practicing was not found to be
defined as p <0.05. Statistical analysis was performed with the use significantly correlated with the surgical approach. The majority
of R software version 4.0.0 (Vienna, Austria). of surgeries performed were by surgeons who trained at tertiary
care centers rather than community hospital residencies
(64.8 vs 35.2%; p = 0.054), but training at a tertiary center alone
results was not correlated with surgical approach. The hospital level of
A total of 219 observations were included (52 laparoscopic, 167 care (community hospital vs tertiary care center) was relatively
open) that underwent surgical management of ulcer disease. evenly split, with 53% of surgeries being performed at community
There were a total of 77 unique surgeons in the data set. Of these, hospitals and 47% at tertiary care centers; however, the level of
25 surgeons were responsible for the 52 laparoscopic repairs care was not significant in the choice of a laparoscopic approach
performed. The maximum number of PPU repairs performed by (33 laparoscopic repairs in community hospitals vs 20 in tertiary
a single surgeon was 11 and the average was 2.84. The maximum centers).
number of laparoscopic repairs of PPU by a single surgeon was 5 As there were not many factors specific to the demographics
with an average of 2.08 laparoscopic repairs. or patient presentation that were clinically significant in choosing a
laparoscopic over open repair, an additional analysis was conducted
Intention to Treat Data Analysis that excluded patients with missing variables, namely, INR (n = 49)
Surgeries that started laparoscopic but converted open were and hypotension (n = 9) (Table 4). Since most patients presumably
analyzed on an intention to treat basis with respect to the patient had their coagulopathy or hypotension corrected before
demographic and presentation data. Overall, the groups were proceeding to the operating room, these factors were deemed
comparable in terms of their presentation and demographics clinically irrelevant. A new forward stepwise regression analysis
with no statistically significant factors distinguishing the was then conducted excluding the variables INR and hypotension.
laparoscopic and open repair groups. The median age of the The final model contained five relevant factors: BMI, comorbidities,
patients undergoing laparoscopic vs open surgery was (59.9 vs residency type—tertiary vs community, fellowship, MIS fellowship.
World Journal of Laparoscopic Surgery, Volume 15 Issue 1 (January–April 2022) 41