A Retrospective Study on Use of Neutrophil‐Lymphocyte Ratio as a Prognostic Measurement to Assess Acute Diverticulitis

ABSTRACT

Background

Acute diverticulitis (AD) is a common surgical condition and the Neutrophil‐lymphocyte ratio (NLR) is an emerging biomarker ratio used to guide its management. The aim of this study is to validate and assess the utility of the NLR in AD in the Australian population.

Methods

This is a single centre retrospective observational study of patients who presented to the emergency department with the diagnosis of AD between September 2018 and September 2023, in Ipswich, Queensland. One thousand five hundred and forty patients were screened against exclusion/inclusion criteria and 634 patients were available for analysis.

Results

The study identified NLR, CRP (C‐reactive protein) and age to be significant coefficients in predicting length of stay (LOS) in regression analysis. NLR (OR1.06, p < 0.001) and CRP (OR1.01, p < 0.001) were significant predictors for surgical management of diverticulitis. NLR was found to be superior predictor of surgical management in ROC analysis (AUC 0.75, sensitivity 65%, specificity 75%, p < 0.001) compared to CRP, but both were equivalent in predicting for diverticulitis severity and percutaneous drainage. Further analysis revealed NLR between those receiving surgery, percutaneous drainage and readmission (One‐way ANOVA) and NLR between modified Hinchey classifications were also significantly different (Mann–Whitney U).

Conclusion

In this study, we have further validated the effectiveness of NLR as a diagnostic marker. In particular, NLR is superior to CRP in predicting surgical management. It has also proven useful to predict for LOS, disease severity and percutaneous drainage. NLR usage should be encouraged in the clinical setting as it is simple and effective.

Graphical abstract representation of Neutrophil‐lymphocyte ratio use as a prognostic marker in acute diverticulitis.

1. Introduction

Acute diverticulitis (AD) is a widely prevalent condition in western countries. In the United States, AD represents the third most common gastrointestinal illness that requires hospitalisation, and is a leading cause of elective colonic resection [1]. Over half of all adults will have diverticulosis by 60 years of age, and 5%–10% of these patients will develop AD at some point in their lifetime [2]. The overall incidence of diverticulitis has been on the rise over time, with estimated aggregate national costs in the United States ranging from $1.8–2.6 billion per year [3]. Frequently, uncomplicated AD can be treated conservatively with intravenous antibiotics and bowel rest. However, complications of AD can result in diverticular bleeding, abscess, fistula, stricture, perforation and sepsis.

The World Society of Emergency Surgery (WSES) recommends a comprehensive evaluation of patients including clinical history, physical examination and recommends the use of contrast enhanced computed tomography (CT) [4]. A variety of classifications can be derived based on CT and clinical findings. The Hinchey classification, Modified Hinchey Classification, American Association for Surgery of Trauma (AAS) and World Society of Emergency Surgery (WSES) classifications are similar in performance with regards to predicting complications of AD.

There has been growing interest in the potential of biomarkers in diverticular disease as non‐invasive, reliable, and cost‐effective tools for early diagnosis of complicated AD. Neutrophil‐lymphocyte ratio (NLR) is one such emerging biomarker used for prognosis. The effectiveness of NLR has been acknowledged across various conditions, both benign and malignant [5, 6, 7]. This biomarker reflects two key aspects of the immune system: the innate immune response, primarily driven by neutrophils, and adaptive immunity, mediated by lymphocytes [8]. Systemic inflammatory response syndrome (SIRS) states are correlated with enhanced neutrophil‐driven killing as part of the innate immune response. As a result, NLR typically shows an increase in neutrophil levels and a decrease in lymphocytes.

Despite the prevalent investigation of this biomarker across a variety of different pathologies, literature regarding the use of the NLR in AD remains relatively understudied. It is particularly unclear regarding its predictivity of disease severity, clinical outcomes, and the need for minimally invasive or emergency surgical interventions. Furthermore, the cut‐off values for what is deemed an abnormal NLR is not clearly known with variability in this value.

This is a single centre retrospective study that assesses the clinical usefulness and diagnostic accuracy of the NLR in the Australian population. The aim of this study is to validate the clinical utility of NLR with AD and to compare it to existing biomarkers such as C‐reactive protein (CRP).

2. Methods

This is a retrospective observational study of all patients who presented to the emergency department with the diagnosis of AD between September 2018 to September 2023. This study was approved as a low and negligible risk study by the West Moreton Hospital and Health Service (WMHHS) Human Research Ethics Committee. The inclusion criteria included patients (i) ≥ 18 years old, (ii) referred from Ipswich emergency department for AD, (iii) with baseline investigations performed such as basic blood test (full blood count, basic biochemistry test) and a CT abdomen with contrast.

Patients are excluded if they demonstrated the following (i) active haematological disorder (such as leukaemia or lymphoma), (ii) recent use of cytotoxic chemotherapy, granulocyte colony stimulating factor (G‐CSF), (iii) recent use of any exogenous corticosteroid, (iv) active human immunodeficiency virus (HIV) or acquired immune deficiency syndrome (AIDS). These conditions were chosen as they can affect the patient's neutrophil or lymphocyte numbers [9, 10]. Lastly, patients with incomplete health information (such as incomplete medical history or lack of investigations or results) are also excluded.

2.1. Variables—Patient Characteristics

Individual electronic health information on admission was collected and these include the following. Demographic data included age, gender, body mass index (BMI) and active smoking status. Past or medical history of type 2 diabetes mellitus (T2DM), myocardial infarction (MI), cancer and recent stroke (within 4 weeks); these medical conditions were selected as potential confounders to affect NLR [6, 7, 9, 11, 12].

Radiological outcomes were based on the modified Hinchey classification for diverticulitis severity. The classification is as follows: Class 0—uncomplicated diverticulitis/uAD. The complicated diverticulitis included Class I (confined pericolic abscess) and Class II (abscess distant from primary site), while Class III (generalised purulent peritonitis) and Class IV (generalised faecal peritonitis) [13].

2.2. Variables—Outcome Measures

The dependent quantitative (continuous) variables included the length of stay (LOS), NLR, lymphocyte count (× 10⁹/L), neutrophil count (× 10⁹/L) and CRP level (mg/L). The dependent qualitative (binary) variables are the clinical outcomes which are stratified to either the patient being (i) discharged with antibiotics or (ii) hospital admission; any additional management received such as (i) intravenous antibiotic, (ii) surgical procedure and (iii) percutaneous drainage. Additional clinical outcome measures include those re‐admitted for diverticulitis and those with fistula formation.

2.3. Statistical Analysis

The data were analysed using the JASP 0.17.3 (intel, University of Amsterdam, Netherlands) and PSPP 2.0.1.2 (GNU project, USA) statistical software. Descriptive statistics were used to describe the demographic details based on modified Hinchey classifications. Kendall tau correlation was used to assess for correlations between continuous variables. A linear regression was used to assess for regression coefficients that could predict LOS. Logistic regression analysis was used to predict clinical outcomes based on inflammatory markers; confounds or covariates were excluded in a stepwise approach to identify significant regression coefficients.

One‐way ANOVA test was used to compare the differences in means of dependent continuous variables between Hinchey classifications; the nonparametric equivalent Kruskal–Wallis test was used if the distribution was not normal, followed by Games–Howell post hoc test to identify which groups were significantly different. The Mann–Whitney U test was used to compare the differences in means of dependent continuous variables between clinical outcomes (binary variables ~ intravenous antibiotic, surgery, percutaneous drainage, fistula and readmission).

The receiver operating characteristic (ROC) curve analysis was performed for CRP and NLR. The analysis assessed the effectiveness of the markers to predict diverticulitis severity (simple/complicated diverticulitis) and clinical outcomes (intravenous antibiotic, surgery and percutaneous drainage). The Youden index (J) was used to determine the optimal cut‐off value for the diagnostic marker.

3. Results

The electronic health information of 1540 patients met the inclusion criteria and was screened accordingly. Five hundred and fifty seven patients met the exclusion criteria and an additional 349 patients were removed from analysis due to incomplete or missing data. A total of 634 patients were available for the analysis.

3.1. Descriptive Statistics

The demographic details are described in Table 1 and organised as modified Hinchey classification. From the table, the LOS and NLR progressively increase with the severity of modified Hinchey classification. The frequency counts on additional demographic details and known confounders are described in Table 2.

TABLE 1.

Demographics details based on modified Hinchey classifications.

	Modified Hinchey classification
	0	1	2	3	4
Age (years)
Mean (SD)	56.8(13.3)	56.5(14.3)	60.7(13.1)	58.1(14.7)	74.0(18.4)
Range	27.0–91.0	26.0–94.0	30.0–83.0	33.0–86.0	61.0–87.0
LOS (days)
Mean (SD)	2.7(2.0)	4.5(5.0)	6.9(6.0)	9.6(7.1)	12.4(2.5)
Range	0.1–19.4	0.0–54.5	0.5–31.7	2.0–37.1	10.6–14.1
BMI
Mean (SD)	32.8(7.8)	31.8(7.4)	30.3(6.5)	30.3(5.1)	29.2(5.9)
Range	17.0–63.4	15.8–64.2	15.0–45.0	20.7–40.3	25.0–33.3
CRP
Mean (SD)	67.6(59.4)	106.9(88.6)	124.1(82.1)	113.7(75.9)	47.5(64.4)
Range	1.9–337.0	0.4–426.0	2.4–298.0	2.4–313.0	1.9–93.0
NLR
Mean (SD)	5.6(6.5)	6.6(5.4)	6.9(4.7)	9.8(6.7)	11.7(8.6)
Range	0.7–63.4	0.1–50.2	1.2–20.1	2.1–31.1	5.6–17.8

Abbreviation: SD, standard deviation.

TABLE 2.

Demographic details by frequency counts (binary variables) based on modified Hinchey classifications.

	Modified Hinchey classification
	0	1	2	3	4
Gender
Male	106	154	21	20	1
Female	160	145	19	7	1
Hospital status
Admission	249	290	40	27	2
Discharge	17	9	0	0	0
Readmission
Yes	23	20	11	0	0
No	243	279	29	27	2
Surgery
Yes	2	21	7	20	2
No	264	278	33	7	0
Percutaneous drainage
Yes	0	7	13	2	0
No	266	292	27	25	2
Intravenous antibiotic
Yes	254	276	31	14	0
No	12	23	9	13	2
Fistula
Yes	0	5	8	1	0
No	266	294	32	26	2
Active smoker
Yes	71	104	15	9	1
No	195	195	25	18	1
MI history
Yes	19	23	2	2	0
No	247	276	38	25	2
T2DM history
Yes	47	31	8	5	0
No	219	268	32	22	2
Cancer history
Yes	15	16	5	2	0
No	251	282	35	25	2
Stroke history
Yes	0	1	0	0	0
No	266	297	49	27	2

Abbreviations: MI, myocardial infarction; T2DM, type 2 Diabetes Mellitus.

3.2. Correlation Studies

Kendall's tau showed significant correlation between continuous variables, see Table 3. NLR correlates positively with LOS (0.24, p < 0.001) and CRP (0.25, p < 0.001). LOS correlates positively with CRP (0.23, p < 0.001) and age (0.059, p = 0.027).

TABLE 3.

Kendall tau correlations among continuous variables.

	Age	BMI	LOS	CRP	NLR
Age	1	—	—	—	—
BMI	−0.15 p < 0.001	1	—	—	—
LOS	0.059 p = 0.027	−0.031 p = 0.242	1		—
CRP	−0.014 p = 0.611	−0.042 p = 0.118	0.23 p < 0.001	1	—
NLR	0.04 p = 0.134	−0.096 p < 0.001	0.24 p < 0.001	0.25 p < 0.001	1

3.3. Mann–Whitney U Test

The Mann–Whitney U test demonstrated significant differences in the NLR, CRP and LOS between patients who had received surgery, received percutaneous drainage, and those who were re‐admitted for diverticulitis (Table 4).

TABLE 4.

Mann–Whitney U test showing the mean differences of CRP, NLR and LOS among clinical outcomes.

Clinical outcome		Mean (SD)
		CRP	NLR	LOS
Surgery	No	87.37 (76.28)	5.99 (5.77)	3.39 (2.78)
	Yes	139.07 (95.48)	10.48 (6.65)	12.28 (9.76)
	p	< 0.001	< 0.001	< 0.001
Percutaneous drainage	No	89.78 (78.49)	6.31 (6.01)	3.95 (4.30)
	Yes	142.28 (84.57)	7.63 (4.81)	8.88 (7.76)
	p	< 0.001	< 0.001	< 0.001
Readmission	No	94.14 (80.37)	—	—
	Yes	64.40 (59.78)	—	—
	p	0.008	—	—

Abbreviation: SD, standard deviation.

3.4. ANOVA

We found significant differences in the NLR between modified Hinchey classifications with one‐way ANOVA. Follow‐up post hoc comparisons mean difference between Hinchey Class 0 & 3 was −4.15 (±1.35), p_tukey = 0.033. The Kruskal–Wallis test was used for CRP and LOS, which also showed significant differences between the severity of diverticulitis. Notably, for CRP, the post hoc mean difference between Hinchey Class 0 & 1 was −39.37 (±6.29), p_tukey < 0.001; Class 0 & 2, mean difference, −56.48 (±13.48), p_tukey = 0.001; and Class 0 & 3, mean difference was −46.09 (±15.05), p_tukey = 0.035.

3.5. Linear Regression

We performed a multilinear regression model for length of stay, which demonstrated to be a weak but significant model. By a stepwise exclusion of covariates and confounds, we identified age, NLR, and CRP to be significant coefficients in the regression model with adjusted R ² = 0.17, p < 0.001. We can model as LOS = −0.799 + (0.042 × age) + (0.015 × CRP) + (0.182 × NLR).

3.6. Logistic Regression

We identified a weak but significant model for prediction for surgery. The two significant predictors are NLR with OR 1.06 (95% CI 1.03–1.10), p < 0.001 and CRP with OR 1.01 (95% CI 1.00–1.01), p < 0.001. With prediction for percutaneous drainage, we identified CRP as a weak predictor with OR 1.01 (95% CI 1.00–1.01), p = 0.001 and history of T2DM with OR 3.10 (95% CI 1.12–8.35), p = 0.029. There were no significant logistic regression models identified for intravenous antibiotic usage, fistula formation and readmission.

3.7. ROC Curve Analysis

The analysis showed NLR to be an effective diagnostic marker for surgery and equivalent to CRP when assessing for complicated diverticulitis and percutaneous drainage (Table 5 and Figure 1). No significant finding was identified for either fistula or readmission.

TABLE 5.

ROC analysis and Youden index values on diagnostic markers in predicting clinical outcomes.

Prediction for	Marker	Cut off	Youden index	AUC	Sensitivity (%)	Specificity (%)	p
Surgery	NLR	6.75	0.4	0.75	65	75	< 0.001
Surgery	CRP	132	0.28	0.67	52	76	< 0.001
Complicated AD	NLR	5.62	0.21	0.61	48	73	< 0.001
Complicated AD	CRP	109	0.26	0.64	46	80	< 0.001
Percutaneous drainage	NLR	4.84	0.29	0.64	77	52	0.03
Percutaneous drainage	CRP	43	0.31	0.69	95	36	< 0.001

Abbreviations: AD, acute diverticulitis; AUC, area under curve.

FIGURE 1.

ROC curve for prediction for (i) complicated Acute Diverticulitis (cAD), (ii) Surgery, (iii) Percutaneous Drainage.

4. Discussion

Prompt clinical recognition of complicated AD is required to inform appropriate clinical decision making and therefore prevent morbidity and mortality in patients. This large retrospective cohort study aimed to assess the role of NLR in predicting adverse clinical outcomes in AD. This cohort of 634 patients represents the largest cohort evaluated within the current literature and the only cohort from Australia. Our study supports the use of NLR, as a predictive factor of severity of diverticulitis. A logistic regression model demonstrated NLR was significantly associated with the likelihood of surgical intervention (OR 1.06, 95% CI 1.03–1.10, p < 0.001). CRP was also a significant predictor of surgical intervention after adjustment of confounders (OR 1.01, 95% CI 1.00–1.01, p = 0.008). Following adjustment for known disease confounders (presence of T2DM, prior MI, smoking status, prior malignancy and stroke), there was no change to our findings from the regression model. Furthermore, NLR as well as CRP was also positively correlated with increased LOS. Our study is largely consistent with the existing body of literature.

The first study that described the use of NLR was Zahorec et al. [14]. The authors demonstrated utility in NLR as predictor of clinical course in intensive care unit patients following sepsis or unscheduled surgery or major surgery. Since then, haemogram based biomarkers have been widely investigated, and shown to reliably predict disease severity in a variety of different disease processes such as acute cholecystitis, coronary artery disease and malignancies [15, 16, 17]. Reynolds et al. [18] was the first study to investigate the role of NLR in AD. Their single centre prospective cohort of 101 patients demonstrated the efficacy of NLR in predicting severity of diverticulitis. The cut‐off value for NLR with the best diagnostic yield of NLR was 5.34 (sensitivity 90.48%, specificity 55%). Similarly, they found that NLR demonstrated a high predictive value for surgical procedure compared to CRP.

Following on from this, Mari et al. [19] conducted a single centre retrospective study of 225 patients in 2019 which demonstrated the NLR was also positively correlated with advanced Hinchey classification. Mari et al. [19] reported a cut‐off to be 6.68 (sensitivity 68.75%, specificity 79.21%) as the highest accuracy to predict complicated diverticulitis. Other biomarkers such as CRP were not evaluated in this study. Other studies have derived lower cut‐off values compared to this study.

In 2021, Palacios Huatuco et al. [20] conducted a single centre retrospective study of 325 patients, which identified a cut‐off value of 4.2 (sensitivity 80%, specificity 64.1%) to predict complicated diverticulitis. Similar to our study, their cohort also demonstrated a positive correlation between NLR and the need for percutaneous drainage or surgery. Despite the study demonstrating both mean values of CRP (119.60 vs. 56.36, p < 0.01) and NLR (7.61 vs. 4.04 p < 0.01) were significantly higher in complicated diverticulitis episodes, no direct comparison was made between the different biomarkers.

Sabo et al. [21] conducted a retrospective cohort study of 147 patients and found a similar NLR cut‐off of 4.06 (sensitivity 80%, specificity of 69.3%) to predict complicated diverticulitis. CRP values were not evaluated in this study; however, other biomarkers such as platelet‐to‐lymphocyte ratio (PLR) and monocyte‐to‐lymphocyte ratio (MLR) were compared to NLR. The authors demonstrated that NLR was superior to PLR and MLR in predicting diverticulitis severity. Interestingly, this particular study also followed up patients in the short term and demonstrated NLR was not associated with recurrent episodes of diverticulitis [21]. In contrast, Zager et al. [22], demonstrated a positive correlation between NLR and recurrent episodes of diverticulitis, an increased number of readmissions, and a longer hospital stay. This single‐centre retrospective study (n = 456) identified a cut‐off value of 5.4 for severe diverticulitis.

Recently, Aydin et al. [23] conducted a retrospective cohort study of 82 patients within Türkiye, which demonstrated that a NLR cut‐off of 3.82 (sensitivity 82.4%, specificity 64.6%) was a significant predictor of complicated diverticulitis and outperformed other biomarkers such as CRP and Systemic Inflammatory Index (SII).

This in contrast to our study that CRP was slightly better than NLR in terms of predicting AD severity and the need for percutaneous drainage. Despite this observation, our study demonstrated NLR was a stronger predictor for the need for emergency surgery; we identified 6.75 (AUC = 0.75, sensitivity 65%, specificity 75%) as the NLR cut‐off value to indicate for surgery. A similar cut‐off for the need for surgical intervention was derived in a study conducted by Gonzalez et al. [24]. This single‐centre retrospective cohort study with 102 Mexican patients demonstrated a cut‐off of 5.1 to predict complicated diverticulitis (AUC = 0.633, sensitivity 90%, specificity 43%); as well as a cut‐off of 6.4 to predict the need for surgical intervention (AUC = 0.790, sensitivity 71%, specificity 56%). Arguably the role of NLR in predicting surgical/percutaneous intervention is more clinically relevant. Computed tomography will give clinicians stratification on the radiological severity of disease; however, it is more impactful for a clinician to have a tool to predict the need for intervention.

The results of the current study, along with those from previous studies, suggest that NLR could serve as an important biomarker in the management of diverticulitis. The NLR has an accurate ability to predict disease severity, the need for surgical intervention, and the length of hospital stay, which would significantly improve clinical decision‐making. NLR is an easily obtainable, cost‐effective, and non‐invasive marker, making it an attractive tool for routine clinical use. The strengths of this study are that this cohort represents the largest cohort in the current literature and the first to evaluate Australian patients. Secondly, confounding factors were adjusted for in our analyses. Finally, our study also compares the performance of NLR to CRP, which is the current standard of care [4]. The results suggest NLR can be reliably used as a cost‐effective alternative biomarker to CRP. Despite these promising findings, it is important to acknowledge the limitations of the study.

The retrospective nature of data collection is a potential source for selection or recall bias to be introduced. Furthermore, whilst this study represents a large cohort of patients, the results may have limited generalisability to non‐Caucasian populations, as this was a single‐centre study conducted within Australia. For example, the frequency of right‐sided diverticulitis among Caucasians is reported to be 1.5% compared to 55%–70% amongst the Asian populations [25]. Literature investigating the role of NLR in right‐sided diverticulitis has been carried out in Asian populations with differing results. Su et al. [26] conducted a retrospective cohort study of 289 Chinese patients with right‐sided diverticulitis, which demonstrated that NLR was significantly correlated with complicated diverticulitis (cut‐off 5.7, sensitivity 87.10%, specificity 78.30%). On the contrary, Park et al. [27] conducted a multi‐centre retrospective cohort study of 490 Korean patients with right‐sided colonic diverticulitis, which demonstrated no association between NLR and diverticulitis severity.

Hence, a clear future direction would be to investigate NLR as well as other biomarkers at an international multi‐centre level including both Asian and Caucasian populations. Whilst NLR and CRP are the most prominent markers, other biomarkers such as SII, PLR, delta‐neutrophil index (DNI) should be investigated comparatively and concurrently [28]. There also appears to be significant variability in defining the optimal cut‐off value for NLR in discriminating between simple and complicated diverticulitis. From our review of the current literature, the optimal NLR cut‐offs in predicting complicated diverticulitis range between 3.82–6.40 [23, 24]. However, the mean cut‐off from our review of all studies is greater than 5. This aligns with the current literature, as the general consensus from non‐diverticulitis studies, suggest the NLR cut‐off value of ≥ 5 to determine adverse outcomes in a variety of conditions [29]. This also represents a gap in current research and further research is warranted to further define this cut‐off. As aforementioned, a key cut‐off value that should be determined, is the ability of the biomarker to predict clinical intervention (i.e., percutaneous drainage or emergency surgery), as this provides greater clinical value for the clinician.

Our study demonstrates that NLR is a cost‐effective and reliable prognostic biomarker in AD. Moving forward, NLR can be routinely incorporated into clinical algorithms not only as a predictor of disease severity, but also as a predictive tool for the need for surgery or intervention in complicated diverticulitis. NLR can reliably assist clinicians in the early recognition of patients who may fail conservative management, and as such, facilitate earlier intervention for ‘at‐risk’ patients, leading to better clinical outcomes.

5. Conclusion

In conclusion, the NLR serves as a reliable, accurate, and cost‐effective predictor of complicated diverticulitis and its associated complications. This single‐centre retrospective study of 634 Australian patients represents the largest cohort evaluated and validates the use of NLR through our regression and ROC curve analyses. The NLR can be a useful yet simple auxiliary tool that can be utilised by clinicians to stratify those patients at risk of suffering complications from AD.

Author Contributions

Yam Ting Ho: conceptualization, methodology, writing – original draft, writing – review and editing, formal analysis, data curation, investigation, project administration. Omattage M. Perera: conceptualization, methodology, writing – original draft, writing – review and editing, validation, visualization, resources, project administration. Femi E. Ayeni: methodology, writing – review and editing, validation, funding acquisition. Harrison Gregory: data curation, investigation. Hugh McMahon: investigation, data curation. Punkaja M. S. Amarasekera: investigation, data curation. Yi‐Che Chan: data curation, investigation. Jia Han Chang: investigation, data curation. Tzu Yi Chuang: conceptualization, methodology, validation, writing – review and editing. Peter J. Coverdale: conceptualization, methodology, validation, supervision.

Ethics Statement

Ethics approved as a low and negligible risk study by the West Moreton Hospital and Health Service (WMHHS) Human Research Ethics Committee, reference HREC/2023/QWMS/101750.

Conflicts of Interest

The authors declare no conflicts of interest.

Ho Y. T., Perera O. M., Ayeni F. E., et al., “A Retrospective Study on Use of Neutrophil‐Lymphocyte Ratio as a Prognostic Measurement to Assess Acute Diverticulitis,” ANZ Journal of Surgery 95, no. 7‐8 (2025): 1532–1539, 10.1111/ans.70220.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.