Abstract

Background:

Accurate distinction between complicated acute appendicitis (CAA) and simple acute appendicitis (SAA) is crucial for guiding treatment decisions. This study aimed to systematically evaluate the diagnostic value of peripheral blood inflammatory composite markers and develop a predictive model for CAA.

Methods:

In this retrospective study, 533 patients with acute appendicitis (429 in the training set and 104 in the validation set) admitted between January 2018 and December 2024 were enrolled. Various peripheral blood composite inflammatory markers were calculated. Multivariate stepwise logistic regression identified independent predictors. Based on these predictors, a diagnostic nomogram was constructed and its performance was evaluated using receiver operating characteristic (ROC) curve analysis, calibration curves, and decision curve analysis (DCA).

Results:

The Systemic Inflammation Response Index (SIRI, OR = 2.972, 95% CI: 1.722–5.187, P < 0.001), Neutrophil-to-Albumin Ratio (NAR, OR = 3.099, 95% CI: 1.693–5.723, P < 0.001), and Prognostic Nutritional Index (PNI, OR = 0.553, 95% CI: 0.309–0.969, P = 0.041) were identified as independent predictors. The areas under the ROC curve (AUC) for SIRI, NAR, and PNI were 0.739, 0.742, and 0.535, respectively. The nomogram model incorporating these three markers showed an AUC of 0.758 (95% CI: 0.714–0.803) in the training set and 0.756 (95% CI: 0.656–0.857) in the validation set. Calibration curves indicated good agreement between predicted and observed outcomes (mean absolute error [MAE] = 0.022 and 0.025, respectively), and DCA confirmed the clinical utility of the model across a range of threshold probabilities in both cohorts.

Conclusion:

This study demonstrates that SIRI, NAR, and PNI are independently associated with acute appendicitis severity. The developed nomogram, integrating these three markers, offers a simple, cost-effective tool with good diagnostic performance for distinguishing CAA from SAA, potentially aiding in timely clinical decision-making and individualized treatment planning.

Introduction

Acute appendicitis is one of the most common surgical acute abdominal conditions, with its global occurrence rate having increased by 63.5% over the past decade (1). Acute appendicitis can be further classified into simple acute appendicitis and complicated acute appendicitis. The occurrence rate of complicated appendicitis is as high as 28%–29%, imposing a considerable burden on healthcare systems (2, 3). Currently, the standard treatment for acute appendicitis is surgical resection via appendectomy (3). However, a comparative study published in The Lancet on antibiotic therapy vs. appendectomy for acute appendicitis demonstrated that antibiotic treatment is non-inferior to appendectomy for patients with uncomplicated acute appendicitis (4). Meanwhile, with the emergence of endoscopic retrograde appendicitis therapy, treatment options for acute appendicitis have significantly expanded (5). Additionally, amid increasing social stress and an accelerated pace of life, a growing number of young patients with appendicitis prefer conservative treatment. However, conservative management for complicated appendicitis is often less effective and carries a higher risk of severe complications, leading to prolonged hospitalization, increased mortality, and elevated medical costs (6, 7). Therefore, accurate classification of acute appendicitis subtypes is crucial for guiding clinical decision-making. Several scoring systems have been developed to aid in the diagnosis of acute appendicitis (8), nevertheless, these tools are limited in their ability to effectively predict disease severity. Composite inflammatory markers derived from peripheral blood tests are characterized by their simplicity and cost-effectiveness and are widely used for disease diagnosis and prognosis. Inflammatory indices such as the neutrophil-to-albumin ratio (NAR) (9), Systemic Inflammation Response Index (SIRI) (10), platelet-to-lymphocyte ratio (PLR) (11) and neutrophil-to-lymphocyte ratio (NLR) (12) have been shown in various studies to possess certain predictive value for the severity of appendicitis. Nevertheless, many of these studies suffer from the limitation of not comprehensively including a wide range of inflammatory markers. Thus, this study aims to systematically investigate key predictors of complicated acute appendicitis by incorporating multiple peripheral blood-derived composite inflammatory markers and to develop a diagnostic nomogram model. This approach is expected to provide a clinically practical solution for the precise stratification of acute appendicitis severity.

Materials and methodsStudy subjects

This retrospective study enrolled patients who were treated for acute appendicitis at the 906th Hospital of Joint Logistics Support Force of Chinese People's Liberation Army (PLA) between January 2018 and December 2024 (Full dataset). Among them, patients from January 2018 to December 2022 were included in the training set, and patients from January 2023 to December 2024 were assigned to the validation set. All subjects met the following inclusion and exclusion criteria. Inclusion criteria: 1) Definitive diagnosis of acute appendicitis confirmed by imaging examinations (including appendiceal ultrasound and abdominal computed tomography) combined with laboratory test results; 2) Age ranging from 18 to 60 years; 3) No history of digestive tract diseases, e.g., inflammatory bowel disease, gastrointestinal tumors, irritable bowel syndrome; 4) No use of antibiotics or hormonal medications within the 3 months prior to enrollment; 5) Good compliance and cooperation with diagnosis and treatment. Exclusion criteria: 1) Refusal to undergo surgical appendectomy; 2) Pregnancy or lactation; 3) Incomplete clinical data. Based on the final postoperative pathological reports, patients were divided into two groups: the simple acute appendicitis (SAA) group, pathologically diagnosed with simple acute appendicitis, and the complicated acute appendicitis (CAA) group, pathologically diagnosed with acute suppurative appendicitis, gangrenous appendicitis, or appendiceal perforation.

Data collection

General data collected included age, gender, smoking history (defined as smoking >1 cigarette per day for >6 months), and alcohol consumption history (defined as weekly alcohol intake >10 g for >1 year). Laboratory test data comprised white blood cell count, neutrophil count, lymphocyte count, monocyte count, red blood cell count, platelet count, red blood cell distribution width, and albumin level. Laboratory test data, collected within 24 h of hospital admission, included white blood cell count, neutrophil count, lymphocyte count, monocyte count, red blood cell count, platelet count, red blood cell distribution width, and serum albumin level. Clinical data encompassed preoperative diagnostic imaging results, detailed surgical records, and postoperative pathological reports.

Definition of peripheral blood composite inflammatory markers

The following composite inflammatory markers were calculated based on peripheral blood test results, with all cell counts measured in *109/L, except for red blood cell count in *1012/L, and red blood cell distribution width (RDW) in %, serum albumin level in g/L. NLR (12), Neutrophil-to-lymphocyte ratio; PLR (11), Platelet-to-lymphocyte ratio; LMR (13), Lymphocyte-to-monocyte ratio; SII (14), Systemic immune-inflammation index, calculated as platelet count * monocyte count/lymphocyte count; SIRI (10), Systemic inflammation response index, calculated as neutrophil count * monocyte count/lymphocyte count; PNI (15), Prognostic nutritional index, calculated as albumin level +5 * lymphocyte count; NAR (9), Neutrophil-to-albumin ratio; PAR (16), Platelet-to-albumin ratio; NPR (13), Neutrophil-to-platelet ratio; RLR (17), RDW-to-lymphocyte ratio.

Statistical analysis

Statistical analyses were performed using R software (version 4.3.2). Continuous data were expressed as median (interquartile range, IQR:Q1, Q3). The Kolmogorov–Smirnov test was used to assess the normality of data distribution. For normally distributed continuous data, comparisons between the two groups were conducted using the independent samples t-test, and for non-normally distributed continuous data, the Wilcoxon rank-sum test was applied. Count data were described as n (%), and intergroup comparisons were performed using the Pearson chi-square test. The optimal cut-off values for the continuous variables were calculated using the “logrank” function from the “cutoff” package in R. In the training set, univariate logistic regression analysis was utilized to explore factors distinguishing CAA from SAA. Variables with a P-value <0.1 in univariate analysis were included in the subsequent multivariate stepwise logistic regression analysis. The stepwise selection was based on the Akaike Information Criterion (AIC), with the model having the lowest AIC selected as the optimal model to reduce multicollinearity. The variance inflation factor (VIF) was used to quantify multicollinearity among independent variables, a VIF <10 indicated no significant multicollinearity. A nomogram was constructed based on the final multivariate logistic regression model to visualize the combined prediction model. The diagnostic performance of each independent predictor and the nomogram model for differentiating CAA from SAA was evaluated using receiver operating characteristic (ROC) curves, with the area under the curve (AUC). The calibration of the nomogram was assessed using a calibration curve generated via 1000 bootstrap resamples, and the consistency between predicted probabilities and actual outcomes was quantified using the Hosmer-Lemeshow test. Decision curve analysis (DCA) was performed to evaluate the clinical utility of the nomogram. The validation set was used to further validate the actual performance of the model. Additionally, 10-fold cross-validation was performed on the full dataset to further evaluate the stability and generalizability of the prediction model. All statistical tests were two-sided, with a significance level set at α = 0.05. A P-value <0.05 was considered statistically significant.

Ethics approval and consent to participate

This study was approved by the Ethics Committee of the 906th Hospital of Joint Logistic Support Force of PLA (Ethics Approval Number: PLA906-Research-20250806). All patients signed informed consent documentation before their inclusion in the study and all the procedures followed the ethical standards of the World Medical Association Declaration of Helsinki.

ResultsBaseline characteristics of the study population

A total of 706 patients were initially considered for this study. Among them, 16 patients were under 18 years of age, 68 patients were over 60 years of age, 52 patients declined surgical intervention, and 37 patients had incomplete clinical data. Consequently, 533 patients who met the inclusion and exclusion criteria were ultimately enrolled (full dataset, 429 in the training set and 104 in the validation set, Figure 1). Table 1 presents the differences between patients in the training set and validation set. No significant differences were observed in age, history of smoking, or history of smoking between the two groups. However, the proportion of males in the training set was higher than that in the validation set (72.5% vs. 56.7%, P = 0.028), and there was a significant difference in the incidence of complicated acute appendicitis between the two groups (59.0% vs. 22.1%, P < 0.001). In addition, significant differences were noted in peripheral blood indices and composite inflammatory markers between the two groups (all P < 0.05), except for red blood cell distribution width (RDW) and platelet (PLT).

Participant flow in the study.

VariableTraining set (N = 429)Validation set (N = 104)P valueAge (years)26 (22, 37)30 (20, 36)0.924Gender/Male311 (72.5%)59 (56.7%)0.028Type/CAA253 (59.0%)23 (22.1%)<0.001WBC (*109/L)9.6 (6.3, 13.57)6.19 (5.08, 8.54)<0.001Neutrophil (*109/L)7.4 (3.97, 11.3)4.12 (3.09, 6.68)<0.001Lymphocyte (*109/L)1.5 (1.11, 1.91)1.45 (1, 1.65)0.019Monocyte (*109/L)0.5 (0.34, 0.72)0.35 (0.25, 0.52)<0.001RBC (*1012/L)4.76 (4.34, 5.05)4.48 (4.11, 4.74)<0.001RDW (%)12.9 (12.5, 13.3)13 (12, 14)0.809Platelet (*109/L)199 (169, 228)205.5 (167, 247.25)0.148Albumin (g/L)46.7 (43.4, 49.2)40.6 (36.68, 44.15)<0.001History of smoking123 (28.6%)35 (33.6%)0.380History of alcohol85 (19.8%)23 (22.1%)0.169NLR4.53 (2.29, 8.76)3.01 (1.95, 5.28)0.001PLR130.77 (100, 173.68)152.95 (121.04, 206.46)<0.001LMR3 (1.75, 4.9)3.93 (2.19, 6.08)0.003PNI54.4 (50.95, 57.8)46.88 (43.16, 52.16)<0.001SII67.44 (39.82, 110.13)51.79 (29.78, 100.86)0.033SIRI2.43 (0.8, 5.89)1.07 (0.49, 2.69)<0.001NAR0.16 (0.09, 0.24)0.11 (0.07, 0.18)<0.001PAR4.25 (3.64, 4.95)4.99 (4.19, 6.28)<0.001NPR0.04 (0.02, 0.06)0.02 (0.01, 0.03)<0.001RLR8.53 (6.6, 11.73)9.30 (7.57, 12.80)0.012

Baseline characteristics of the study population .

Baseline characteristics of patients in the training set

The training cohort comprised 311 males (72.5%) and 118 females (27.5%), with a median age of 26 years (IQR: 22, 37). A history of smoking was reported in 123 patients (28.7%), and alcohol consumption was reported in 85 patients (19.8%). Based on postoperative pathological reports, patients were categorized into two groups: 176 patients (41.0%) with simple acute appendicitis (SAA group) and 253 patients (59.0%) with complicated acute appendicitis (CAA group). All patients were successfully treated. Comparison of preoperative baseline characteristics between the two groups revealed no significant differences in age, gender distribution, proportion of smokers, or proportion of alcohol consumers (P > 0.05). However, white blood cell (WBC), neutrophil (Neu), and monocyte (Mon) counts were significantly lower in SAA group compared to CAA group, while the lymphocyte (Lym) count was slightly higher in SAA group. These differences were statistically significant (P < 0.001). In contrast, no significant differences were observed between the two groups in red blood cell (RBC) count, PLT, RDW, or albumin (ALB) levels (P > 0.05). Peripheral blood composite inflammatory indices were calculated according to the specified formulas. Comparative analysis demonstrated that the NLR, PLR, SII, SIRI, NAR, NPR, and RLR were significantly lower in SAA group than in CAA group. Conversely, the LMR was significantly higher in SAA group. All these differences were statistically significant (P < 0.001). No significant differences were found between the two groups for the PNI or PAR. Details are presented in Table 2.

VariableSAA group (N = 176)CAA group (N = 253)P valueAge (years)26 (22, 35)27 (21, 39)0.439Gender/Male123 (69.9%)188 (74.3%)0.369WBC (*109/L)7.1 (5.64, 9.71)12.08 (7.96, 14.87)<0.001Neutrophil (*109/L)4.69 (3.36, 7.74)9.88 (6.15, 12.67)<0.001Lymphocyte (*109/L)1.64 (1.33, 2.09)1.4 (1.04, 1.81)<0.001Monocyte (*109/L)0.42 (0.3, 0.6)0.6 (0.4, 0.83)<0.001RBC (*1012/L)4.75 (4.37, 5.06)4.77 (4.33, 5.04)0.762RDW (%)12.8 (12.5, 13.23)12.9 (12.5, 13.3)0.467Platelet (*109/L)196 (169, 226)200 (170, 229)0.387Albumin (g/L)46.4 (43.3, 49.03)47.1 (43.5, 49.6)0.207History of smoking44 (25.0%)79 (31.2%)0.196History of alcohol30 (17.0%)55 (21.7%)0.282NLR2.88 (1.74, 4.65)6.6 (3.65, 11.11)<0.001PLR116.38 (91.33, 151.47)140.83 (108.1, 190.74)<0.001LMR4.17 (2.59, 5.8)2.33 (1.54, 3.7)<0.001PNI54.88 (51.4, 58.3)54.35 (50.8, 57.5)0.213SII47.88 (31.35, 79.96)82.2 (51.79, 129.64)<0.001SIRI1.19 (0.53, 2.52)3.91 (1.83, 7.49)<0.001NAR0.1 (0.07, 0.16)0.21 (0.14, 0.27)<0.001PAR4.18 (3.59, 4.9)4.26 (3.66, 4.95)0.871NPR0.03 (0.02, 0.04)0.05 (0.03, 0.06)<0.001RLR8.02 (6.12, 9.75)9.07 (7.06, 12.65)<0.001

Baseline characteristics of patients in the training set.

Risk factors for differentiating simple and complicated acute appendicitis

The optimal cutoff values for continuous variables were determined using R software, and these continuous variables were then converted into binary categorical variables based on the identified cutoffs (detailed cutoff values and categorization criteria are presented in Supplementary Table S1). All categorical variables were included in univariate logistic regression analysis to screen for factors associated with distinguishing CAA from SAA (Table 3). Variables with a P-value <0.1 in the univariate analysis were further included in multivariate stepwise logistic regression analysis. Using the backward stepwise method, variables that resulted in the largest increase in the AIC were sequentially removed from the full model until the optimal AIC was achieved (Supplementary Table S2). The final multivariate logistic regression results demonstrated that an elevated NAR (OR = 3.099, 95% CI: 1.693–5.723, P < 0.001), an elevated SIRI (OR = 2.972, 95% CI: 1.722–5.187, P < 0.001), and a decreased PNI (OR = 0.553, 95% CI: 0.309–0.969, P = 0.041) were independent risk factors for distinguishing CAA from SAA (Table 4). Additionally, VIF was used to assess the degree of multicollinearity among the three risk factors. All VIF values were less than 3 (2.434, 1.750 and 2.186, respectively), indicating no significant multicollinearity among them.

VariableOR95% CIP valueAge (years)1.3630.922–2.0150.120Gender/Male1.2460.811–1.9110.313WBC (*109/L)5.7993.798–8.992<0.001Neutrophil (*109/L)6.043.922–9.472<0.001Lymphocyte (*109/L)0.420.273–0.639<0.001Monocyte (*109/L)2.9881.994–4.524<0.001RBC (*1012/L)1.1940.807–1.7670.375RDW (%)1.2840.872–1.8950.207Platelet (*109/L)1.4580.986–2.160.059Albumin (g/L)1.5051.023–2.2210.038NLR6.3624.135–9.964<0.001PLR2.2951.549–3.421<0.001LMR0.2510.167–0.376<0.001PNI0.6970.471–1.0280.069SII4.192.794–6.341<0.001SIRI6.454.231–9.974<0.001NAR6.9474.36–11.399<0.001PAR1.2250.833–1.8020.302NPR5.1613.393–7.967<0.001RLR2.4121.575–3.745<0.001History of smoking1.3620.887–2.110.161History of alcohol1.3520.83–2.2350.231

Univariate logistic regression analysis for discriminating simple and complicated acute appendicitis.

VariableOR95% CIP valueAlbumin (g/L)1.7260.985–3.0860.061PNI0.5530.309–0.9690.041SIRI2.9721.722–5.187<0.001NAR3.0991.693–5.723<0.001

Multivariate stepwise logistic regression analysis for discriminating simple and complicated acute appendicitis.

Construction and efficacy evaluation of the diagnostic model for differentiating simple and complicated acute appendicitis

Based on the results of the multivariate stepwise logistic regression analysis for differentiating simple from complicated acute appendicitis, a prediction model was constructed using the nomogram method (Figure 2). The discriminatory efficacy of each independent risk factor and the nomogram model was subsequently analyzed using ROC curves. As shown in Figure 3 and Table 5, among the independent influencing factors, the SIRI and the NAR demonstrated relatively good discriminatory diagnostic performance, with AUC of 0.739 (95% CI: 0.693–0.785) and 0.742 (95% CI: 0.696–0.787), respectively. In contrast, the PNI exhibited relatively weaker discriminatory ability, with an AUC of only 0.535 (95% CI: 0.480–0.591). Notably, the nomogram model constructed by combining the three predictors achieved a higher diagnostic efficacy, with an AUC of 0.758 (95% CI: 0.714–0.803). The Hosmer-Lemeshow test was used to assess the calibration of the nomogram model, and the result showed P = 0.872, indicating the good calibration. Furthermore, the bootstrap-based calibration curve of the nomogram model demonstrated excellent predictive performance, with a mean absolute error (MAE) of 0.022 (Figure 4A), suggesting that the predicted risk of CAA generated by the model is highly consistent with the actual pathological diagnosis. And the decision curve analysis (DCA) based on the nomogram indicated that within the clinically relevant threshold probability range of 40% to 80%, applying this model to guide clinical decisions yielded a higher net clinical benefit (Figure 4B).

Nomogram predicting complicated acute appendicitis.

ROC of risk factors and nomogram.

VariableCut-offAUC95% CISensitivitySpecificityP valuePNI55.5750.5350.480–0.59147.2%61.7%<0.001SIRI2.1970.7390.693–0.78572.7%70.8%<0.001NAR0.1950.7420.696–0.78784.7%55.7%<0.001Nomogram0.4910.7580.714–0.80372.2%73.5%<0.001

AUC of the risk factors and nomogram prediction model for predicting complicated acute appendicitis.