Using the GEPR model to predict outcomes of geriatric emergency general surgery

Using the GEPR model to predict outcomes of geriatric emergency general surgery | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Using the GEPR model to predict outcomes of geriatric emergency general surgery Dequan Xu, Haoxin Zhou, Jie Rong, Xin Xie, Limin Hou This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3725510/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose The present study aimed to develop artificial intelligence (AI)-based model and GEPR, derived from geriatric data, to predict the outcomes of geriatric EGS. Methods A retrospectively database of geriatric EGS patients who underwent emergency surgery was used for the development of the AI model and GEPR. The study employed a specialized algorithm, comprising of four sequential steps: scale prototype selection, clinical data collection and collation, AI model development, and GEPR development. Results In total, 1500 patients were enrolled with mean age of 69.8 years in the study.RandomForestClassifier algorithm outperformed the other AI models. Based on the feature importance, GEPR was derived with a total score range of 0–26. The GEPR has a c-statistic of 0.872 for mortality in hospital (95%CI 0.840–0.905). The observed probability of mortality in hospital gradually increased from 0% at a score of 0 to 63.3% at a score of 12 and 100% at a score of 15 . Conclusion Using patient-related and technical parameters, an GEPR derived from AI algorithms for prediction of surgical complications in geriatric EGS was developed. GEPR reliably predicts postoperative the mortality in hospital in geriatric EGS patients. Further prospective multicenter trials are needed to externally validate the model developed. geriatrics perioperative risk model risk prediction emergency general surgery Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction There are 280 million people aged beyond 60 living in China by the end of 2022. Although they account for just 19.8 percent of the total population, they receive more than one third of all inpatient surgeries. By the middle of this century, people aged 60 and over will take up 30 percent of the Chinese population and an increasing number of surgeries[ 1] . The increasing average age of emergency surgical patients amplifies the impact of functional decline on outcomes. Previous research has highlighted elderly patients undergo emergency general surgery (EGS) and accrue a greater risk of fatal outcomes and postoperative complications than the general population. The enhancement of outcomes in general surgery poses a particularly formidable challenge when it comes to geriatric emergency patients. The accurate prognostic assessment and predictive models are crucial in assisting physicians to make informed decisions. So far, several scoring tools have been developed to predict outcomes of the surgical patient including POSSUM (Physiological and Operative Severity Score for the enumeration of Mortality and morbidity) , SRS (Surgical Risk Scale), PMP (Pre-operative Mortality Predictor), CCI (Charlson comorbidity Index), ASA (American Society of Anaesthesiology) classification and APACHE (Acute Physiology and Chronic Health Evaluation). The previous studies have revealed both strengths and weaknesses in these instruments; however, there remains a lack of consensus regarding their reliability in the geriatric population undergoing emergency laparotomy or how they compare to each other in terms of predicting mortality. Theses scoring systems may be easily evaluated in an elective setting, but their reliability becomes challenging when assessing patients with acute diseases, particularly geriatric emergency cases. Although previous study has found that the Emergency Surgery Score (ESS) can predict mortality in the elderly emergency general surgery (EGS) patients, it is not specifically designed to predict risk in geriatric patients and its accuracy in predicting morbidity decreases for nonagenarians[ 2] . Our hypothesis is that a novel geriatric index, specifically derived from comprehensive geriatric data, will effectively capture the distinct response of this population to various risk factors. The aims of this study were (1) to describe the clinical features and prognosis of acute abdomen in elderly patients, (2) to develop an artificial intelligence (AI)-based model to predict the outcomes of geriatric emergency general surgery, and (3) to derive a geriatric emergency perioperative risk (GEPR) index specifically designed for geriatric patients, taking into account their unique clinical and physiological characteristics. Methods STEP 1: Prototype of the risk score The number of scientific publications on predictive modeling has experienced a significant surge in recent years. Over 280,000 papers had been published with the terms “prognostic factor” or “predictor” in PubMed by December 2022. The cumulative count of publications featuring the term “prediction model” or “prognostic model” had surpassed 27,000 by that point; thus, there were approximately tenfold more studies on prognostic factors published compared to studies on prognostic modeling. Far too frequently , researchers in this particular field adhere to formulaic methodologies without adequately delving into the complexities of selecting prognostic variables and presenting data in a meaningful manner. The era of “big data” has brought about a situation where even minor perturbations, which hold no clinical significance, can attain extreme levels of statistical significance; Clinically significant variables may be statistically validated as invalid data. To prevent this from happening, all prognostic variables in our study were deemed important, and none was excluded. The variables utilized in the GEPR were selected based on existing literature and other currently employed risk indices. The selection of prognostic variables is particularly important in this study. It must comply with the following principles: predictors that are well-defined and reliably conducted Independent prognostic factor that has been confirmed in other studies Prognostic factors recommended by guidelines or consensus Entries in existing models for predicting mortality Age-based regression equations were considered in particular Finally , we first produced a prototype of risk score. In order to facilitate further research, the variables were named uniformly at the same time (Table 1). Description of special variables (220-age)*85% (factor 2) Maximal heart rate (HR max ) is a crucial metric for healthcare professionals to assess cardiovascular compliance during exercise testing and development of personalized exercise prescriptions. The past few years have witnessed the emergence of several supplementary age-based regression equations. Of the available equations, the Fox equation (HR max =220-Age) may be considered the optimal choice for a general population due to its reduced likelihood of under or overestimating individual HR max [ 3] . HR max declines substantially with age, due to stiffening of the aging ventricle, the heart takes longer to fill. Therefore , for the elderly, calculation of the HR max is based on 85% of HR max [(220−age)×85 %][ 4] . The presence of tachycardia exceeding the maximum heart rate can contribute to reduction in cardiac output and result in inadequate tissue perfusion. GNRI (factor 10) The prevalence of protein-energy malnutrition is high among the elderly population. The presence of malnutrition is undeniably linked to a heightened occurrence of mortality due to infection. Recently, the Geriatric Nutritional Risk Index (GNRI) has been reported to be a useful tool to assess older patients’ adverse outcomes, including mortality, a longer length of hospital stay, surgical site infection, and cardiovascular events[ 5] . The GNRI is a straightforward and precise tool that only necessitates the routine measurement of albumin, weight, and height upon admission. Preoperative nutritional status was assessed with GNRI, which was calculated as GNRI=[1.489×albumin (g/L)]+[41.7× actual/ideal body weight ]. In our study, we calculated the ideal body weight through the following Lorentz equations:0.75×height (cm)-62.5 for men, 0.60×height (cm)-40 for women[ 6] . When the actual body weight exceeded the ideal weight, the ratio was adjusted to 1. The patients were classified based on the specified threshold values: major risk (GNRI:<82), moderate risk (GNRI:82 to <92), low risk (GNRI:92 to≤ 98), other results were considered normal in our study. GFR (factor 11) The glomerular filtration rate (GFR) is widely regarded as the most comprehensive indicator of renal function in both healthy individuals and those with pathological conditions. GFR decreases with age, resulting from both physiologic aging of the kidney and specific pathologic influences[ 7] . Earlier reports showed low GFR as well as increasing age being independent risk factors associated with mortality[ 8] . The measurement of GFR is not readily achievable in clinical practice; instead, it is estimated through the utilization of equations incorporating serum creatinine levels, age, and sex. We estimated GFR through the following newly developed CKD-EPI(Chronic Kidney Disease Epidemiology Collaboration) equation[ 9] : GFR=140×(SC÷0.7) -0.176 ×0.993 Age (if SC≦0.7mg/dl, Female); GFR=127×(SC÷0.7) -0.616 ×0.993 Age (if SC>0.7mg/dl, Female); GFR=128×(SC÷0.9) -0.015 ×0.993 Age (if SC≦0.9mg/dl, Male); GFR=119×(SC÷0.9) -0.688 ×0.993 Age (if SC>0.9mg/dl, Male) The patients were classified based on the following threshold values: major risk (GFR [ml/min/1.73m 2 ]:≤44), moderate risk (GFR [ml/min/1.73m 2 ]:≥45 ≤59), other results were considered normal in our study. STEP 2: Collection of clinical data This retrospective study enrolled 1500 elderly patients with abdominal pain (excluding trauma) who underwent emergency surgery and whose information was maintained in the First Affiliated Hospital of Harbin Medical University from January 2017 to January 2022. The hospital provides a full 24-h emergency service with surgery, X-ray, intensive care and has a dedicated acute care surgery (ACS) team. The current study underwent a thorough review and received approval from the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University. The necessity of obtaining individual patient consent for participation is deemed unnecessary. We collected and analyzed data of patients anonymously. Once the study cohort was finalised, the pertinent information was extracted from hospital’s databases: 1)demographic variables (sex and age); 2)anthropometric parameters (height and weight); 3)comorbidities; 4) American society of Anesthesiologist (ASA)’s physical status classification; 5) radiological investigations and preoperative laboratory indexes measured on the day closest to surgery (serum creatinine concentration, WBC count, INR, alkaline phosphatase, serum glutamic-oxaloacetic transaminase(SGOT), serum albumin levels , X rays and CT); 6) preoperative vitals (pulse rate, systolic and diastolic blood pressure); 7) additional data (post-operative complications and mortality rate, stay in hospital and final diagnosis). The concept of functional dependence in our study encompasses both partial and complete dependency. WBC was further divided into 4.5 and 11&15&25*10 3 /mm 3 . Patients with incomplete primary information were excluded from the study . Patients who underwent surgical procedures involving local or regional anesthesia, such as those for anal and subcutaneous abscesses, were also excluded. Additionally, patients who voluntarily discontinued their treatment without following medical advice were not included in the study. The primary outcome was the mortality in hospital. Secondary outcomes included the occurrence of severe postoperative complications, defined as Clavien-Dindo grade Ⅳ or Ⅴ complications (life-threatening and requiring intensive care unit management or death). Serious complications encompass unplanned intubation, prolonged mechanical ventilation, pulmonary embolism, acute renal failure necessitating dialysis, stroke, myocardial infarction, cardiac arrest, septic shock, and mortality. STEP 3: Development of the AI-Based Model The prediction models were constructed using all variables to optimize model performance, without conducting any feature selection preprocessing. The patients’ data underwent refinement to enable machine learning. The dataset were randomly divided into two parts: 80% for training and 20% for testing. The model of each outcome was built with 7 machine learning algorithms, including (1) DecisionTreeClassifier,(2) RandomForestClassifier, (3)GradientBoostingClassifier, (4)KNeighborsClassifier, (5)LogisticRegression, (6)MLPClassifier, and (7)XGBClassifier. The study employed various learning algorithms, and the optimal model demonstrating superior performance was chosen for subsequent stages of machine learning . The performance of the models was evaluated based on metrics including accuracy rate, F1-Score, Precision rate, Recall rate and the area under the receiver operator curve (AUC) . STEP 4: Development of the GEPR The variables were ranked and prioritized based on their feature importance using the best AI-based model. Next, the variables were re-scored according to the features importance derived from the AI-based analysis. The factor will be assigned a score of 1 if it falls within the range of 0 to 0.2. A score of 2 will be assigned if the factor falls within the range of 0.2 to 0.4, and a score of 3 will be assigned for the range of 0.4 to 0.6. This logic continues for subsequent ranges, with each empty interval being assigned a score one less according to the same logic. A new scoring system called GEPR would be easy to utilize. The performance of the GEPR was judged based on the AUC. At last, the scores of 1,500 patients were recalculated based on the new coefficients to characterize clinical significance of the different scores in the GEPR. The overall process of AI-based model and GEPR model development is illustrated in Figure 1, providing a comprehensive summary . Statistical Analysis Descriptive statistics were used to report baseline patient and injury characteristics. Means and SDs or medians and interquartile ranges (IQRs) were calculated for continuous variables, as appropriate. Analysis of data was conducted using SPSS program (version 23, IBM) for Windows. The machine learning model was trained using the keras open source python package. Results One thousand and five hundred patients were enrolled with mean age of 69.8 years and male to female ratio of 1.6:1. when combined, a total number of 904 (60.3%) patients had an age range of 60-70 years, 438 (29.2%) patients were between the ages of 71-80 years, and 158 (10.5%) patients were 81 years and older (Table 2). There were 137 patients died in hospital (mortality rate 9.1%). Median hospital LOS was 11.1 days (IQR,4-18 ). The patients were categorized into three age groups:60-70, 70-80, and >80. The mortality rate in hospital increases proportionally with the age group (6.5% , 11.2%, and 18.4% respectively). The most common surgery indications were infection,obstruction,perforation and bleeding in our study. The final clinical diagnosis is shown in Table 2. The most common cause of acute abdomen in our elderly patients was appendicitis, seen in 428 patients (28.5%), followed by bowel or gastric perforation, seen in 297 patients (19.8%) then large bowel obstruction in 202 patients (13.5%), small bowel obstruction in 134 patients (8.9%), hermia in 127 patients (8.5%), biliary tract diseases in 95 patients (6.3%), gastrointestinal bleeding seen in 83 patients (5.5%), and lastly soft-tissue infection and ischaemia. The overall risk of cancer increases with age, and consequently emergency presentation with malignant disease does too. Demographics and clinical outcome of our study are shown in Table 2 . The results showed that the RandomForestClassifier and XGBClassifier algorithm both obtained the highest AUC value (0.97) for the mortality prediction model(see Figure 2). In terms of accuracy rate, F1-Score, precision rate and the Recall rate, the RandomForestClassifier algorithm is more dominant (see Figure 3). Therefore, the RandomForestClassifier algorithm outperformed the other models. The model achieved an Accuracy of 92.7%, F1-Score of 92.7% , Precision of 92.3% and Recall of 92.6% . Barplot function plotting was used to visualize the predictors while contributing to the decision-making process. The predictors were arranged into descending order depending on the importance estimates shown in Figure 4. As an initial step to check for predictors’ importance, all predictors were deemed important, and none was excluded. GNRI<82(factor 10-major risk), GFR≤44(factor 11-major risk), GFR≥45 ≤59(factor 11-moderate risk), GNRI≥82 15≤25(factor 17-moderate risk) were ranked as the top five important predictors. The factor 10-major risk was assigned a score of 4; factor 11-major risk was assigned a score of 3; factor 17-moderate risk, factor 11-moderate risk,factor 10-moderate risk and factor 16 were assigned a score of 2; the remaining factors were assigned a score of 1. Based on the feature importance of these 19 factors, a score was derived that ranges from 0 to 26 points (Table 3). Therefore , we propose the simplest possible score, with an integer point scale, as the GEPR. This score has a c-statistic of 0.872 for mortality (95%CI 0.840-0.905). The observed probability of mortality in hospital gradually increased from 0% at a score of 0 to 63.3% at a score of 12 and 100% at a score of 15 (Figure 5). There were no observations for scores greater than 16. Discussion The literature provides various definitions for elderly patients, yet there is a lack of clear and definitive criteria. The definition of elderly should not solely rely on chronological age but rather be based on a combination of factors that determine biological aging. The measurement of these factors poses challenges, as clear and objective definitions are not readily available. The term “elderly” is defined in our study as individuals aged 60 years or above. The incidence of elderly patients presenting as an emergency with a surgical condition rises proportionally with age. Our previous research had shown that more than 42% of hospitalizations for emergency surgery conditions occur in older adults (60 years or older), and this proportion is expected to rise significantly over the upcoming decade[ 10] . The proportion of hospital admissions attributed to Emergency General Surgery (EGS) ranges from 8% to 26%[ 11] . The incidence of surgical emergencies of the abdomen is significantly higher among the elderly population compared to other demographic groups. In our study, EGS conditions such as appendicitis,obstruction, and biliary disease are responsible for 3 in every 5 emergency surgical hospitalizations. Emergency surgery in geriatric patients has long been acknowledged to result in higher morbidity and mortality rates compared to younger patients, primarily due to advanced age, diminished physiological reserves and increased burden of medical comorbidities.Van Geloven et al[ 12] reported on patients over age 80 who presented to the Emergency Department (ED) with abdominal pain and found that 27% required surgery, with an overall mortality of 17% that increased to 34% among those who required operative intervention. In our study, the patients were categorized into three age groups:60-70, 70-80, and >80. The mortality rate in hospital increases proportionally with the age group (6.5% , 11.2%, and 18.4% respectively). Risk prediction in healthcare is a complex yet crucial task, particularly within the context of escalating demand and limited or even diminishing resources available. The utilization of risk modeling is beneficial for enhancing the identification of individuals vulnerable to adverse healthcare outcomes, thereby facilitating resource allocation optimization through targeted allocation of limited resources. The literature on prognostic factors for morbidity and mortality in elderly patients undergoing acute abdominal surgery is insufficient, despite the increasing prevalence of this procedure. Recent systematic reviews have revealed a scarcity of risk-prediction instruments specifically designed for the assessment of geriatric emergency patients, and those that are available have limited diagnostic accuracy[ 13] . The estimation of mortality risk for geriatric emergency patients by physicians is predominantly subjective. The utilization of a probability model to predict mortality risk provides a rational and objective approach for quantifying the severity. The previous predictive models, which were all statistical models, relied on multivariate logistic regression analyses. It is worth noting that this approach has been widely used in the field. However, the utilization of the stepwise approach in multivariable regression may lead to model instability and make it highly sensitive to even minor changes in data, such that the inclusion or exclusion of certain observations can significantly alter the model[ 14] . Machine learning and AI are rapidly growing in capability and increasingly applied to model outcomes and complications within medicine[ 15] . The present model is founded upon a machine-learning AI framework. Machine learning is a set of methods that computers use to generate and enhance predictions or behaviors based on data. The AI model is actually a black-box model that cannot be routinely interpreted. Such as the best random forest model in our study, it consist of hundreds of decision trees that “vote” for predictions. This study effectively addresses the limitation of artificial intelligence in medical prediction by utilizing feature importance to assign scores to factors, thereby rendering complex algorithms more comprehensible. The success of this study is attributed to the extensive independent factor screening conducted by doctors based on clinical practice. The present study validates our hypothesis regarding the necessity of conducting specific geriatric analysis and deriving a corresponding model, ultimately resulting in the development of the GEPR. The utilization of GEPR model in predicting post-emergency operation outcomes among elderly patients is crucial for healthcare system organization, clinical decision-making, and effective communication with patients and their families. The specific application process as shown in figure 6. In the following discussion, we will briefly introduce the use of GEPR. 1. Surgical plan The anticipated surge in the elderly population, the heightened rate of emergency presentations, and the augmented risks associated with unplanned surgeries present a formidable challenge to healthcare systems, particularly surgical services.It is important to evaluate the patient’s condition by emergency physicians. In fact, the suitability of elderly emergency patients for emergency surgery necessitates consideration of multiple factors, encompassing age, comorbidities, preoperative vital signs as well as laboratory and imaging findings. Several scores have been assessed that might predict poor outcome following emergency surgery. Many risk scores strive for excessive generality, while others are excessively disease-specific. The novel model incorporates a diverse range of factors utilizing a specialized algorithm. By employing the scoring system derived from this new model, we can identify low-risk groups eligible for emergency operations, such as those with scores below 8, resulting in an overall hospital mortality rate lower than 20%. Likewise, our work also suggests that the decision to deny surgery should not be solely based on age . Indeed, some studies have shown that selected elderly patients may tolerate major emergency surgery and recover well[ 16] . The presence of advanced age alone does not serve as a contraindication for surgical intervention; however, specific circumstances and medical conditions contribute to an increased surgical risk among the geriatric population. Analysis of GEPR scoring results shows that even in elderly patients over 80 years old , if their preoperative score was less than 6, these elderly patients may had a lower mortality rate after surgery. Notable, too, even though the maximum score that can be achieved using GEPR is 26, none of the patients in our cohort scored higher than 16 and none of those scoring 15 survived. We believe that GEPR will take personalized medicine to an entirely new level. 2. Non-surgical plan Emergency Medicine has traditionally emphasised fast-paced and protocolised care for time-dependent health threats. Although the philosophy of Emergency Medicine, initially based on a singular ‘rule-out worst-case scenario’ approach applicable to all age groups, gradually recognized the suboptimal nature of this strategy for many aging populations[ 17] . Implementing early intervention strategies, such as deferring surgical procedures for patients with modifiable risk factors, has the potential to mitigate morbidity and mortality rates within the geriatric surgical patient population. In our study, we have incorporated certain modifiable risk factors like preoperative vital signs and GNRI into the GEPR. Therefore, the scale can be used for dynamic assessment in clinical practice. For patients who exhibit hesitancy or are deemed unsuitable for surgical intervention, non-surgical treatment modalities can be employed. During this phase, the utilization of GEPR enables dynamic evaluation of the patient's condition, and once optimal score results are achieved, surgical procedures or alternative therapeutic approaches can be pursued. 3. palliative care The consideration of palliative care is an alternative for these patients with modifiable risk factors. Avoiding unnecessary prolongation of the dying process, particularly in the ICU, not burdening family;and being aware have all been demonstrated to be of paramount importance to seriously ill surgical patients[ 18] . The main objective of palliative care is to mitigate the suffering associated with invasive treatment in cases where there is a significantly unfavorable prognosis. The provision of expert palliative care services concurrent with disease-directed therapies is known to effectively alleviate patient suffering and avert goal- and value-incongruent care, as well as improve quality of life[ 19] . 4. MAID The decriminalization of Medical Assistance in Dying (MAID), which encompasses both euthanasia and assisted suicide, has been implemented in several countries[ 20] . It is often accepted that we may legitimately speak about euthanasia only in cases of patient who has an incurable medical condition and is experiencing constant and unbearable physical or psychological suffering that cannot be alleviated. Considering recent bioethics literature, a distinction can be made between two main approaches to defining the goals of medicine. On the one hand , the objectives of medicine, such as the promotion of health, alleviation of suffering, and enhancement of wellbeing, are commonly acknowledged. On the other hand, philosophers stressing the value of individual autonomy in biomedical ethics are committed to accepting that the proper goals of medicine are ultimately determined by the autonomous decisions of patients[ 21] . In light of this, it becomes evident that physicians can ethically perform euthanasia upon their patients’ request when they are enduring unbearable physical or psychological suffering. A federal government of Canada report on MAID revealed that in 2019, 65% of MAID procedures took place with the involvement of family medicine and emergency medicine[ 22] . Ball et al. reported that availability of MAID can be therapeutic. They found that MAID deaths provide a greater level of patient comfort than even the deaths from the withdrawal of life support in intensive care units; the availability of MAID has improved the outlook of many patients who have not chosen the procedure. It is important to note, the legalization of euthanasia requires not only patients’ willingness but also an objective evaluation system such as GEPR. The actual GEPR equation is quite complicated and fit only for being deposited within the bowels of a computer, but once one enters some simple information into the computer, data such as age, sex, and serum creatinine, one can readily get a predicting result. The calculation of GEPR may initially appear more time-consuming compared to other existing risk-classification systems; however, we believe that the development of an automated GEPR calculator could facilitate process optimization in hospitals. we currently intend to develop an online calculator with the aim of enhancing the practicality of GEPR for medical professionals. Our study has some limitations. This is not a multicenter study, implying that further researches including various medical centers and larger sample sizes are needed. If the number of training cohort and validation cohort is further expanded, the model will further improve its performance and stability. Moreover, the GEPR model has not yet been compared with other predictive tools currently utilized in geriatric settings. Conclusion The demographic shift towards an ageing population has transformed the clinical populations for all non-paediatric physicians into one in which older people predominate. The physiological changes associated with aging, including diminished organ function and pharmacokinetic and pharmacodynamic variability, along with compromised functional status, necessitate a more personalized approach to treatment decisions in the elderly patients. As discussed above, the diagnosis of acute abdomen in the elderly remains a clinical challenge due to a very various differential diagnosis; the surgical treatment remains the primary choice and approach, however, not all elderly patients may be deemed suitable candidates for surgery and alternative treatments should be considered; the non-operative management should be kept in mind with all its well-known limitations and risks. The GEPR model will help implement geriatric ED interventions to improve emergency care for older patients. Abbreviations EGS emergency general surgery POSSUM Physiological and Operative Severity Score for the enumeration of Mortality and morbidity SRS Surgical Risk Scale PMP Pre-operative Mortality Predictor CCI Charlson comorbidity Index ASA American Society of Anaesthesiology APACHE Acute Physiology and Chronic Health Evaluation ESS Emergency Surgery Score GEPR geriatric emergency perioperative risk HR max Maximal heart rate GNRI Geriatric Nutritional Risk Index GFR glomerular filtration rate CKD-EPI Chronic Kidney Disease Epidemiology Collaboration ACS acute care surgery IQRs interquartile ranges AUC area under the receiver operator curve ED Emergency Department MAID Medical Assistance in Dying Declarations Ethics approval and consent to participate The current study underwent a thorough review and received approval from the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University. The necessity of obtaining individual patient consent for participation is deemed unnecessary. (IRB-AF/SC-04/02.0). Consent for publication Not applicable. Availability of data and materials Part of the data analysed during this study are inclueded in the supplementary information files. Complete datasets are not publicly available, but are available from the corresponding author on reasonable request. Funding No funding for this study. Authors’ contributions D.X: planning and execution of work, manuscript writing. H. Z: data analysis. J. R and X. X: data collection. L.H：planning of study and manuscript editing. All authors discussed and revised the manuscript for submission. References World Health Organization. Populations are getting older. 12 February 2020. https://www.who.int/multi-media/details/populations-are-getting-older. Gaitanidis, A; Mikdad, S; Breen, K; et al.The Emergency Surgery Score (ESS) accurately predicts outcomes in elderly patients undergoing emergency general surgery.[J].Am J Surg.2020,220(4):1052-1057 Shookster, D; Lindsey, B; Cortes, N; et al.Accuracy of Commonly Used Age-Predicted Maximal Heart Rate Equations.[J].Int J Exerc Sci.2020,13(7):1242-1250. Hamrick, I; Meyer, F; Perioperative management of delirium and dementia in the geriatric surgical patient.[J].Langenbecks Arch Surg.2013,398(7):947-55. Bouillanne, O; Morineau, G; Dupont, C; et al.Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients.[J].Am J Clin Nutr.2005,82(4):777-83 Jia, Z; El Moheb, M; Nordestgaard, A; et al.The Geriatric Nutritional Risk Index is a powerful predictor of adverse outcome in the elderly emergency surgery patient.[J].J Trauma Acute Care Surg.2020,89(2):397-404 Lengnan, X, Aiqun, C, Ying, S, et al. The effects of aging on the renal function of a healthy population in Beijing and an evaluation of a range of estimation equations for glomerular filtration rate. Aging (Albany NY). 2021; 13 (5): 6904-6917. doi: 10.18632/aging.202548 Tanos, P, Ablett, AD, Carter, B, et al. SHARP risk score: A predictor of poor outcomes in adults admitted for emergency general surgery: A prospective cohort study. ASIAN J SURG. 2022; 46 (7): 2668-2674. doi: 10.1016/j.asjsur.2022.10.049 Liu, Xun, Wang, Yanni, Wang, Cheng, et al. A new equation to estimate glomerular filtration rate in Chinese elderly population. PloS one. 2013; 8 (11): e79675. doi: 10.1371/journal.pone.0079675 Xu, D; Yin, Y; Hou, L; et al.A special acute care surgery model for dealing with dilemmas involved in emergency department in China.[J].Sci Rep.2021,11(1):1723 Bruns, B, Tesoriero, R, Narayan, M, et al. Emergency General Surgery: Defining Burden of Disease in the State of Maryland AM SURGEON. 2020; 81 (8): 829-834. doi: 10.1177/000313481508100825 van Geloven, AA, Biesheuvel, TH, Luitse, JS, et al. Hospital admissions of patients aged over 80 with acute abdominal complaints. EUR J SURG. 2000; 166 (11): 866-71. doi: 10.1080/110241500447254 O'Caoimh, R, Cornally, N, Weathers, E, et al. Risk prediction in the community: A systematic review of case-finding instruments that predict adverse healthcare outcomes in community-dwelling older adults. MATURITAS. 2015; 82 (1): 3-21. doi: 10.1016/j.maturitas.2015.03.009 Grant, SW, Hickey, GL, Head, SJ. Statistical primer: multivariable regression considerations and pitfalls. EUR J CARDIO-THORAC. 2019; 55 (2): 179-185. doi: 10.1093/ejcts/ezy403 Desai, GS. Artificial Intelligence: The Future of Obstetrics and Gynecology. J OBSTET GYN INDIA. 2018; 68 (4): 326-327. doi: 10.1007/s13224-018-1118-4 Subramanian, A, Balentine, C, Palacio, CH, et al. Outcomes of damage-control celiotomy in elderly nontrauma patients with intra-abdominal catastrophes. AM J SURG. 2010; 200 (6): 783-8; discussion 788-9. doi: 10.1016/j.amjsurg.2010.07.027 Mooijaart, SP, Carpenter, CR, Conroy, SP. Geriatric emergency medicine-a model for frailty friendly healthcare. AGE AGEING. 2022; 51 (3): doi: 10.1093/ageing/afab280 Supiano, KP, McGee, N, Dassel, KB, et al. A Comparison of the Influence of Anticipated Death Trajectory and Personal Values on End-of-Life Care Preferences: A Qualitative Analysis. CLIN GERONTOLOGIST. 2017; 42 (3): 247-258. doi: 10.1080/07317115.2017.1365796 Spencer, AL, Miller, PR, Russell, GB, et al. Timing is everything: Early versus late palliative care consults in trauma. J TRAUMA ACUTE CARE. 2022; 94 (5): 652-658. doi: 10.1097/TA.0000000000003881 Oczkowski Simon J W,Ball Ian,Saleh Carol,et al.The provision of medical assistance in dying: protocol for a scoping review.BMJ open.2017;7 (8)：e017888 Varelius Jukka.Illness, suffering and voluntary euthanasia.BIOETHICS.2007;21 (2)：75-83. First annual report on medical assistance in dying in Canada, 2019. Ottawa, ON: Government of Canada; 2019 Tables Table 1 . Prototype of GEPR Variable Named variable Demographics Age > 80 factor 1 Preoperative vital signs Heart rate >(220-age)*85% factor 2 Systolic blood pressure≤90mmHg or Diastolic blood pressure≤60mmHg factor 3 Comorbidities Neoplastic comorbidity 1* factor 4 History of COPD factor 5 Hypertension factor 6 Diabetes factor 7 Ventilator requirement within 48 hours preoperatively factor 8 Functional dependence factor 9 GNRI ≥92 ≤98 factor 10-low risk ≥82 <92 factor 10-moderate risk 125U/L factor 12 INR > 1.5 factor 13 SGOT > 40U/L factor 14 Platelets 145 mg/dL factor 16 WBC count (×10 3 /mm 3 ) 15≤25 factor 17-moderate risk >25 factor 17-major risk Imaging techniques Pleural effusion factor 18 free air in the abdominal factor 19 Table 2. Demographics and clinical outcome of elderly patients with acute abdomen Variables No.(%) of patients n=1500 Sex, n(%) Female 583 (38.9) Male 917 (61.1) ASA score , n(%) Ⅱ 239 (15.9 ) Ⅲ 971 (64.7) Ⅳ 277 (18.5) Ⅴ 13 (0.9) Age(years), (mean±SD) ≧60 1500 (69.8±7.456) ≧60 ≦70 904 (64.8±7.46) ＞70 ≦80 438 (75.0±7.44) ＞80 158 (84.7±7.49) LOS(days),(IQR) 11.1 (4-18) ICU stay (h), n(%) ≧48 440（29.3） Surgery indication,n(%) infection 600 (40.0) obstruction 348 (23.2) perforation 277 (18.5) bleeding 127 (8.4) incarcerated hermia 114 (7.6) ischaemia 18 (1.2) other 16 (1.1) EGS diagnosis*, n(%) appendicitis 428 (28.5) bowel/gastric perforation 297 (19.8) large bowel obstruction 202 (13.5) small bowel obstruction 134 (8.9) hermia 127 (8.5) biliary tract disease 95 (6.3) gastrointestinal bleeding 83 (5.5) Soft-tissue infection 54 (3.6) ischaemia 18 (1.2) neoplasm 235 (15.7) otherϕ 67 (4.5) Mortality, n(%) ≧60 ≦70 59 (6.5) ＞70 ≦80 49 (11.2) ＞80 29 (18.4) * some patients had more than 1 diagnosis . ϕ postoperative bleeding, diverticulitis,ulcerative colitis,et al. SD=standard deviation Table 3 . Development of the the Geriatric Emergency Perioperative Risk Index Variable Points* Demographics Age >80 1 Preoperative vital signs Heart rate >(220-age)*85% 1 Systolic blood pressure≤90mmHg or Diastolic blood pressure≤60mmHg 1 Comorbidities Neoplastic comorbidity 1* 1 History of COPD 1 Hypertension 1 Diabetes 1 Ventilator requirement within 48 hours preoperatively 1 Functional dependence 1 GNRI ≥92 ≤98 1 ≥82 <92 2 125U/L 1 INR > 1.5 1 SGOT > 40U/L 1 Platelets 145 mg/dL 2 WBC count (×10 3 /mm 3 ) 15≤25 2 >25 1 Imaging techniques Pleural effusion 1 free air in the abdominal 1 Additional Declarations No competing interests reported. Supplementary Files rawdata.xlsx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3725510","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":261194836,"identity":"d6a95ca7-baef-484d-92fc-33688b8a9255","order_by":0,"name":"Dequan Xu","email":"","orcid":"","institution":"the First Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Dequan","middleName":"","lastName":"Xu","suffix":""},{"id":261194838,"identity":"76b11413-2587-4725-b3b4-7291ded8d7f7","order_by":1,"name":"Haoxin Zhou","email":"","orcid":"","institution":"the First Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Haoxin","middleName":"","lastName":"Zhou","suffix":""},{"id":261194840,"identity":"984a871c-8799-4f42-b5b5-7f992293ed74","order_by":2,"name":"Jie Rong","email":"","orcid":"","institution":"the First Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Rong","suffix":""},{"id":261194843,"identity":"0ba455e9-1c6d-4898-89d8-3a0f498fad97","order_by":3,"name":"Xin Xie","email":"","orcid":"","institution":"the First Affiliated Hospital of Harbin Medical University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xin","middleName":"","lastName":"Xie","suffix":""},{"id":261194845,"identity":"945967e1-12a5-446f-bad4-825d49b8f676","order_by":4,"name":"Limin Hou","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxElEQVRIiWNgGAWjYBACfmb24z8+VNjw2B9vIFKLZDtPguSMM2kyDGcOEKnF4DyDgTRv22EbhhsJxGo5zJBgCLSFh3Hm4403GGpsogk77DDjgQSQX5il04otGI6l5TYQ0sIHtOUgyBY26RwzCcaGw4S1MBxmMGwG+oWHR/IMkVoEDjMYM4O0SEjwEKlFspknjRHkMAMeoF8SiPELP//xYwxA79sbsB/eeONDjQ0RfkECBhIJpCiHaCFVxygYBaNgFIwMAAAl2z6mjXVHpAAAAABJRU5ErkJggg==","orcid":"","institution":"the First Affiliated Hospital of Harbin Medical University","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Limin","middleName":"","lastName":"Hou","suffix":""}],"badges":[],"createdAt":"2023-12-08 12:29:19","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false,"coiExplicitlySet":false},"doi":"10.21203/rs.3.rs-3725510/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3725510/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":48688000,"identity":"fd449f37-f11c-490f-9331-963bf2a6b1ca","added_by":"auto","created_at":"2023-12-22 16:03:02","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":460494,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/b8512d157542e5d37b86bdce.png"},{"id":48688006,"identity":"12baf398-4afc-4a68-985a-e4d5bb90c149","added_by":"auto","created_at":"2023-12-22 16:03:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":506431,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/b781ef8016a471ceb57942f9.png"},{"id":48688515,"identity":"654b053e-0f33-4a34-a4ba-7863fe4688d3","added_by":"auto","created_at":"2023-12-22 16:11:02","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":33497,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/c6be985e3f24d478ca4a4a24.png"},{"id":48688002,"identity":"5057080e-207b-466a-b9ad-9386bc75d8b1","added_by":"auto","created_at":"2023-12-22 16:03:02","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":58667,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/072ab978bb508ac727946c1e.png"},{"id":48688514,"identity":"e35b6580-2df0-4150-b4b6-8ebc973cbee7","added_by":"auto","created_at":"2023-12-22 16:11:02","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":39015,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/6c8aac638cde6c48e121958a.png"},{"id":48688004,"identity":"412288e5-f8c0-4fb1-a65d-92801497b606","added_by":"auto","created_at":"2023-12-22 16:03:02","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":152865,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/58c1e49eb63b48d8541adf87.png"},{"id":50578625,"identity":"84b54e06-5487-42d8-b25e-abd573c763ea","added_by":"auto","created_at":"2024-02-02 18:08:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1349053,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/aebc504f-f5a5-4401-b8f5-08a79b9f5140.pdf"},{"id":48688005,"identity":"5459708b-dd08-4dcc-ab76-80c7fe7a8297","added_by":"auto","created_at":"2023-12-22 16:03:02","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":149916,"visible":true,"origin":"","legend":"","description":"","filename":"rawdata.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3725510/v1/4286023b98739eae834b9aa2.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Using the GEPR model to predict outcomes of geriatric emergency general surgery","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThere are 280 million people aged beyond 60 living in China by the end of 2022. Although they account for just 19.8 percent of the total population, they receive more than one third of all inpatient surgeries. By the middle of this century, people aged 60 and over will take up 30 percent of the Chinese population and an increasing number of surgeries[\u003csup\u003e1]\u003c/sup\u003e.\u0026nbsp;The increasing average age of emergency surgical patients amplifies the impact of functional decline on outcomes. Previous research has highlighted elderly patients undergo emergency general surgery (EGS) and accrue a greater risk of fatal outcomes and postoperative complications than the general population. The enhancement of outcomes in general surgery poses a particularly formidable challenge when it comes to geriatric emergency patients. The accurate prognostic assessment and predictive models are crucial in assisting physicians to make informed decisions.\u003c/p\u003e\n\u003cp\u003eSo far, several scoring tools have been developed to predict outcomes of the surgical patient including POSSUM (Physiological and Operative Severity Score for the enumeration of Mortality and morbidity) , SRS (Surgical Risk Scale), PMP (Pre-operative Mortality Predictor), CCI (Charlson comorbidity Index), ASA (American Society of Anaesthesiology) classification and APACHE (Acute Physiology and Chronic Health Evaluation). The previous studies have revealed both strengths and weaknesses in these instruments; however, there remains a lack of \u0026nbsp;consensus regarding their reliability in the geriatric population undergoing emergency laparotomy or how they compare to each other in terms of predicting mortality. Theses scoring systems may be easily evaluated in an elective setting, but their reliability becomes challenging when assessing patients with acute diseases, particularly geriatric emergency cases. Although previous study has found that the Emergency Surgery Score (ESS) can predict mortality in the elderly emergency general surgery (EGS) patients, it is not specifically designed to predict risk in geriatric patients and its accuracy in predicting morbidity decreases for nonagenarians[\u003csup\u003e2]\u003c/sup\u003e. Our hypothesis is that a novel geriatric index, specifically derived from comprehensive geriatric data, will effectively capture the distinct response of this population to various risk factors.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;The aims of this study were (1) to describe the clinical features and prognosis of acute abdomen in elderly patients, (2) to develop an artificial intelligence (AI)-based model to predict the outcomes of geriatric emergency general surgery, and (3) to derive a geriatric emergency perioperative risk (GEPR) index specifically designed for geriatric patients, taking into account their unique clinical and physiological characteristics.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eSTEP \u0026nbsp;1:\u003c/strong\u003e\u003cem\u003e\u0026nbsp;\u003cstrong\u003ePrototype of the risk score\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe number of scientific publications on predictive modeling has experienced a significant surge in recent years. Over 280,000 papers had been published with the terms \u0026ldquo;prognostic factor\u0026rdquo; or \u0026ldquo;predictor\u0026rdquo; in PubMed by December 2022. The cumulative count of publications featuring the term \u0026ldquo;prediction model\u0026rdquo; or \u0026ldquo;prognostic model\u0026rdquo; had surpassed 27,000 by that point; thus, there were approximately tenfold more studies on prognostic factors published compared to studies on prognostic modeling. Far too frequently , researchers in this particular field adhere to formulaic methodologies without adequately delving into the complexities of selecting prognostic variables and presenting data in a meaningful manner. The era of \u0026ldquo;big data\u0026rdquo; has brought about a situation where even minor perturbations, which hold no clinical significance, can attain extreme levels of statistical significance; Clinically significant variables may be statistically validated as invalid data. To prevent this from happening, all prognostic variables in our study were deemed important, and none was excluded. The variables utilized in the GEPR were selected based on existing literature and other currently employed risk indices.\u0026nbsp;The\u0026nbsp;selection of prognostic variables is particularly important in this study.\u0026nbsp;It must\u0026nbsp;comply with\u0026nbsp;the following principles:\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003epredictors that are well-defined and reliably conducted\u003c/li\u003e\n \u003cli\u003eIndependent prognostic factor that has been confirmed in other studies\u003c/li\u003e\n \u003cli\u003ePrognostic factors recommended by guidelines or consensus\u003c/li\u003e\n \u003cli\u003eEntries in existing models for predicting mortality\u003c/li\u003e\n \u003cli\u003eAge-based regression equations were considered in particular\u0026nbsp;\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u0026nbsp; Finally , we first produced a prototype of risk score. In order to facilitate further research, the variables were named uniformly at the same time (Table 1).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eDescription of special variables\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(220-age)*85% \u0026nbsp;(factor 2)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMaximal heart rate (HR\u003csub\u003emax\u003c/sub\u003e) is a crucial metric for healthcare professionals to assess cardiovascular compliance during exercise testing and development of personalized exercise prescriptions. The past few years have witnessed the emergence of several supplementary age-based regression equations. Of the available equations, the Fox equation (HR\u003csub\u003emax\u003c/sub\u003e=220-Age) may be considered the optimal choice for a general population due to its reduced likelihood of under or overestimating individual HR\u003csub\u003emax\u003c/sub\u003e[\u003csup\u003e3]\u003c/sup\u003e. HR\u003csub\u003emax\u003c/sub\u003e declines substantially with age, due to stiffening of the aging ventricle, the heart takes longer to fill. Therefore , for the elderly, calculation of the HR\u003csub\u003emax\u003c/sub\u003e is based on 85% of HR\u003csub\u003emax\u003c/sub\u003e [(220\u0026minus;age)\u0026times;85 %][\u003csup\u003e4]\u003c/sup\u003e. The presence of tachycardia exceeding the maximum heart rate can contribute to reduction in cardiac output and result in inadequate tissue perfusion.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGNRI \u0026nbsp;(factor 10)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe prevalence of protein-energy malnutrition is high among the elderly population. The presence of malnutrition is undeniably linked to a heightened occurrence of mortality due to infection. Recently, the Geriatric Nutritional Risk Index (GNRI) has been reported to be a useful tool to assess older patients\u0026rsquo; adverse outcomes, including mortality, a longer length of hospital stay, surgical site infection, and cardiovascular events[\u003csup\u003e5]\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe GNRI is a straightforward and precise tool that only necessitates the routine measurement of albumin, weight, and height upon admission. Preoperative nutritional status was assessed with GNRI, which was calculated as GNRI=[1.489\u0026times;albumin (g/L)]+[41.7\u0026times; actual/ideal body weight ]. In our study, we calculated the ideal body weight through the following Lorentz equations:0.75\u0026times;height (cm)-62.5 for men, 0.60\u0026times;height (cm)-40 for women[\u003csup\u003e6]\u003c/sup\u003e. When the actual body weight exceeded the ideal weight, the ratio was adjusted to 1. The patients were classified based on the specified threshold values: major risk (GNRI:\u0026lt;82), \u0026nbsp; moderate risk (GNRI:82 to \u0026lt;92), low risk (GNRI:92 to\u0026le; 98), other results were considered normal in our study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGFR\u003c/strong\u003e \u003cstrong\u003e\u0026nbsp;(factor 11)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe glomerular filtration rate (GFR) is widely regarded as the most comprehensive indicator of renal function in both healthy individuals and those with pathological conditions. GFR decreases with age, resulting from both physiologic aging of the kidney and specific pathologic influences[\u003csup\u003e7]\u003c/sup\u003e. Earlier reports showed low GFR as well as increasing age being independent risk factors associated with mortality[\u003csup\u003e8]\u003c/sup\u003e. The measurement of GFR is not readily achievable in clinical practice; instead, it is estimated through the utilization of equations incorporating serum creatinine levels, age, and sex. We estimated GFR through the following newly developed CKD-EPI(Chronic Kidney Disease Epidemiology Collaboration) equation[\u003csup\u003e9]\u003c/sup\u003e:\u003c/p\u003e\n\u003cp\u003eGFR=140\u0026times;(SC\u0026divide;0.7)\u003csup\u003e-0.176\u003c/sup\u003e\u0026times;0.993\u003csup\u003eAge\u003c/sup\u003e (if SC≦0.7mg/dl, Female); GFR=127\u0026times;(SC\u0026divide;0.7)\u003csup\u003e-0.616\u003c/sup\u003e\u0026times;0.993\u003csup\u003eAge\u003c/sup\u003e (if SC\u0026gt;0.7mg/dl, Female); GFR=128\u0026times;(SC\u0026divide;0.9)\u003csup\u003e-0.015\u003c/sup\u003e\u0026times;0.993\u003csup\u003eAge\u003c/sup\u003e (if SC≦0.9mg/dl, Male);\u003c/p\u003e\n\u003cp\u003eGFR=119\u0026times;(SC\u0026divide;0.9)\u003csup\u003e-0.688\u003c/sup\u003e\u0026times;0.993\u003csup\u003eAge\u003c/sup\u003e (if SC\u0026gt;0.9mg/dl, Male)\u003c/p\u003e\n\u003cp\u003eThe patients were classified based on the following threshold values: major risk (GFR [ml/min/1.73m\u003csup\u003e2\u003c/sup\u003e]:\u0026le;44), moderate risk (GFR [ml/min/1.73m\u003csup\u003e2\u003c/sup\u003e]:\u0026ge;45 \u0026le;59), other results were considered normal in our study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSTEP \u0026nbsp;2:\u003c/strong\u003e\u003cem\u003e\u0026nbsp; \u003cstrong\u003eCollection of clinical data\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThis retrospective study enrolled 1500 elderly patients with abdominal pain (excluding trauma) who underwent emergency surgery and whose information was maintained in the First Affiliated Hospital of Harbin Medical University from January 2017 to January 2022. The hospital provides a full 24-h emergency service with surgery, X-ray, intensive care and has a dedicated acute care surgery (ACS) team.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe current study underwent a thorough review and received approval from the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University. The necessity of obtaining individual patient consent for participation is deemed unnecessary. We collected and analyzed data of patients anonymously.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOnce the study cohort was finalised, the pertinent information was extracted from hospital\u0026rsquo;s databases: 1)demographic variables (sex and age); 2)anthropometric parameters (height and weight); 3)comorbidities; 4) American society of Anesthesiologist (ASA)\u0026rsquo;s physical status classification; 5) radiological investigations and preoperative laboratory indexes measured on the day closest to surgery (serum creatinine concentration, WBC count, INR, alkaline phosphatase, serum glutamic-oxaloacetic transaminase(SGOT), serum albumin levels , X rays and CT); 6) preoperative vitals (pulse rate, systolic and diastolic blood pressure); 7) additional data (post-operative complications and mortality rate, stay in hospital and final diagnosis).\u003c/p\u003e\n\u003cp\u003eThe concept of functional dependence in our study encompasses both partial and complete dependency. WBC was further divided into \u0026lt;=4.5,\u0026gt;4.5 and \u0026lt;=11, \u0026gt;11\u0026amp;\u0026lt;=15, \u0026gt;15\u0026amp;\u0026lt;=25, and \u0026gt;25*10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e. Patients with incomplete primary information were excluded from the study . Patients who underwent surgical procedures involving local or regional anesthesia, such as those for anal and subcutaneous abscesses, were also excluded. Additionally, patients who voluntarily discontinued their treatment without following medical advice were not included in the study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe primary outcome was the mortality in hospital. Secondary outcomes included the occurrence of severe postoperative complications, defined as Clavien-Dindo grade Ⅳ or Ⅴ complications (life-threatening and requiring intensive care unit management or death). Serious complications encompass unplanned intubation, prolonged mechanical ventilation, pulmonary embolism, acute renal failure necessitating dialysis, stroke, myocardial infarction, cardiac arrest, septic shock, and mortality. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSTEP \u0026nbsp;3:\u003c/strong\u003e\u003cem\u003e\u0026nbsp; \u003cstrong\u003eDevelopment of the AI-Based Model\u003c/strong\u003e\u003c/em\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe prediction models were constructed using all variables to optimize model performance, without conducting any feature selection preprocessing. The patients\u0026rsquo; data underwent refinement to enable machine learning. The dataset were randomly divided into two parts: 80% for training and 20% for testing. The model of each outcome was built with 7 machine learning algorithms, including (1) DecisionTreeClassifier,(2) RandomForestClassifier, (3)GradientBoostingClassifier, (4)KNeighborsClassifier, (5)LogisticRegression, (6)MLPClassifier, and (7)XGBClassifier.\u003c/p\u003e\n\u003cp\u003eThe study employed various learning algorithms, and the optimal model demonstrating superior performance was chosen for subsequent stages of machine learning . The performance of the models was evaluated based on metrics including accuracy rate, F1-Score, Precision rate, Recall rate and the area under the receiver operator curve (AUC) .\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSTEP \u0026nbsp;4:\u003c/strong\u003e\u003cem\u003e\u0026nbsp; \u003cstrong\u003eDevelopment of the GEPR\u003c/strong\u003e\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eThe variables were ranked and prioritized based on their feature importance using the best AI-based model. Next, the variables were re-scored according to the features importance derived from the AI-based analysis. The factor will be assigned a score of 1 if it falls within the range of 0 to 0.2. A score of 2 will be assigned if the factor falls within the range of 0.2 to 0.4, and a score of 3 will be assigned for the range of 0.4 to 0.6. This logic continues for subsequent ranges, with each empty interval being assigned a score one less according to the same logic. A new scoring system called GEPR would be easy to utilize. The performance of the GEPR was judged based on the AUC. At last, the scores of 1,500 patients were recalculated based on the new coefficients to characterize clinical significance of the different scores in the GEPR. The overall process of AI-based model and GEPR model development is illustrated in Figure 1, providing a comprehensive summary .\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eDescriptive statistics were used to report baseline patient and injury characteristics. Means and SDs or medians and interquartile ranges (IQRs) were calculated for continuous variables, as appropriate. Analysis of data was conducted using SPSS program (version 23, IBM) for Windows. The machine learning model was trained using the keras open source python package.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eOne thousand and five hundred patients were enrolled with mean age of 69.8 years and male to female ratio of 1.6:1. when combined, a total number of 904 (60.3%) patients had an age range of 60-70 years, 438 (29.2%) patients were between the ages of 71-80 years, and 158 (10.5%) patients were 81 years and older (Table 2). There were 137 patients died in hospital (mortality rate 9.1%). Median hospital LOS was 11.1 days (IQR,4-18 ). The patients were categorized into three age groups:60-70, 70-80, and \u0026gt;80. The mortality rate in hospital increases proportionally with the age group (6.5% , 11.2%, and 18.4% respectively). \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe most common surgery indications were infection,obstruction,perforation and bleeding in our study. The final clinical diagnosis is shown in Table 2. The most common cause of acute abdomen in our elderly patients was appendicitis, seen in 428 patients (28.5%), followed by bowel or gastric perforation, seen in 297 patients (19.8%) then large bowel obstruction in 202 patients (13.5%), small bowel obstruction in 134 patients (8.9%), hermia in \u0026nbsp;127 patients (8.5%), biliary tract diseases in 95 patients (6.3%), gastrointestinal bleeding seen in 83 patients (5.5%), and lastly soft-tissue infection and ischaemia. The overall risk of cancer increases with age, and consequently emergency presentation with malignant disease does too. Demographics and clinical outcome of our study are shown in Table 2 .\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe results showed that the RandomForestClassifier and XGBClassifier algorithm both obtained the highest AUC value (0.97) for the mortality prediction model(see Figure 2). In terms of accuracy rate, F1-Score, precision rate and the Recall rate, the RandomForestClassifier algorithm is more dominant (see Figure 3). Therefore, the RandomForestClassifier algorithm outperformed the other models. The model achieved an Accuracy of 92.7%, F1-Score of 92.7% , Precision of 92.3% and Recall of 92.6% .\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBarplot function plotting was used to visualize the predictors while contributing to the decision-making process. The predictors were arranged into descending order depending on the importance estimates shown in Figure 4. As an initial step to check for predictors\u0026rsquo; importance, all predictors were deemed important, and none was excluded. GNRI\u0026lt;82(factor 10-major risk), GFR\u0026le;44(factor 11-major risk), GFR\u0026ge;45 \u0026le;59(factor 11-moderate risk), GNRI\u0026ge;82 \u0026lt;92(factor 10-moderate risk) and WBC count\u0026gt;15\u0026le;25(factor 17-moderate risk) were ranked as the top five important predictors. The factor 10-major risk was assigned a score of 4; factor 11-major risk was assigned a score of 3; factor 17-moderate risk, factor 11-moderate risk,factor 10-moderate risk and factor 16 were assigned a score of 2; the remaining factors were assigned a score of 1. Based on the feature importance of these 19 factors, a score was derived that ranges from 0 to 26 points (Table 3). Therefore , we propose the simplest possible score, with an integer point scale, as the GEPR. This score has a c-statistic of 0.872 for mortality (95%CI 0.840-0.905). The observed probability of mortality in hospital gradually increased from 0% at a score of 0 to 63.3% at a score of 12 and 100% at a score of 15 (Figure 5). There were no observations for scores greater than 16. \u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe literature provides various definitions for elderly patients, yet there is a lack of clear and definitive criteria. The definition of elderly should not solely rely on chronological age but rather be based on a combination of factors that determine biological aging. The measurement of these factors poses challenges, as clear and objective definitions are not readily available. The term \u0026ldquo;elderly\u0026rdquo; is defined in our study as individuals aged 60 years or above.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe incidence of elderly patients presenting as an emergency with a surgical condition rises proportionally with age. Our previous research had shown that more than 42% of hospitalizations for emergency surgery conditions occur in older adults (60 years or older), and this proportion is expected to rise significantly over the upcoming decade[\u003csup\u003e10]\u003c/sup\u003e. The proportion of hospital admissions attributed to Emergency General Surgery (EGS) ranges from 8% to 26%[\u003csup\u003e11]\u003c/sup\u003e. The incidence of surgical emergencies of the abdomen is significantly higher among the elderly population compared to other demographic groups. In our study, EGS conditions such as appendicitis,obstruction, and biliary disease are responsible for 3 in every 5 emergency surgical hospitalizations. Emergency surgery in geriatric patients has long been acknowledged to result in higher morbidity and mortality rates compared to younger patients, primarily due to advanced age, diminished physiological reserves and increased burden of medical comorbidities.Van Geloven et al[\u003csup\u003e12]\u003c/sup\u003e reported on patients over age 80 who presented to the Emergency Department (ED) with abdominal pain and found that 27% required surgery, with an overall mortality of 17% that increased to 34% among those who required operative intervention. In our study, the patients were categorized into three age groups:60-70,\u0026nbsp;70-80, and \u0026gt;80. The mortality rate in hospital increases proportionally with the age group (6.5% , 11.2%, and 18.4% respectively).\u003c/p\u003e\n\u003cp\u003eRisk prediction in healthcare is a complex yet crucial task, particularly within the context of escalating demand and limited or even diminishing resources available. The utilization of risk modeling is beneficial for enhancing the identification of individuals vulnerable to adverse healthcare outcomes, thereby facilitating resource allocation optimization through targeted allocation of limited resources. The literature on prognostic factors for morbidity and mortality in elderly patients undergoing acute abdominal surgery is insufficient, despite the increasing prevalence of this procedure. Recent systematic reviews have revealed a scarcity of risk-prediction instruments specifically designed for the assessment of geriatric emergency patients, and those that are available have limited diagnostic accuracy[\u003csup\u003e13]\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe estimation of mortality risk for geriatric emergency patients by physicians is predominantly subjective. The utilization of a probability model to predict mortality risk provides a rational and objective approach for quantifying the severity. The previous predictive models, which were all statistical models, relied on multivariate logistic regression analyses. It is worth noting that this approach has been widely used in the field. However, the utilization of the stepwise approach in multivariable regression may lead to model instability and make it highly sensitive to even minor changes in data, such that the inclusion or exclusion of certain observations can significantly alter the model[\u003csup\u003e14]\u003c/sup\u003e. Machine learning and AI are rapidly growing in capability and increasingly applied to model outcomes and complications within medicine[\u003csup\u003e15]\u003c/sup\u003e. The present model is founded upon a machine-learning AI framework. Machine learning is a set of methods that computers use to generate and enhance predictions or behaviors based on data. The AI model is actually a black-box model that cannot be routinely interpreted. Such as the best random forest model in our study, it consist of hundreds of decision trees that \u0026ldquo;vote\u0026rdquo; for predictions.\u0026nbsp;This study effectively addresses the limitation of artificial intelligence in medical prediction by utilizing feature\u0026nbsp;importance\u0026nbsp;to assign scores to factors, thereby rendering complex algorithms more comprehensible. The success of this study is attributed to the extensive independent factor screening conducted by doctors based on\u0026nbsp;clinical practice.\u003c/p\u003e\n\u003cp\u003eThe present study validates our hypothesis regarding the necessity of conducting specific geriatric analysis and deriving a corresponding model, ultimately resulting in the development of the GEPR. The utilization of GEPR model in predicting post-emergency operation outcomes among elderly patients is crucial for healthcare system organization, clinical decision-making, and effective communication with patients and their families. The specific application process as shown in figure 6. In the following discussion, we will briefly introduce the use of GEPR.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e1. Surgical plan\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe anticipated surge in the elderly population, the heightened rate of emergency presentations, and the augmented risks associated with unplanned surgeries present a formidable challenge to healthcare systems, particularly surgical services.It is important to evaluate the patient\u0026rsquo;s condition by emergency physicians.\u0026nbsp;In fact, the suitability of elderly emergency patients for\u0026nbsp;emergency\u0026nbsp;surgery necessitates consideration of multiple factors, encompassing age, comorbidities, preoperative vital signs\u0026nbsp;as well as laboratory and imaging findings.\u0026nbsp;Several scores have been assessed that might predict poor outcome following emergency surgery. Many risk scores strive for excessive generality, while others are excessively disease-specific.\u0026nbsp;The novel model incorporates a diverse range of factors utilizing a specialized algorithm. By employing the scoring system derived from this new model, we can identify low-risk groups eligible for emergency operations, such as those with scores below 8, resulting in an overall hospital mortality rate lower than 20%.\u003c/p\u003e\n\u003cp\u003eLikewise, our work also suggests that the decision to deny surgery should not be solely based on age . Indeed, some studies have shown that selected elderly patients may tolerate major emergency surgery and recover well[\u003csup\u003e16]\u003c/sup\u003e. The presence of advanced age alone does not serve as a contraindication for surgical intervention; however, specific circumstances and medical conditions contribute to an increased surgical risk among the geriatric population. Analysis of GEPR scoring results shows that even in elderly patients over 80 years old , if their preoperative score was less than 6, these elderly patients may had a lower mortality rate after surgery. Notable, too, even though the maximum score that can be achieved using GEPR is 26, none of the patients in our cohort scored higher than 16 and none of those scoring 15 survived.\u0026nbsp;We believe that GEPR will take personalized medicine to an entirely new level.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e2. Non-surgical plan\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEmergency Medicine has traditionally emphasised fast-paced and protocolised care for time-dependent health threats. Although the philosophy of Emergency Medicine, initially based on a singular \u0026lsquo;rule-out worst-case scenario\u0026rsquo; approach applicable to all age groups, gradually recognized the suboptimal nature of this strategy for many aging populations[\u003csup\u003e17]\u003c/sup\u003e. Implementing early intervention strategies, such as deferring surgical procedures for patients with modifiable risk factors, has the potential to mitigate morbidity and mortality rates within the geriatric surgical patient population. In our study, we have incorporated certain modifiable risk factors like preoperative vital signs and GNRI into the GEPR. Therefore, the scale can be used for dynamic assessment in clinical practice.\u0026nbsp;For patients who exhibit hesitancy or are deemed unsuitable for surgical intervention, non-surgical treatment modalities can be employed. During this phase, the utilization of\u0026nbsp;GEPR\u0026nbsp;enables dynamic evaluation of the patient\u0026apos;s condition, and once optimal score results are achieved, surgical procedures or alternative therapeutic approaches can be pursued.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e3. palliative care\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe consideration of palliative care is an alternative for these patients with modifiable risk factors. Avoiding unnecessary prolongation of the dying process, particularly in the ICU, not burdening family;and being aware have all been demonstrated to be of paramount importance to seriously ill surgical patients[\u003csup\u003e18]\u003c/sup\u003e. The main objective of palliative care is to mitigate the suffering associated with invasive treatment in cases where there is a significantly unfavorable prognosis. The provision of expert palliative care services concurrent with disease-directed therapies is known to effectively alleviate patient suffering and avert goal- and value-incongruent care, as well as improve quality of life[\u003csup\u003e19]\u003c/sup\u003e. \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003e4. MAID\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe decriminalization of Medical Assistance in Dying (MAID), which encompasses both euthanasia and assisted suicide, has been implemented in several countries[\u003csup\u003e20]\u003c/sup\u003e. It is often accepted that we may legitimately speak about euthanasia only in cases of patient who has an incurable medical condition and is experiencing constant and unbearable physical or psychological suffering that cannot be alleviated. Considering recent bioethics literature, a distinction can be made between two main approaches to defining the goals of medicine. On the one hand , the objectives of medicine, such as the promotion of health, alleviation of suffering, and enhancement of wellbeing, are commonly acknowledged. On the other hand, philosophers stressing the value of individual autonomy in biomedical ethics are committed to accepting that the proper goals of medicine are ultimately determined by the autonomous decisions of patients[\u003csup\u003e21]\u003c/sup\u003e. In light of this, it becomes evident that physicians can ethically perform euthanasia upon their patients\u0026rsquo; request when they are enduring unbearable physical or psychological suffering.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eA federal government of Canada report on MAID revealed that in 2019, 65% of MAID procedures took place with the involvement of family medicine and emergency medicine[\u003csup\u003e22]\u003c/sup\u003e. Ball et al. reported that availability of MAID can be therapeutic. They found that MAID deaths provide a greater level of patient comfort than even the deaths from the withdrawal of life support in intensive care units; the availability of MAID has improved the outlook of many patients who have not chosen the procedure. It is important to note, the legalization of euthanasia requires not only patients\u0026rsquo; willingness but also an objective evaluation system such as GEPR.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe actual GEPR equation is quite complicated and fit only for being deposited within the bowels of a computer, but once one enters some simple information into the computer, data such as age, sex, and serum creatinine, one can readily get a predicting result. The calculation of GEPR may initially appear more time-consuming compared to other existing risk-classification systems; however, we believe that the development of an automated GEPR calculator could facilitate process optimization in hospitals.\u0026nbsp;we currently intend to develop an online calculator with the aim of enhancing the practicality of GEPR for medical professionals.\u003c/p\u003e\n\u003cp\u003eOur study has some limitations. This is not a multicenter study, implying that further researches including various medical centers and larger sample sizes are needed. If the number of training cohort and validation cohort is further expanded, the model will further improve its performance and stability. Moreover, the GEPR model has not yet been compared with other predictive tools currently utilized in geriatric settings.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThe demographic shift towards an ageing population has transformed the clinical populations for all non-paediatric physicians into one in which older people predominate. The physiological changes associated with aging, including diminished organ function and pharmacokinetic and pharmacodynamic variability, along with compromised functional status, necessitate a more personalized approach to treatment decisions in the elderly patients. As discussed above, the diagnosis of acute abdomen in the elderly remains a clinical challenge due to a very various differential diagnosis;\u003c/p\u003e \u003cp\u003ethe surgical treatment remains the primary choice and approach, however, not all elderly patients may be deemed suitable candidates for surgery and alternative treatments should be considered; the non-operative management should be kept in mind with all its well-known limitations and risks. The GEPR model will help implement geriatric ED interventions to improve emergency care for older patients.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eEGS\u003c/p\u003e\n\u003cp\u003eemergency general surgery\u003c/p\u003e\n\n\u003cp\u003ePOSSUM\u003c/p\u003e\n\u003cp\u003ePhysiological and Operative Severity Score for the enumeration of Mortality and morbidity\u003c/p\u003e\n\n\u003cp\u003eSRS\u003c/p\u003e\n\u003cp\u003eSurgical Risk Scale\u003c/p\u003e\n\n\u003cp\u003ePMP\u003c/p\u003e\n\u003cp\u003ePre-operative Mortality Predictor\u003c/p\u003e\n\n\u003cp\u003eCCI\u003c/p\u003e\n\u003cp\u003eCharlson comorbidity Index\u003c/p\u003e\n\n\u003cp\u003eASA\u003c/p\u003e\n\u003cp\u003eAmerican Society of Anaesthesiology\u003c/p\u003e\n\n\u003cp\u003eAPACHE\u003c/p\u003e\n\u003cp\u003eAcute Physiology and Chronic Health Evaluation\u003c/p\u003e\n\n\u003cp\u003eESS\u003c/p\u003e\n\u003cp\u003eEmergency Surgery Score\u003c/p\u003e\n\n\u003cp\u003eGEPR\u003c/p\u003e\n\u003cp\u003egeriatric emergency perioperative risk\u003c/p\u003e\n\n\u003cp\u003eHR\u003csub\u003emax\u003c/sub\u003e\u003c/p\u003e\n\u003cp\u003eMaximal heart rate\u003c/p\u003e\n\n\u003cp\u003eGNRI\u003c/p\u003e\n\u003cp\u003eGeriatric Nutritional Risk Index\u003c/p\u003e\n\n\u003cp\u003eGFR\u003c/p\u003e\n\u003cp\u003eglomerular filtration rate\u003c/p\u003e\n\n\u003cp\u003eCKD-EPI\u003c/p\u003e\n\u003cp\u003eChronic Kidney Disease Epidemiology Collaboration\u003c/p\u003e\n\n\u003cp\u003eACS\u003c/p\u003e\n\u003cp\u003eacute care surgery\u003c/p\u003e\n\n\u003cp\u003eIQRs\u003c/p\u003e\n\u003cp\u003einterquartile ranges\u003c/p\u003e\n\n\u003cp\u003eAUC\u003c/p\u003e\n\u003cp\u003earea under the receiver operator curve\u003c/p\u003e\n\n\u003cp\u003eED\u003c/p\u003e\n\u003cp\u003eEmergency Department\u003c/p\u003e\n\n\u003cp\u003eMAID\u003c/p\u003e\n\u003cp\u003eMedical Assistance in Dying\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\n\u003cp\u003eThe current study underwent a thorough review and received approval from the Research Ethics Committee of the First Affiliated Hospital of Harbin Medical University. The necessity of obtaining individual patient consent for participation is deemed unnecessary. (IRB-AF/SC-04/02.0).\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\n\u003cp\u003ePart of the data analysed during this study are inclueded in the supplementary information files. Complete datasets are not publicly available, but are available from the corresponding author on reasonable request.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\n\u003cp\u003eNo funding for this study.\u003c/p\u003e\n\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e\u003c/p\u003e\n\n\u003cp\u003eD.X: planning and execution of work, manuscript writing. H. Z: data analysis.\u003c/p\u003e\n\u003cp\u003eJ. R and X. X: data collection. L.H：planning of study and manuscript editing.\u003c/p\u003e\n\u003cp\u003eAll authors discussed and revised the manuscript for submission.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eWorld Health Organization. Populations are getting older. 12 February 2020. https://www.who.int/multi-media/details/populations-are-getting-older.\u003c/li\u003e\n \u003cli\u003eGaitanidis, A; Mikdad, S; Breen, K; et al.The Emergency Surgery Score (ESS) accurately predicts outcomes in elderly patients undergoing emergency general surgery.[J].Am J Surg.2020,220(4):1052-1057\u003c/li\u003e\n \u003cli\u003eShookster, D; Lindsey, B; Cortes, N; et al.Accuracy of Commonly Used Age-Predicted Maximal Heart Rate Equations.[J].Int J Exerc Sci.2020,13(7):1242-1250.\u003c/li\u003e\n \u003cli\u003eHamrick, I; Meyer, F; Perioperative management of delirium and dementia in the geriatric surgical patient.[J].Langenbecks Arch Surg.2013,398(7):947-55.\u003c/li\u003e\n \u003cli\u003eBouillanne, O; Morineau, G; Dupont, C; et al.Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients.[J].Am J Clin Nutr.2005,82(4):777-83\u003c/li\u003e\n \u003cli\u003eJia, Z; El Moheb, M; Nordestgaard, A; et al.The Geriatric Nutritional Risk Index is a powerful predictor of adverse outcome in the elderly emergency surgery patient.[J].J Trauma Acute Care Surg.2020,89(2):397-404\u003c/li\u003e\n \u003cli\u003eLengnan, X, Aiqun, C, Ying, S, et al. The effects of aging on the renal function of a healthy population in Beijing and an evaluation of a range of estimation equations for glomerular filtration rate. Aging (Albany NY). 2021; 13 (5): 6904-6917. doi: 10.18632/aging.202548\u003c/li\u003e\n \u003cli\u003eTanos, P, Ablett, AD, Carter, B, et al. SHARP risk score: A predictor of poor outcomes in adults admitted for emergency general surgery: A prospective cohort study. ASIAN J SURG. 2022; 46 (7): 2668-2674. doi: 10.1016/j.asjsur.2022.10.049\u003c/li\u003e\n \u003cli\u003eLiu, Xun, Wang, Yanni, Wang, Cheng, et al. A new equation to estimate glomerular filtration rate in Chinese elderly population. PloS one. 2013; 8 (11): e79675. doi: 10.1371/journal.pone.0079675\u003c/li\u003e\n \u003cli\u003eXu, D; Yin, Y; Hou, L; et al.A special acute care surgery model for dealing with dilemmas involved in emergency department in China.[J].Sci Rep.2021,11(1):1723\u003c/li\u003e\n \u003cli\u003eBruns, B, Tesoriero, R, Narayan, M, et al. Emergency General Surgery: Defining Burden of Disease in the State of Maryland AM SURGEON. 2020; 81 (8): 829-834. doi: 10.1177/000313481508100825\u003c/li\u003e\n \u003cli\u003evan Geloven, AA, Biesheuvel, TH, Luitse, JS, et al. Hospital admissions of patients aged over 80 with acute abdominal complaints. EUR J SURG. 2000; 166 (11): 866-71. doi: 10.1080/110241500447254\u003c/li\u003e\n \u003cli\u003eO\u0026apos;Caoimh, R, Cornally, N, Weathers, E, et al. Risk prediction in the community: A systematic review of case-finding instruments that predict adverse healthcare outcomes in community-dwelling older adults. MATURITAS. 2015; 82 (1): 3-21. doi: 10.1016/j.maturitas.2015.03.009\u003c/li\u003e\n \u003cli\u003eGrant, SW, Hickey, GL, Head, SJ. Statistical primer: multivariable regression considerations and pitfalls. EUR J CARDIO-THORAC. 2019; 55 (2): 179-185. doi: 10.1093/ejcts/ezy403\u003c/li\u003e\n \u003cli\u003eDesai, GS. Artificial Intelligence: The Future of Obstetrics and Gynecology. J OBSTET GYN INDIA. 2018; 68 (4): 326-327. doi: 10.1007/s13224-018-1118-4\u003c/li\u003e\n \u003cli\u003eSubramanian, A, Balentine, C, Palacio, CH, et al. Outcomes of damage-control celiotomy in elderly nontrauma patients with intra-abdominal catastrophes. AM J SURG. 2010; 200 (6): 783-8; discussion 788-9. doi: 10.1016/j.amjsurg.2010.07.027\u003c/li\u003e\n \u003cli\u003eMooijaart, SP, Carpenter, CR, Conroy, SP. Geriatric emergency medicine-a model for frailty friendly healthcare. AGE AGEING. 2022; 51 (3): doi: 10.1093/ageing/afab280\u003c/li\u003e\n \u003cli\u003eSupiano, KP, McGee, N, Dassel, KB, et al. A Comparison of the Influence of Anticipated Death Trajectory and Personal Values on End-of-Life Care Preferences: A Qualitative Analysis. CLIN GERONTOLOGIST. 2017; 42 (3): 247-258. doi: 10.1080/07317115.2017.1365796\u003c/li\u003e\n \u003cli\u003eSpencer, AL, Miller, PR, Russell, GB, et al. Timing is everything: Early versus late palliative care consults in trauma. J TRAUMA ACUTE CARE. 2022; 94 (5): 652-658. doi: 10.1097/TA.0000000000003881\u003c/li\u003e\n \u003cli\u003eOczkowski Simon J W,Ball Ian,Saleh Carol,et al.The provision of medical assistance in dying: protocol for a scoping review.BMJ open.2017;7 (8)：e017888\u003c/li\u003e\n \u003cli\u003eVarelius Jukka.Illness, suffering and voluntary euthanasia.BIOETHICS.2007;21 (2)：75-83.\u003c/li\u003e\n \u003cli\u003eFirst annual report on medical assistance in dying in Canada, 2019. Ottawa, ON: Government of Canada; 2019\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1 . \u0026nbsp; Prototype of GEPR\u003c/strong\u003e\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eNamed variable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDemographics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eAge \u0026gt; 80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003ePreoperative vital signs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHeart rate \u0026nbsp; \u0026nbsp; \u0026gt;(220-age)*85%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSystolic blood pressure\u0026le;90mmHg or\u003c/p\u003e\n \u003cp\u003eDiastolic blood pressure\u0026le;60mmHg\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eComorbidities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eNeoplastic comorbidity\u003csup\u003e1*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHistory of COPD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eDiabetes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 7\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eVentilator requirement within 48 hours preoperatively\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 8\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eFunctional dependence\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eGNRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;92\u0026nbsp;\u0026le;98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 10-low risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;82 \u0026lt;92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 10-moderate risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 10-major risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eGFR [ml/min/1.73m\u003csup\u003e2\u003c/sup\u003e]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;45\u0026nbsp;\u0026le;59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 11-moderate risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026le;44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 11-major risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eLaboratory values\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eAlkaline phosphatase \u0026gt; 125U/L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 12\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eINR \u0026gt; 1.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 13\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSGOT \u0026gt; 40U/L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 14\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003ePlatelets \u0026lt; 150*10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 15\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSodium \u0026gt; 145 mg/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 16\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eWBC count (\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;4.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 17-low risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026gt;15\u0026le;25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 17-moderate risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026gt;25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 17-major risk\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eImaging techniques\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003ePleural effusion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 18\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003efree air in the abdominal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\"\u003e\n \u003cp\u003efactor 19\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003eTable 2. \u0026nbsp;Demographics and clinical outcome of elderly patients with acute abdomen\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eVariables\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003eNo.(%) of patients\u003c/p\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp;n=1500\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eSex, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e583 (38.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e917 (61.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eASA score , n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eⅡ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e239 (15.9 )\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eⅢ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e971 (64.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eⅣ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e277 (18.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eⅤ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e13 (0.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eAge(years), (mean\u0026plusmn;SD)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e≧60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e1500 (69.8\u0026plusmn;7.456)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e≧60\u0026nbsp;≦70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e904 (64.8\u0026plusmn;7.46)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e＞70\u0026nbsp;≦80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e438 (75.0\u0026plusmn;7.44)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e＞80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e158 (84.7\u0026plusmn;7.49)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eLOS(days),(IQR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e11.1 (4-18)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eICU stay (h), n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e≧48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e440（29.3）\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eSurgery indication,n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003einfection\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e600 (40.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eobstruction\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e348 (23.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eperforation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e277 (18.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003ebleeding\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e127 (8.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eincarcerated hermia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e114 (7.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eischaemia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e18 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eother\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e16 (1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eEGS diagnosis*, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eappendicitis\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e428 (28.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003ebowel/gastric perforation\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e297 (19.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003elarge bowel obstruction\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e202 (13.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003esmall bowel obstruction\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e134 (8.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003ehermia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e127 (8.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003ebiliary tract disease\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e95 (6.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003egastrointestinal bleeding\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e83 (5.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eSoft-tissue infection\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e54 (3.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eischaemia\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e18 (1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eneoplasm\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e235 (15.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003eotherϕ\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e67 (4.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003eMortality, n(%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e≧60\u0026nbsp;≦70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e59 (6.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e＞70\u0026nbsp;≦80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e49 (11.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"33.274647887323944%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"39.08450704225352%\" valign=\"top\"\u003e\n \u003cp\u003e＞80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.640845070422536%\" valign=\"top\"\u003e\n \u003cp\u003e29 (18.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e* some patients had more than 1 diagnosis .\u003c/p\u003e\n\u003cp\u003eϕ postoperative bleeding, diverticulitis,ulcerative colitis,et al.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eSD=standard \u0026nbsp;deviation\u003c/p\u003e\n\u003cp\u003eTable 3 . Development of the the Geriatric Emergency Perioperative Risk Index\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003ePoints*\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDemographics\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eAge\u0026nbsp;\u0026gt;80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003ePreoperative vital signs\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHeart rate \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026gt;(220-age)*85%\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSystolic blood pressure\u0026le;90mmHg or\u003c/p\u003e\n \u003cp\u003eDiastolic blood pressure\u0026le;60mmHg\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eComorbidities\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eNeoplastic comorbidity\u003csup\u003e1*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHistory of COPD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eHypertension\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eDiabetes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eVentilator requirement within 48 hours preoperatively\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eFunctional dependence\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eGNRI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;92\u0026nbsp;\u0026le;98\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;82 \u0026lt;92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;82\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 4\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eGFR [ml/min/1.73m\u003csup\u003e2\u003c/sup\u003e]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;45\u0026nbsp;\u0026le;59\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026le;44\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eLaboratory values\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eAlkaline phosphatase \u0026gt; 125U/L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eINR \u0026gt; 1.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSGOT \u0026gt; 40U/L\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003ePlatelets \u0026lt; 150*10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eSodium \u0026gt; 145 mg/dL\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003eWBC count (\u0026times;10\u003csup\u003e3\u003c/sup\u003e/mm\u003csup\u003e3\u003c/sup\u003e)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026lt;4.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026gt;15\u0026le;25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 2\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026gt;25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eImaging techniques\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003ePleural effusion\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003efree air in the abdominal\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"50%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; 1 \u0026nbsp;\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"geriatrics, perioperative risk model, risk prediction, emergency general surgery","lastPublishedDoi":"10.21203/rs.3.rs-3725510/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3725510/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003ePurpose\u003c/h2\u003e \u003cp\u003eThe present study aimed to develop artificial intelligence (AI)-based model and GEPR, derived from geriatric data, to predict the outcomes of geriatric EGS.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eA retrospectively database of geriatric EGS patients who underwent emergency surgery was used for the development of the AI model and GEPR. The study employed a specialized algorithm, comprising of four sequential steps: scale prototype selection, clinical data collection and collation, AI model development, and GEPR development.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eIn total, 1500 patients were enrolled with mean age of 69.8 years in the study.RandomForestClassifier algorithm outperformed the other AI models. Based on the feature importance, GEPR was derived with a total score range of 0\u0026ndash;26. The GEPR has a c-statistic of 0.872 for mortality in hospital (95%CI 0.840\u0026ndash;0.905). The observed probability of mortality in hospital gradually increased from 0% at a score of 0 to 63.3% at a score of 12 and 100% at a score of 15 .\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eUsing patient-related and technical parameters, an GEPR derived from AI algorithms for prediction of surgical complications in geriatric EGS was developed. GEPR reliably predicts postoperative the mortality in hospital in geriatric EGS patients. Further prospective multicenter trials are needed to externally validate the model developed.\u003c/p\u003e","manuscriptTitle":"Using the GEPR model to predict outcomes of geriatric emergency general surgery","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2023-12-22 16:02:57","doi":"10.21203/rs.3.rs-3725510/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ee48e93f-7990-4530-a378-7572cac537f1","owner":[],"postedDate":"December 22nd, 2023","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2024-02-02T17:59:55+00:00","versionOfRecord":[],"versionCreatedAt":"2023-12-22 16:02:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3725510","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3725510","identity":"rs-3725510","version":["v1"]},"buildId":"WrCJVZZCHTDjtuVLN7oU0","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}