Fungal Isolates in Peritoneal Fluid Culture Do Not Impact Peri-Operative Outcomes of Peptic Ulcer Perforation

Abstract

Background:

Fungal isolates from peritoneal fluid sampling in patients with perforated peptic ulcer (PPU) is not uncommon and its management unclear. This study aims to evaluate whether the presence of fungus in peritoneal fluid cultures is a predictor of morbidity and mortality after laparotomy for PPU.

Method:

This is a single-center retrospective study including adult patients with perforated gastric and duodenal ulcers over a 10-year period (January 2004 to January 2014). Evaluation of predictors contributing to fungal growth was conducted using multiple logistic regression analysis. Operative factors and 30-day mortality and morbidity outcomes were compared against fungal growth using a multivariable generalized linear mixed model analysis.

Results:

The median age was 58 (interquartile range [IQR] 44–70) years with 110 (20.3%) females. In addition to hypertension and hyperlipidemia, diabetes mellitus (13.5%), ischemic heart disease (2.6%), and heart failure (2.4%) were common. Fungus was cultured from peritoneal fluid in 209 (38.6%) patients. Median American Society of Anesthesiology (ASA) score was 2 (IQR 2–3) and median Mannheim peritonitis index (MPI) score was 15 (IQR 10–20). Free air was detected in 323 (59.6%) patients and 52 (9.6%) patients had gastrectomy. Median length of stay was 7 (IQR 6–11) days. All-cause complications were seen in 53 (9.8%) patients, of whom 37 patients (6.8%) developed intra-abdominal collection, 20 patients (3.7%) had anastomotic leakage, and 12 patients (2.2%) required repeat operation. Thirty-day mortality was seen in 47 (8.7%) patients. Multivariable analysis showed age (median age, 64; IQR 53–74) as a predictor of fungal growth (p < 0.001) but fungal growth not a predictor of adverse peri-operative outcomes.

Conclusion:

Fungal isolates in peritoneal fluid cultures are more likely to occur in older patients who have PPU. Presence of fungal isolates does not impact peri-operative outcomes.

Peptic ulcer disease has a high prevalence. Bleeding and perforation are two commonly encountered complications. Perforated peptic ulcer (PPU) is a surgical emergency and omental patch repair is the most common procedure performed. Despite advances in critical care, understanding of sepsis and surgical techniques, peri-operative morbidity (20%–50%) and mortality (1.3%–20%) remain high [1 –6], with sepsis being the predominant contributor. At outset, PPU is a sterile phenomenon with gastric acid causing chemical peritonitis. With time, secondary infection with sepsis ensues [7]. This is supported by evidence reporting worse outcomes in patients with delayed presentation [5,7]. Hence, it is likely that the peritoneal fluid bacteriology evolves with time. Perforated peptic ulcer is classified as contaminated or dirty-infected surgical procedure and empiric antibiotic agents are prescribed routinely with therapeutic intent. In patients who have PPU, peritoneal fluid culture is sampled and is predominant source of microbial isolation. Evidence-based standard clinical practice recommends changing empirical antibiotic agents when the culture reports are available. It remains the duty of acute care surgeons to follow local antibiogram and antibiotic stewardship program recommendations [8,9].

Although the management guidelines for bacterial isolate are clear, fungal isolates from the peritoneal fluid create a clinical dilemma with regard to anti-fungal therapy. Fungal isolates could represent oral commensal flora and it is unclear if antifungal therapy should be administered to patients with fungal isolates. Furthermore, not all patients undergo fungal testing of the peritoneal fluid. Fungal growth in peritoneal fluid cultures has been associated with poorer surgical outcomes [10 –16]. Current mortality risk prediction models of PPU do not include positive fungal isolates as part of the criteria [10]. It is unknown whether positive fungal cultures are merely an association or a predictor of poorer outcomes. In 2003, a retrospective study by Shan et al. [11] including 145 patients with PPU reported 43.4% incidence of positive fungal isolates and that it was an adverse risk factor for surgical site infection, length of hospital stay, and mortality. In a matched retrospective study from Taiwan, Li et al. [12] reported that antifungal therapy did not improve outcomes including 30-day all-cause mortality in patients and it could be reserved for patients who are critically ill and/or severely immunocompromised. In 2018, a retrospective study by Horn et al. [13] that included 107 patients with PPU reported 25.2% administration of pre-operative antifungal therapy; 30.8% patients having peritoneal fluid cultures tested of which 51.5% were tested positive for fungus. The presence of fungal isolates did not impact length of stay, intensive care unit stay, ventilator days, 30-day re-admission rates, or intra-abdominal abscess formation. Routine pre-operative antifungal therapy for all patients with PPU is not evidence based and we do prescribe anti-fungal therapy selectively. Because of conflicting reports, more evidence is needed. Our study aims to evaluate whether the presence of fungus in peritoneal fluid cultures is a predictor of poor surgical outcomes. We hypothesize that fungus is a contaminant of normal oral microflora, and a positive peritoneal fungal isolate should not affect surgical outcomes of PPU.

Patients and Methods

A single-center retrospective study was conducted and included all adult patients with perforated gastric and duodenal ulcers over a 10-year period (January 2004 to January 2014). Electronic medical records were retrieved for International Statistical Classification of Disease (ICD)-9 codes: 531.10, 532.10, 567.21, 567.29, 531.50, 532.50, 151.9 for primary or secondary diagnosis of PPU, and ICD 10 codes (used in the hospital since January 1, 2012): gastric ulcer (coded K25, subcategories 25.1, 25.2, 25.5, and 25.6) or duodenal ulcer (coded K26, subcategories 26.1, 26.2, 26.5, and 26.6). Moribund patients, patients who did not undergo surgery, patients in whom PPU could not be confirmed intra-operatively, and patients with perforation of other organs were excluded from analysis.

An intra-operative peritoneal fluid sample is taken at the discretion of the operating surgeon. During weekday office hours, the sample is processed immediately by the microbiology laboratory. Over weekends and after office hours, the sample is processed at the next immediate available opportunity. The specimens obtained were processed for gram stain, bacterial cultures (aerobic an anaerobic), and fungal smears. Isolates were identified based on VITEK 2 system (bioMérieux, Marcy l'Etoile, France). All isolates were sub-cultured on to CHROMagar Candida (Becton Dickinson, Franklin Lakes, NJ). We did not record the type of Candida species separately, however, we have isolated Candida albicans, Candida tropicalis, Candida glabrate, and Candida parapsilosis. Routine anti-fungal susceptibility testing is not done and is additionally processed at the discretion of clinician. All patients who have yeast isolated and are admitted to intensive care units are managed by antimicrobial stewardship programs and routinely intravenous fluconazole 200–400 mg per 24 hours is prescribed depending on body weight and creatinine clearance. Routine anti-fungal agents are not prescribed to patients admitted to general wards. In patients who deviate from the expected norms of recovery, abdominal imaging was performed to check for any intra-abdominal collections that in turn would be drained and fluid tested for culture and sensitivity. Intravenous anidulafungin is reserved to be prescribed strictly by an infectious disease specialist.

Statistical analysis

Normal distribution of continuous variables was determined using the Shapiro-Wilk normality test. Continuous variables were presented as mean ± standard deviation or median with interquartile range (IQR) where applicable. Statistical analysis was performed with the Student t-test or Mann-Whitney U test for continuous variables where appropriate and Pearson χ² test for nominal variables to look for comparability in both groups. Evaluation of predictors contributing to fungal growth was conducted using multiple logistic regression analysis. Operative factors and 30-day mortality and morbidity outcomes were compared against fungal growth using a multivariate generalized linear mixed model analysis. Variables with p ≤ 0.1 in the univariable analysis were included in the multiple or multivariable model where applicable. All statistical analysis was performed using SPSS version 23.0 (SPSS Inc., Chicago, IL). The level of statistical significance was set at 0.05.

Results

A total of 673 patients were managed for PPU over the study duration. Seven patients with missing data and 124 patients who did not have abdominal fluid sent for culture were excluded. The remaining 542 patients were included in the final analysis. The median age was 58 years (IQR 44 − 70) with 110 females (20.3%). Diabetes mellitus (13.5%), ischemic heart disease (2.6%), and heart failure (2.4%) were the most common comorbidities in addition to hypertension and hyperlipidemia. Fungus was cultured from peritoneal fluid in 209 patients (38.6%).

Regarding operative factors and outcomes, the median American Society of Anesthesiology (ASA) score was 2 (IQR 2–3) while median Mannheim peritonitis index (MPI) score was 15 (IQR 10–20). Free air was demonstrated on radiography in 323 patients (59.6%) and 52 patients (9.6%) required gastrectomy. Median length of stay was 7 days (IQR 6–11). All-cause complications were seen in 53 patients (9.8%) patients, of whom 37 (6.8%) developed intra-abdominal collections, 20 (3.7%) had anastomotic leakage, and 12 (2.2%) required repeat operation. After incident operation, 30-day mortality was seen in 47 patients (8.7%).

On univariable analysis, older age (p < 0.01) female gender (p < 0.01), use of steroids (p = 0.03), chronic kidney disease (p < 0.01), low hemoglobin (p < 0.01), elevated creatinine (p < 0.01), low albumin (p = 0.09) at admission, MPI score ≥20, ASA score ≥2 (p < 0.01), and higher physiology score (p < 0.01) predicted presence of fungal isolates. On multivariable analysis, only age remained a predictor for fungal growth (Table 1).

Table 1.

Demographic Predictors of Fungal Growth

Demographics	No (%)			Univariable p value	Multiple logistic regression
Demographics	Fungal (n = 209)	No fungal (n = 333)	Total (n = 542)	Univariable p value	p	OR	95% CI
Age¹	64 (53.0–74.3)	53 (40.0–66.4)	58 (44–70)	<0.001	<0.001	1.031	1.01–1.047
Albumin¹	34 (34–34)	34 (34–34)	34 (34–34)	0.091	0.471	1.017	0.972–1.064
ASA score (>2)	96 (46.3)	107 (32.4)	203 (37.5)	0.001	0.487	0.851	0.540–1.342
Cardiac failure	7 (3.3)	7 (2.1)	14 (2.6)	0.373
Cirrhosis	2 (1.0)	3 (0.9)	5 (0.9)	0.947
COPD	4 (1.9)	7 (2.1)	11 (2)	0.880
Creatinine¹	97 (80–146)	88 (76–109)	91 (78–120)	<0.001	0.054	1.004	1.000–1.007
CKD	10 (4.8)	3 (0.9)	13 (2.4)	0.007	0.361	1.940	0.457– 8.048
DM	35 (16.7)	38 (11.4)	73 (13.5)	0.077	0.913	0.970	0.558–1.684
Free air on CXR	134 (64.1)	189 (56.8)	323 (59.6)	0.212
Gender				<0.01	0.103	1.593	0.911– 2.787
Male	150 (71.8)	282 (84.7)	432 (79.7)
Female	59 (28.2)	51 (15.3)	110 (20.3)
Haemoglobin¹	13.1 (10.8–15.2)	14.1 (12.4–15.2)	14.0 (11.8–15.2)	0.001	0.387	0.967	0.897–1.043
History of gastrectomy	1 (0.5)	5 (1.5)	6 (1.1)	0.266
History of NSAIDs	10 (4.8)	10 (3)	20 (3.7)	0.284
History of steroids	6 (2.9)	2 (0.6)	8 (1.5)	0.033	0.176	3.230	0.592–17.622
History of smoking	66 (31.6)	108 (32.4)	174 (32.1)	0.836
History of ulcer disease	18 (8.6)	30 (9)	48 (8.9)	0.874
IHD	14 (6.7)	14 (4.2)	28 (5.2)	0.202
MPI score (≥ 20)	77 (36.8)	74 (22.2)	151 (27.9)	<0.001	0.532	0.839	0.484–1.455
Operative score¹	19 (15–19)	19 (15–19)	19 (15–19)	0.274
Physiology score¹	20 (16–26)	17 (15–23)	18 (15–24)	<0.001	0.608	0.989	0.949–1.031
Pre-operative shock
SBP <91mm Hg	16 (7.7)	18 (5.4)	34 (6.3)	0.293
SBP <100 mm Hg and HR >100	16 (7.7)	19 (5.7)	35 (6.5)	0.369
SIRS criteria (/2)	69 (33)	89 (26.7)	158 (29.2)	0.117
Presentation (>24 h)	108 (51.7)	159 (47.8)	267 (49.3)	0.373
Urea¹	6.4 (4.4–10.6)	5.1 (3.7–7.1)	5.500 (3.900–9.025)	<0.001	>0.999	1.000	0.967–1.034
WBC¹	11.1 (7.8–15.7)	12.0 (8.6–15.8)	11.7 (8.3–15.8)	0.174

All descriptive statistics presented as number (percentage). P value calculated using Pearson χ² test unless otherwise indicated.

Median values presented (interquartile range), p value calculated using Mann-Whitney U test

ASA = American Society of Anesthesiologists score; COPD = chronic obstructive pulmonary disease; CI = confidence interval; CKD = chronic kidney disease; DM = diabetes mellitus; CXR = chest radiograph; HR = heart rate; NSAIDs = non-steroidal anti-inflammatory drugs; IHD = ischemic heart disease; MPI = Mannheim peritonitis index; OR = odds ratio; SBP = systolic blood pressure; SIRS = systemic inflammatory response syndrome; WBC = total white blood cell count.

On univariable analysis, peritoneal fluid fungal isolates contributed to the need for gastrectomy (versus omental patch repair) (p < 0.01), surgical intensive care unit (SICU) admission (p < 0.01), ulcer size larger than 10 mm (p < 0.01), total length of hospital stay more than 9 days (p < 0.01), mortality (p = 0.01), intra-abdominal collection (p < 0.01), and re-operation (p = 0.04). On multivariable analysis, none of the outcomes demonstrated significance (Table 2).

Table 2.

Operative Factors and Surgical Outcomes Affected by Presence of Fungal Growth

Operative	No (%)			Univariable p value	Generalized linear mixed model
Operative	Fungal (n = 209)	No fungal (n = 333)	Total (n = 542)	Univariable p value	p	OR	95% CI
Access				0.224
Open	207 (99.0)	325 (97.6)	532 (98.2)
Laparoscopic	2 (1.0)	8 (2.4)	10 (1.8)
Blood Transfusion	20 (9.6)	28 (8.4)	58 (10.7)	0.643
Intra-abdominal collection	24 (11.5)	13 (3.9)	37 (6.8)	0.001	0.157	0.614	−0.238–1.465
Leakage	11 (5.2)	9 (2.7)	20 (3.7)	0.124
Mortality (30 d)	26 (12.4)	21 (6.3)	47 (8.7)	0.014	0.667	0.172	−0.613–0.957
Re-operation	8 (3.8)	4 (1.2)	12 (2.2)	0.043	0.969	−0.029	−1.467–1.409
SICU stay	54 (25.8)	62 (18.6)	116 (21.4)	<0.01	0.137	0.412	−0.132–0.955
Size of ulcer (mm)¹	10 (5–15)	5 (5–10)	7 (IQR5–10)	<0.01	0.089	0.022	−0.003–0.047
Total length of stay (d)¹	9 (6–18)	7 (6–10)	7 (IQR6–11)	<0.01	0.576	0.004	−0.010–0.017
TPN	31 (14.8)	26 (7.8)	57 (10.5)	0.247
Type of surgery				0.002	0.328	0.375	−0.377–1.126
Gastrectomy	30 (14.4)	22 (6.6)	52 (9.6)
Omental patch	172 (82.3)	305 (91.6)	477 (88.0)

All descriptive statistics presented as number (percentage), P value calculated using Pearson χ² test unless otherwise indicated.

Median values presented (IQR), P value calculated using Mann-Whitney U test.

IQR = interquartile range; SICU = surgical intensive care unit; TPN = total parenteral nutrition.

Discussion

Our study has shown that older age is a predictor of fungal isolates from peritoneal fluid and such isolates do not impact peri-operative outcomes. Fungi are considered natural commensal organisms of the gastrointestinal tract, but in the setting of PPU, it remains unknown whether fungal growth in peritoneal fluid cultures represents a contaminant or a true infection. Previous studies have demonstrated that Candida species in the peritoneal cavity is cleared without complications in most healthy patients and antifungal therapy is not indicated as the mere presence of fungus does not affect outcome [14,15]. However, there are an increasing number of studies reporting associations between fungal growth in peritoneal fluid and poorer clinical outcomes [10,16 –18]. There are also suggestions that only nosocomial fungal peritonitis is associated with higher mortality but not community-acquired fungal peritonitis (19). Regardless, the clinical significance of fungal growth in peritoneal fluid remains unclear. The Surgical Infection Society and the Infectious Disease Society of America recommends antifungal therapy in only clinically severe cases and their recommendations are not based on strong evidence (20, 21). Our study concludes that on multivariate analysis, presence of fungal isolates does not impact morbidity or mortality in PPU patients managed with emergency surgery at our institution.

In our study, 81.6% (524/664) patients had their peritoneal fluid sampled, and this is higher compared with the 30.8% reported by Horn et al. [13]. This is significant because blood cultures may not have been done or it may not detect bacteria. In our experience of managing patients with pyogenic liver abscesses, we have shown that blood culture is positive in only 41.6% of patients (22). Hence, if body fluids are not subjected to microscopy routinely in patients with suspected sepsis, microbial therapy may not truly be optimized. The incidence of fungal isolates in peritoneal fluid in our study is consistent with other reports [15,19]. We have validated several PPU risk prediction models previously against our cohort. (23) The factors affecting mortality (female gender, liver cirrhosis, SICU stay, ASA score >2) demonstrate statistically significant association with fungal growth on univariable analysis but not on multivariable analysis. In our study, patients with fungal isolates were generally 10 years older than those without fungal isolated and only older age was associated with fungal isolates. This is interesting and could be because of various reasons. Older patients may have a delayed presentation to hospital, are more likely to be taking long-term medications, have increased prevalence of diabetes mellitus, or may have a larger ulcer. All of these factors could increase the possibility of oral commensal flora to egress through the PPU and gain access to the peritoneal cavity. There was no detectable difference regarding delayed presentation to hospital or association with diabetes mellitus. However, our study did show a tendency for elevated creatinine and larger size of ulcers in patients with fungal isolates. Both large size and elevated creatinine are associated with poor outcomes in patients with PPU [7]. It has been shown that patients with fungal isolates had a tendency toward longer SICU stay (p = 0.05) [18]. Also, more patients with fungal isolates required SICU admission and the length of hospital stay was longer for patients with fungal isolates, however, this was not statistically significant on multivariable analysis. It is known that immunocompromised patients are predisposed to developing fungal infections and in the present study, steroid use was significant on univariable analysis but not on multivariable analysis [24,25]. This could be the result of small sample size because only eight patients (1.5%) were taking steroids.

The larger the ulcer, the more likely the consideration of resection surgery because of fear of omental patch leak (5). Gastric resection surgeries are associated with increased morbidity and mortality compared with classic omental patch repair [23,26]. In the current study, gastrectomy was more likely required in patients with fungal isolates, however, this was not significant on multivariable analysis. This is consistent with other studies [27 –29]. In the local context, this could be explained by lower comorbidity, fewer patients with pre-operative shock, and fewer patients with resection surgery.

The strengths of our study are large database, a decade of experience, and a liberal policy of sending peritoneal fluid for culture. In addition, multivariable analysis was done to reduce confounding variables. The weakness of our study is that this was conducted in a single center, there were seven (1.9%) missing records, and 124 patients (18.4%) did not have peritoneal fluid sampled. In our opinion, the results are unlikely to be influenced by such weaknesses which is inherent of any retrospective study. Second, we have not made a distinction between community-acquired versus nosocomial PPU unlike some authors [12]. Last, we have not reported the type of bacterial or fungal isolates and their sensitivity pattern. Although our study does provide good evidence that fungal isolates do not impact outcomes; we are unable to answer other questions, such as patient selection for antifungal therapy, duration of antifungal therapy and type of antifungal agent. These answers can only rightfully be obtained by prospective controlled studies. With PPU being an acute pathology with emergency presentation, we foresee that conducting such studies will be challenging, and hence retrospective studies similar to ours are essential to enhance the understanding and evidence base on this topic. We also caution readers with generalization of our results as patient demography and clinical profile is varied not only across continents but also different institutions within same nation.

Conclusion

In conclusion, fungal isolates are more likely to occur in older patients, and outcomes of patients with PPU are not determined by the presence of fungal isolates.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Zelickson

, Bronder

, Johnson

, et al. Helicobacter pylori is not the predominant etiology for peptic ulcers requiring operation. Am Surg, 2011; 77:1054–1060.

Thorsen

, Glomsaker

, von Meer

, et al. Trends in diagnosis and surgical management of patients with perforated peptic ulcer. J Gastrointest Surg, 2011; 15:1329–1335.

Bae

, Shim

K-N

, Kim

, et al. Incidence and Short-term mortality from perforated peptic ulcer in Korea: A population-based study. J Epidemiol, 2012; 22:508–516.

Christensen

, Riis

, Norgaard

, et al. Short-term mortality after perforated or bleeding peptic ulcer among elderly patients: A population-based cohort study. BMC Geriatr, 2007; 7:8.

Boey

, Choi

, Poon

, Alagaratnam

. Risk stratification in perforated duodenal ulcers. A prospective validation of predictive factors. Ann Surg, 1987; 205:22–26.

Hermansson

, von Holstein

, Zilling

. Surgical approach and prognostic factors after peptic ulcer perforation. Eur J Surg, 1999; 165:566–572.

Chung

, Shelat

. Perforated peptic ulcer—An update. World J Gastrointest Surg, 2017; 9:1–12.

Sartelli

, Labricciosa

, Barbadoro

, et al. The Global Alliance for Infections in Surgery: Defining a model for antimicrobial stewardship-results from an international cross-sectional survey. World J Emerg Surg, 2017; 12:34.

Sartelli

, Chichom-Mefire

, Labricciosa

, et al. The management of intra-abdominal infections from a global perspective: 2017 WSES guidelines for management of intra-abdominal infections. World J Emerg Surg, 2017; 12:29.

10.

Katlana

, Vyas

, Jain

, et al. Incidence and significance of intra-operative peritoneal fluid fungal culture in patients of perforated peptic ulcers. J Med Sci Clin Res, 2017; 5:18987–18991.

11.

Shan

, Hsu

, Hsieh

, Sy

, Lee

, Lin

. Significance of intraoperative peritoneal culture of fungus in perforated peptic ulcer. The British journal of surgery. 2003; 90:1215–1219.

12.

, Lee

, Liu

. Antifungal therapy did not improve outcomes including 30-day all-cause mortality in patients suffering community-acquired perforated peptic ulcer-associated peritonitis with Candida species isolated from their peritoneal fluid. J Microbiol Immunol Infect, 2017; 50:370–376.

13.

Horn

, Coleoglou Centeno

, Rasane

, et al. Pre-operative anti-fungal therapy does not improve outcomes in perforated peptic ulcers. Surg Infect, 2018; 19:587–592.

14.

Rex

. Candida in the peritoneum: Passenger or pathogen?. Crit Care Med, 2006; 34:902–903.

15.

Montravers

, Leroy

, Eckmann

. Intra-abdominal candidiasis: It's still a long way to get unquestionable data. Intensive Care Med, 2015; 41:1682–1684.

16.

Carneiro

, Mavrakis

, Mylonakis

. Candida peritonitis: An update on the latest research and treatments. World J Surg, 2011; 35:2650–2659.

17.

Sandven

, Qvist

, Skovlund

, Giercksky

. Significance of Candida recovered from intraoperative specimens in patients with intra-abdominal perforations. Crit Care Med, 2002; 30:541–547.

18.

Bassetti

, Righi

, Ansaldi

, et al. A multicenter multinational study of abdominal candidiasis: Epidemiology, outcomes and predictors of mortality. Intensive Care Med, 2015; 41:1601–1610.

19.

Montravers

, Dupont

, Gauzit

, et al. Candida as a risk factor for mortality in peritonitis. Crit Care Med, 2006; 34:646–652.

20.

Solomkin

, Mazuski

, Bradley

, et al. Diagnosis and management of complicated intra-abdominal infection in adults and children: Guidelines by the Surgical Infection Society and the Infectious Diseases Society of America. Clin Infect Dis, 2010; 50:133–164.

21.

Pappas

, Rex

, Sobel

, et al. Guidelines for treatment of candidiasis. Clin Infect Dis, 2004; 38:161–189.

22.

Shelat

, Chia

, Yeo

, Qiao

, Woon

, Junnarkar

. Pyogenic Liver Abscess: Does Escherichia Coli Cause more Adverse Outcomes than Klebsiella Pneumoniae?. World J Surg. 2015; 39(10):2535–42.

23.

Anbalakan

, Chua

, Pandya

, Shelat

. Five year experience in management of perforated peptic ulcer and validation of common mortality risk prediction models—Are existing models sufficient? A retrospective cohort study. Int J Surg, 2015; 14:38–44.

24.

Low

C-Y

, Rotstein

. Emerging fungal infections in immunocompromised patients. F1000 Med Rep, 2011; 3:14.

25.

Schulze

, Sonnenborn

. Yeasts in the gut: From commensals to infectious agents. Dtsch Arztebl Int, 2009; 106:837–842.

26.

Seow

, Lim

, Shelat

. Low serum albumin may predict the need for gastric resection in patients with perforated peptic ulcer. Eur J Trauma Emerg Surg, 2017; 43:293–298.

27.

Moller

, Adamsen

, Thomsen

, Moller

. Preoperative prognostic factors for mortality in peptic ulcer perforation: A systematic review. Scand J Gastroenter, 2010; 45:785–805.

28.

Noguiera

, Silva

, Santos

, et al. Perforated peptic ulcer: Main factors of morbidity and mortality. World J Surg, 2003; 27:782–787.

29.

Kocer

, Surmeli

, Solak

, et al. Factors affecting mortality and morbidity in patients with peptic ulcer perforation. J Gastroenterol Hepatol, 2007; 22:565–570.