The role of biomarkers in predicting mortality in peptic ulcer perforation

Abstract

Background

Peptic ulcer perforation remains a life-threatening surgical emergency associated with considerable postoperative mortality. Identifying reliable preoperative predictors is crucial for risk stratification. In this study, we aimed to investigate the predictive value of C-reactive protein, albumin/creatinine ratio, defect size, and selected demographic variables for 30-day mortality following surgical treatment of peptic ulcer perforation.

Methods

A retrospective analysis was conducted on 154 patients who underwent emergency surgery for peptic ulcer perforation. The association between 30-day postoperative mortality and C-reactive protein, albumin/creatinine ratio, defect size, as well as selected demographic variables was evaluated. Univariate and multivariate logistic regression analyses were performed. ROC analysis was used to determine optimal cutoff values. p < 0.05 was considered statistically significant.

Results

The mean age of the patients was 56. 102 (66.2%) of the patients were male. The 30-day mortality rate in the study cohort was 16.9%. A CRP level greater than 88 mg/L (p < 0,001) and an albumin/creatinine ratio below 3.37 (p < 0.001) were identified as independent predictors of mortality. While a defect size greater than 5 mm was significant in univariate analysis, it did not retain significance in the multivariate model.

Conclusion

CRP and the albumin/creatinine ratio are easily accessible biomarkers that may support perioperative risk stratification in patients undergoing surgery for perforated peptic ulcer. Defect size alone may be insufficient to reliably reflect prognosis when systemic inflammatory markers are taken into account.

Introduction

Peptic ulcer disease (PUD) is a major global health problem affecting millions of individuals worldwide. It occurs when the protective mechanisms of the gastric or duodenal mucosa are overwhelmed by the corrosive effects of acid and pepsin, leading to tissue erosion and ulceration. While many cases of PUD can be managed conservatively with pharmacological therapy and lifestyle modifications, complications such as perforation are a medical emergency requiring immediate surgical intervention [1]. Perforated peptic ulcers (PPU) are particularly dangerous due to their potential to cause peritonitis, sepsis, and multiple organ failure. The high mortality rates associated with PPU, ranging from 10% to 30%, highlight the critical nature of this complication [2, 3]. These figures are especially concerning in vulnerable populations, including elderly patients and those with pre-existing comorbidities. Factors contributing to the high risk of death include delayed diagnosis, advanced age, the presence of comorbidities, and the physiologic stress of emergency surgery. Therefore, early diagnosis, rapid evaluation, and timely surgical treatment are crucial in improving outcomes in PPU patients [3].

Early recognition and timely surgical intervention are vital in the clinical management of PPU. Predicting postoperative prognosis, however, remains a complex and multifaceted challenge for healthcare professionals. This complexity arises from the interplay of numerous factors, including patient’s pre-existing medical conditions, the nature and extent of the surgical procedure, potential complications, and individual variability in the recovery process. The difficulty of forecasting postoperative outcomes is further increased by the dynamic nature of the recovery process. Patients may experience unexpected deterioration or rapid improvement, requiring continuous reassessment and adjustment of prognoses. Additionally, the increasing prevalence of comorbidities in an aging population and the emergence of more complex surgical interventions contribute to the difficulty of accurately predicting postoperative outcomes. Consequently, there is a growing need for advanced prognostic tools and models capable of integrating diverse data sources and providing more accurate, individualized predictions to support clinical decision-making during the critical postoperative period. Therefore, identifying reliable biomarkers and clinical parameters that can predict mortality is of critical importance for improving patient management and optimizing resource utilization [4].

Recent studies have shown that certain biochemical parameters, particularly C-reactive protein (CRP), albumin, and creatinine levels, may influence surgical outcomes due to their associations with inflammation and organ function [5, 6]. While CRP serves as a marker of the acute-phase response, increasing in cases of infection and tissue injury, albumin levels provide insight into the patient’s nutritional status and the extent of systemic inflammation. Creatinine, a marker of renal function, holds prognostic significance, particularly in elderly patients and those with comorbidities. When evaluated collectively, these biochemical markers offer a comprehensive assessment of the patient’s preoperative condition and potential risk factors. High CRP levels combined with low albumin and abnormal creatinine values may indicate an increased likelihood of postoperative complications and prolonged recovery periods. Consequently, surgeons and anesthesiologists can use these parameters to personalize perioperative management strategies, optimize patient care, and potentially improve surgical outcomes.

However, rather than evaluating these parameters individually, calculating ratios (e.g., the albumin/creatinine ratio) may create more powerful predictive models. Studies investigating the relationship between such ratios and mortality are limited, and research specifically focusing on the postoperative period following PPU surgery is scarce. The albumin/creatinine ratio, in particular, appears promising as a more comprehensive indicator of both renal function and overall health status [7].

In this study it is aimed to identify clinical and biochemical markers capable of predicting 30-day mortality in patients undergoing surgery for PPU. Specifically, the prognostic value of parameters such as CRP levels, albumin/creatinine ratio, and ulcer defect size was statistically analyzed. The findings are intended to contribute to early risk stratification and improve clinical management of patients with PPU.

Materials and methods

This single-center, retrospective study analyzed data from 154 patients who were diagnosed with and surgically treated for peptic ulcer perforation at a tertiary care facility between January 1, 2019, and December 31, 2024. The study was approved by the local ethics committee (Date: 30.06.2025, No: 2025/99). Patients aged 18 years and older with a diagnosis of peptic ulcer perforation confirmed clinically, radiologically (presence of free air), or intraoperatively, who underwent surgery and had complete clinical and laboratory data (including CRP, albumin, creatinine, ASA score, etc.) as well as at least 30 days of postoperative follow-up, were included. Laboratory parameters were consistently collected preoperatively for all patients, and all measurements, including C-reactive protein, were obtained within the final 8 h prior to surgery to ensure timing consistency. The analysis utilized blood parameter values obtained within the final 8 h preceding the operation. Importantly, the biochemical parameters in both centers were measured using the identical analyzer (Siemens Advia Centaur XP). Patients with perforation due to causes other than peptic ulcer, those with pre-existing chronic kidney disease, a history of immunosuppressive disorders, or a diagnosis of malignancy were excluded. The albumin/creatinine ratio was calculated by dividing the serum albumin concentration (g/dL) by the serum creatinine concentration (mg/dL). As both parameters were obtained from the same preoperative blood sample, the resulting value represents a dimensionless ratio used for prognostic comparison rather than a standardized biochemical index.

Demographic data, The American Society of Anesthesiologist (ASA) classifications, comorbidities (diabetes mellitus, hypertension, chronic obstructive pulmonary disease, coronary artery disease, history of cerebrovascular events), ulcer locations (prepyloric, duodenum, antrum, corpus), defect sizes (> 5 mm, 5–15 mm), laboratory values (CRP, albumin, creatinine), and length of hospital stay were obtained from the electronic patient records system. Mortality was defined as death from any cause occurring within 30 days postoperatively.

Statistical analysis

The sample size for this study was determined using a power analysis performed with G*Power 3.1 software. A statistical power (1–β) of 0.80 and a confidence level of 95% were assumed, resulting in a minimum required sample size of 22 patients per group. Accordingly, a total sample size of at least 44 patients was considered sufficient.

The normality of data distribution was assessed using the Shapiro–Wilk test. As most continuous variables did not show normal distribution, nonparametric tests were applied. Continuous variables were expressed as median and interquartile range (IQR) and compared using the Mann–Whitney U test. Categorical variables were analyzed using the chi-square test and presented as frequencies and percentages.

Variables demonstrating statistical significance in univariate analysis were considered for multivariate logistic regression analysis to identify independent predictors of 30-day mortality. Due to the relatively limited number of mortality events, the number of variables included in the multivariate model was intentionally restricted in order to reduce the risk of overfitting. Variable selection was based on clinical relevance and univariate statistical significance, and a parsimonious model was constructed.

The goodness of fit of the logistic regression model was evaluated using the Hosmer–Lemeshow test. No internal validation techniques such as bootstrapping or penalized regression were applied. Statistical significance was defined as p < 0.05. No missing data were observed for the variables included in the statistical analysis.

Results

The median age of patients who underwent surgery due to perforated peptic ulcer was 56 years (IQR: 34–73). Of the population, 66.2% were male. According to the preoperative ASA classification, 70.8% of the patients were classified as ASA 1E or ASA 2E. Among comorbidities, the most common was hypertension (20.1%). Intraoperative evaluations revealed that the ulcer was most frequently located in the prepyloric region of the stomach. The 30-day mortality rate was 26 (16.9%) (Table 1).

Table 1.

Analysis of demographic data

Variables		Median(Q1-Q3)	Count(%)
Age, years		56(34–73)
Gender	Male		102(66.2%)
	Female		52(33.8%)
ASA	ASA1		41(26.6%)
	ASA2		68(44.2%)
	ASA3		41(26.6%)
	ASA4		4(2.6%)
DM			19(12.3%)
HT			31(20.1%)
KAH			17(11%)
CVO			7(4.5%)
KOAH			6(3.9%)
Location	Prepyloric		95(61.7%)
	Duodenum		23(14.9%)
	Antrum		29(18.8%)
	Corpus		7(4.5%)
Mortality			26(16.9%)

In the univariate analysis between groups, statistically significant associations with mortality were found for the following variables: age (p < 0.001), ASA classification (p < 0.001), coronary artery disease (p = 0.032), chronic obstructive pulmonary disease (p = 0.027), creatinine (p = 0.004), albumin (p < 0.001), albumin/creatinine ratio (p < 0.001), CRP (p < 0.001), and defect size (p = 0.002) (Table 2).

Table 2.

Univariate analysis results of the variables

Variables		Median(Q1-Q3)	Count(%)	Median(Q1-Q3)	Count(%)	p- Value
		Alive		Death
Age, year		50(32–68)		74(66–82)		< 0.001
Gender	Male		87(68%)		15(57.7%)	0.312
	Female		41(32%)		11(42.3%)
ASA	ASA1		40(31.3%)		1(3.8%)	< 0.001
	ASA2		58(45.3%)		10(38.5%)
	ASA3		29(22.7%)		12(46.2%)
	ASA4		1(0.8%)		3(11.5%)
DM			15(11.7%)		4(15.4%)	0.604
HT			23(18%)		8(30.8%)	0.138
KAH			11(8.6%)		6(23.1%)	0.032
CVO			4(3.1%)		3(11.5%)	0.06
KOAH			3(2.3%)		3(11.5%)	0.027
Location	Prepyloric		79(61.7%)		16(61.5%)	0.248
	Duodenum		17(13.3%)		6(23.1%)
	Antrum		27(21.1%)		2(7.7%)
	Corpus		5(3.9%)		2(7.7%)
Creatinine, mg/dL		0.81(0.67–0.94)		1.05(0.73–2.06)		0.004
Albumin, gr/dL		3.49(3.03–3.95)		2.55(2.12–3.07)		< 0.001
Albumin/ creatinine		4.46(3.61–5.22)		2.13(1.26–3.38)		< 0.001
CRP, mg/L		44.15(6.56–120)		135(49.85-191.81)		< 0.001
Defect size, mm		5(5–5)		5(5–15)		0.002
Length of hospital stay, day		7(6–8)		3(2–9)		0.016

p-value of < 0.05 was considered statistically significant

Receiver operating characteristic (ROC) analysis demonstrated that the albumin/creatinine ratio showed the highest discriminatory ability for predicting 30-day mortality (AUC: 0.802, 95% CI: 0.687–0.917), with a balanced sensitivity (80.77%) and specificity (80.47%). A cutoff value of < 3.37 was associated with a high negative predictive value (95.37%). CRP levels > 88 mg/L were characterized by very high sensitivity (96.15%) and negative predictive value (98.33%), indicating strong performance in excluding mortality, although specificity was limited (46.09%). Defect size > 5 mm demonstrated high specificity (96.09%) but low sensitivity (30.77%), suggesting limited standalone predictive value. Detailed diagnostic performance metrics are presented in Table 3.

Table 3.

ROC analysis and diagnostic performance of variables predicting 30-day mortality

Variable	Cut-off	AUC	95% CI	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	P-value
Defect size	> 5 mm	0.666	0.633–0.818	30.77	96.09	61.54	87.23	0.008
CRP	> 88 mg/L	0.726	0.633–0.818	96.15	46.09	26.60	98.33	< 0.001
Albumin/Creatinine ratio	< 3.37	0.802	0.687–0.917	80.77	80.47	45.65	95.37	< 0.001

CRP C-reactive protein, AUC Area under the curve, CI Confidence interval, PPV Positive predictive value, NPV Negative predictive value

Cut-off values were determined using Youden’s Index. A p-value of < 0.05 was considered statistically significant

In the multivariate analysis, CRP > 88 mg/L (OR: 3.382, 95% CI: 1.202–9.513; p = 0.021) and albumin/creatinine ratio < 3.37 (OR: 7.769, 95% CI: 2.689–22.448; p < 0.001) were identified as independent risk factors for mortality (Table 4).

Table 4.

Multivariate analysis

Variable	OR	95% CI	P-Value
CAD	3.003	0.792–11.39	0.106
COPD	2.278	0.271–19.164	0.449
Defect size > 5 mm	0.707	0.248–2.017	0.517
CRP > 88 mg/L	3.382	1.202–9.513	0.021
Albumin/Creatinine < 3.37	7.769	2.689–22.448	< 0.001

p-value of < 0.05 was considered statistically significant

Discussion

This study investigated prognostic factors associated with 30-day mortality in patients undergoing surgery for PPU. The research focused on identifying specific clinical and biochemical markers capable of reliably predicting 30-day mortality in this patient population. Among the various parameters examined, two were particularly significant: high CRP levels and low albumin/creatinine ratios. These findings highlight the importance of systemic inflammatory responses and overall nutritional status in determining postoperative outcomes following surgery for perforated peptic ulcers. To the best of our knowledge, evidence regarding the prognostic value of the albumin/creatinine ratio specifically in patients undergoing surgery for perforated peptic ulcer is limited.

High CRP levels observed in patients with higher mortality risk likely reflect the severity of the systemic inflammatory response triggered by the perforation and subsequent peritonitis. CRP is a well-known acute-phase protein and biomarker of inflammation and tissue injury. On the other hand, a low albumin/creatinine ratio indicates impaired nutritional status and potentially compromised renal function, both of which can substantially affect a patient’s ability to tolerate the physiological stress of surgery and postoperative recovery.

In our study, advanced age and high ASA scores were associated with mortality. This can be explained by reduced physiological reserves and the presence of comorbidities in elderly patients, which, when combined with surgical stress, increase the risk of death. Similarly, Burale et al. reported that age ≥ 50 and presentation with shock were independent predictors of mortality [8]. In another study by Menekşe et al., a practical scoring system incorporating age > 65, albumin ≤ 1.5 g/dL, and BUN > 45 mg/dL accurately predicted postoperative mortality in patients undergoing surgery for perforated ulcers. This study shows that age is a strong prognostic marker even on its own [9].

The presence of systemic diseases, such as coronary artery disease and COPD, showed a significant association with mortality in our study. This finding aligns with Burale et al.’s Ethiopia based study, which reported a strong correlation between comorbidities and mortality [8]. In patients with limited cardiopulmonary reserve, sepsis and fluid loss following perforation further increase mortality risk. In a large-scale study conducted by Huang et al. using the Taiwan National Health Database, a scoring system called PPUMS (Perforated Peptic Ulcer Mortality Score) was developed. This system identified comorbidities such as congestive heart failure, advanced liver disease, renal disease, history of malignancy, and obesity as key components increasing mortality risk [10].

In our study, CRP levels > 88 mg/L were significantly associated with 30-day mortality. This finding is consistent with current literature emphasizing the critical role of systemic inflammation in surgical outcomes. CRP is an acute phase reactant synthesized by hepatocytes and rapidly increases in cases of infection, tissue damage or inflammation. In severe intra-abdominal infections like perforated peptic ulcers, CRP levels reflect the severity of peritonitis and sepsis risk, serving as an important prognostic biomarker. In a large-scale study conducted by Huang et al. in 2023, CRP levels were incorporated into the PPUMS system to predict mortality in patients undergoing surgery for perforation. When combined with age and comorbidities, high CRP levels were associated with mortality risks up to 45.9% [10]. High CRP levels are linked to widespread peritonitis, bacterial translocation, and systemic inflammatory response syndrome (SIRS) following perforation. This situation can lead to hemodynamic instability, multiple organ failure, and therefore death, particularly in elderly and comorbid patients. CRP’s rapid increase and ease of measurement make it a practical tool for preoperative risk assessment. From a practical standpoint, these biomarkers may assist surgeons in early postoperative risk assessment and in identifying patients who may benefit from closer monitoring during the perioperative period. Such information may support clinical judgment, particularly in emergency surgical settings.

The albumin/creatinine ratio serves as a composite parameter reflecting both nutritional status and renal function. This ratio provides a dual risk indicator, reflecting malnutrition due to hypoalbuminemia and impaired renal function. Hypoalbuminemia is associated with impaired immune response, delayed wound healing, and increased risk of complications. High creatinine levels indicate reduced renal reserve and the risk of postoperative renal failure. The ratio of these two parameters can serve as a strong prognostic biomarker for surgical outcomes, particularly in elderly and multimorbid patients. In this study, it was found that the risk of mortality increased 7.7 times in patients with an albumin/creatinine ratio < 3.37. Lin et al. demonstrated a similar finding, showing that the albumin/creatinine ratio was an independent risk factor for 28-day mortality in intensive care patients with sepsis, exhibiting a nonlinear negative dose–response relationship [11]. Gibbs et al. reported that low albumin levels are independently associated with surgical mortality [12]. This ratio represents an accessible and cost-effective tool for risk stratification, particularly in resource-limited healthcare systems.

The albumin/creatinine ratio reflects the combined effects of systemic inflammation, nutritional reserve, renal function, and intravascular volume status. Hypoalbuminemia may indicate both malnutrition and an ongoing inflammatory response, whereas elevated creatinine levels reflect impaired renal reserve and are frequently influenced by dehydration and reduced renal perfusion. In the setting of perforated peptic ulcer, perioperative fluid shifts, sepsis-related capillary leakage, and dehydration may further accentuate these changes, making the albumin/creatinine ratio more sensitive to physiological stress than either parameter alone. Therefore, this composite ratio may provide incremental prognostic value beyond individual biomarkers by integrating multiple pathophysiological pathways associated with postoperative mortality.

Although our ROC analysis indicated that defect size > 5 mm was significantly associated with mortality, it did not remain an independent predictor in multivariate analysis. This suggests that the local size of the perforated area alone is insufficient, and systemic responses (such as CRP levels and the albumin/creatinine ratio) are more decisive. Similarly, the literature indicates that defect size may correlate with complications but has limited predictive value for mortality [9].

Several important clinical predictors of mortality in perforated peptic ulcer, such as shock at presentation, time interval between perforation and surgery, established severity scoring systems (e.g., Boey score or Mannheim Peritonitis Index), perioperative hemodynamic instability, intensive care unit requirement, and detailed operative characteristics, were not available for analysis in this retrospective dataset. The absence of these variables may have introduced residual confounding and may partially influence the observed associations between biochemical biomarkers and mortality. Future prospective studies incorporating both clinical severity scores and biochemical markers are needed to better define their relative and combined prognostic value.

Conclusion

The findings of our study are consistent with results reported in recent multicenter and international studies. To the best of our knowledge, evidence regarding the prognostic value of the albumin/creatinine ratio specifically in patients undergoing surgery for perforated peptic ulcer is limited. Even in low-resource countries, simple biomarkers have been shown to effectively predict surgical outcomes. Parameters such as CRP and the albumin/creatinine ratio can be easily evaluated preoperatively and used for risk stratification in surgical decision-making process. In this context, the clinical contribution of our study lies in highlighting the utility of non-invasive, inexpensive, and accessible parameters for predicting mortality. Integrating these biomarkers into routine preoperative assessment protocols may improve patient management and optimize surgical outcomes. These findings may also contribute to the development of simple preoperative risk stratification protocols in emergency surgical settings. In future studies, the dynamic changes of these biomarkers, the effect of surgical techniques, and their relationship with long-term outcomes should be examined in more detail.

Abbreviations

PUD

Peptic ulcer disease

PPU

Perforated peptic ulcers

CRP

C-reactive protein

IQE

Interquartile range

ASA

The American Society of Anesthesiologist

ROC

Receiver operating characteristic

PPUMS

Perforated Peptic Ulcer Mortality Score

SIRS

Systemic inflammatory response syndrome

Authors’ contributions

Conceptualization: F.C., S.O. Data curation: İ.D., S.N.Ö. Formal analysis: C.C., N.C.Ç., Funding acquisition: S.O., F.C. Investigation: İ.D., S.N.Ö. Methodology: C.C., N.C.Ç. Project administration: F.C., İ.D. Resources: İ.D., S.N.Ö. Software: F.C., İ.D. Supervision : C.C., F.C. Validation: N.C.Ç., S.N.Ö. Visualization: S.O., İ.D).

Funding

Authors received no specific funding for this work.

Data availability

Due to institutional regulations and ethical restrictions related to patient confidentiality, the individual-level anonymized data used in this study cannot be publicly shared. However, aggregated data supporting the findings of this study may be made available from the corresponding author upon reasonable request. All data access requests will be evaluated in accordance with institutional data protection policies.

Declarations

Ethics approval and consent to participate

The study was conducted in accordance with the Helsinki Declaration and received ethical approval from the Amasya University Non-Interventional Ethics Committee (Date: 30.06.2025, No: 2025/99). The present study was designed as a retrospective analysis utilizing existing medical records. As the nature of the study involves the use of anonymized and de-identified patient data, with no direct interaction with human subjects and no access to personal identifying information, the requirement for individual participant consent was deemed unnecessary by the Ethics Committee. Consequently, a formal “Consent to Participate” declaration is not applicable to this manuscript.

Consent for publication

The present study was designed as a retrospective analysis utilizing existing medical records. As the nature of the study involves the use of anonymized and de-identified patient data, with no direct interaction with human subjects and no access to personal identifying information, the requirement for individual participant consent was deemed unnecessary by the Ethics Committee. Consequently, a formal “Consent to Participate” declaration is not applicable to this manuscript.

Competing interests

The authors declare no competing interests.