Simple Prediction Score Research Articles

A Simple Prediction Model for Clostridioides difficile Infection: A Hospital-Based Administrative Database Study.

Few prediction scores for Clostridioides difficile infection (CDI), a potentially life-threatening nosocomial diarrhea, combine high accuracy with simplicity. A simple prediction score for routine clinical practice is needed. We conducted a retrospective cohort study of all inpatients aged ≥ 18 at a secondary care hospital in Japan. The derivation and validation cohorts consisted of patients from January 2016 to December 2020 and January 2021 to September 2022, respectively. Demographic and clinical data were retrieved using electronic medical records and an administrative database. The primary outcome was to derive and validate an accurate, simple prediction score for primary hospital-onset CDI. A derived prediction score by logistic regression analysis was calibrated and validated. CDI developed in 102 of 25 517 and 25 of 6259 patients in the derived and validation cohorts (2.7 cases/10 000 patient-days). The derived model for predicting CDI, including antibiotic use, acid suppressant (proton pump inhibitors or vonoprazan) use, Charlson comorbidity index, and Barthel index, yielded c-statistics of 0.89 and 0.82 in the derivation and validation cohort. The model was well calibrated. This simple prediction score enables early medical intervention and modification of treatment plans to reduce the risk of developing primary hospital-onset CDI.

Antiseizure Medication Withdrawal, Risk of Epilepsy, and Longterm EEG Trends in Acute Symptomatic Seizures or Epileptic EEG Patterns.

Patients with acute symptomatic seizures (ASyS) and acute epileptiform findings on EEG are common. They are often prescribed long-term antiseizure medications (ASMs); it is unknown whether or when this is necessary. Primary outcome was late unprovoked seizure occurrence and association with ASM taper. The secondary outcome was EEG pattern evolution over time. This is a retrospective cohort study of patients from 2015 to 2021 with ASyS (clinical or electrographic) and/or epileptiform findings on index hospitalization EEGs who were discharged on ASMs and had subsequent follow-up including an outpatient EEG at Yale New Haven Hospital. All patients were seen in our postacute symptomatic seizure (PASS) clinic after hospital discharge. We also developed a simple predictive score, Epilepsy-PASS (EPI-PASS), using variables independently associated with seizure recurrence based on stepwise regression; each of the 3 identified variables was assigned a score of 0 (absent) or 1 (present), for a total score of 0-3. Of 190 patients screened, 58 were excluded, leaving a final cohort of 112 patients. Twenty-four percent (27/112) patients developed a late unprovoked seizure (i.e., epilepsy). Independent predictors of epilepsy were persistence of epileptiform abnormalities on follow-up EEGs [56% developed epilepsy vs 19% without, 0.002, OR 7.18 (1.36-37.88)], clinical ASyS [32% vs 13%, p = 0.002, OR 7.45 (2.31-54.36)], and cortical involvement on imaging [40% vs 11%, p = 0.003, OR 7.63 (1.96-29.58)]. None of the 23 patients with none of these predictors (0 points on EPI-PASS) developed epilepsy, vs 13% with 1 predictor (EPI-PASS = 1) and 46% with 2 or 3 predictors (EPI-PASS = 2-3) at 1-year follow-up. ASM taper was not associated with seizure recurrence. Abnormal EEG findings in the index hospitalization usually resolved [54/69 (78%) patients] on subsequent EEGs. Most patients with clinical ASyS or acute epileptiform EEG findings do not require long-term ASMs. Index hospitalization EEG findings usually resolve, but if they do not, there is a >50% chance of developing epilepsy. Other predictors of epilepsy are cortical involvement on imaging and clinical ASyS. A simple 4-point scale using these 3 predictors (EPI-PASS) may help predict the risk of developing epilepsy but requires independent validation.

A contemporary simple risk score for prediction of severe acute kidney injury after heart transplantation.

The aim of this study was to develop a simple risk score to estimate severe acute kidney injury (AKI) risk based on a large contemporary heart transplantation (HT) cohort. From 1 January 2015 to 31 December 2021, all consecutive HT recipients in our institute were included and analysed for the occurrence of AKI within the first seven postoperative days. Patients transplanted between 2015 and 2019 comprised the derivation cohort, and those transplanted between 2020 and 2021 formed the validation cohort. The primary endpoint was severe AKI (AKI stages 2-3). The impact of severe AKI on 90day mortality was also evaluated. Overall, 430 HT patients were included in the derivation cohort, and 108 were included in the validation cohort. Postoperative AKI occurred in 388 (72%) patients, including 162 (30%) severe AKI. The risk of 90day mortality significantly increased in patients with severe-AKI. Seven independent predictors of severe AKI were found in the derivation cohort, including recipients' body mass index, history of diabetes, anaemia, preoperative inotropes, estimated glomerular filtration rate, cardiopulmonary bypass duration and intraoperative red blood cell transfusion. The occurrence of severe AKI increased gradually from the lowest to the highest of the four risk score groups in the derivation and validation cohort. The scoring prediction model showed a highly acceptable discriminating power for severe-AKI [C statistic: 0.76, 95% confidence interval (CI): 0.71-0.80 for derivation cohort; C statistic: 0.79, 95% CI: 0.71-0.89 for validation cohort]. A contemporary simple risk score based on available variables from patients undergoing HT can accurately discriminate the risk of severe AKI.

Home Blood Pressure Variability Risk Prediction Score for Cardiovascular Disease Using Data From the J-HOP Study.

Home blood pressure (BP) is more closely associated with cardiovascular event risk than office BP, but cardiovascular risk prediction based on home BP variability is lacking. This study developed a simple cardiovascular event prediction score, including home BP variability data, from the J-HOP study (Japan Morning Surge-Home Blood Pressure). The J-HOP study extended follow-up from December 2017 to May 2018 generated the study data set (4231 patients). Cardiovascular events included fatal/nonfatal stroke (n=94), coronary heart disease (n=124), heart failure (n=42), and aortic dissection (n=8). Cox proportional hazards models were used to predict overall cardiovascular risk. Potential covariates included age, sex, body mass index, smoking, history of diabetes, statin use, history of cardiovascular disease, total cholesterol:high-density lipoprotein cholesterol ratio, office systolic BP (SBP), mean of morning-evening average (MEave), home SBP, and average real variability of MEave home SBP. A risk score and models were constructed, and model performance was assessed. Model performance was best when average real variability of MEave SBP was included (C statistic, 0.760). The risk score assigns points for age (5-year bands), sex, cardiovascular disease history, high-density lipoprotein cholesterol, mean MEave home SBP, and average real variability of MEave home SBP. Estimated 10-year cardiovascular risk ranged from ≤0.6% (score ≤0) to >32% (score ≥26). Calibration 2 statistics values for the model (2.66) and risk score (5.29) indicated excellent goodness of fit. This simple cardiovascular disease prediction algorithm, including day-by-day home BP variability, could be used as part of a home BP-centered approach to hypertension management in clinical practice.

Risk factors and prediction score for new-onset diabetes mellitus after liver transplantation.

New-onset diabetes mellitus is a frequent and severe complication arising after liver transplantation (LT). We aimed to identify the risk factors for new-onset diabetes mellitus after liver transplantation (NODALT) and to develop a risk prediction score system for relevant risks. We collected and analyzed data from all recipients who underwent liver transplantation at the First Affiliated Hospital of Xi'an Jiaotong University. The OR derived from a multiple logistic regression predicting the presence of NODALT was used to calculate the risk prediction score. The performance of the risk prediction score was externally validated in patients who were from the CLTR (China Liver Transplant Registry) database. A total of 468 patients met the outlined criteria and finished the follow-up. Overall, NODALT was diagnosed in 115 (24.6%) patients. Age, preoperative impaired fasting glucose (IFG), postoperative fasting plasma glucose (FPG), and the length of hospital stay were significantly associated with the presence of NODALT. The risk prediction score includes age, preoperative IFG, postoperative FPG, and the length of hospital stay. The risk prediction score of the area under the receiver operating curve was 0.785 (95% CI: 0.724-0.846) in the experimental population and 0.782 (95% CI: 0.708-0.856) in the validation population. Age at the time of transplantation, preoperative IFG, postoperative FPG, and length of hospital stay were independent predictive factors of NODALT. The use of a simple risk prediction score can identify the patients who have the highest risk of NODALT and interventions may start early.

HERMES-24 Score Derivation and Validation for Simple and Robust Outcome Prediction After Large Vessel Occlusion Treatment.

Clinicians need simple and highly predictive prognostic scores to assist practical decision-making. We aimed to develop a simple outcome prediction score applied 24 hours after anterior circulation acute ischemic stroke treatment with endovascular thrombectomy and validate it in patients treated both with and without endovascular thrombectomy. Using the HERMES (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials) collaboration data set (n=1764), patients in the endovascular thrombectomy arm were divided randomly into a derivation cohort (n=430) and a validation cohort (n=441). From a set of candidate predictors, logistic regression modeling using forward variable selection was used to select a model that was both parsimonious and highly predictive for modified Rankin Scale (mRS) ≤2 at 90 days. The score was validated in validation cohort, control arm (n=893), and external validation cohorts from the ESCAPE-NA1 (Efficacy and Safety of Nerinetide for the Treatment of Acute Ischaemic Stroke; n=1066) and INTERRSeCT (Identifying New Approaches to Optimize Thrombus Characterization for Predicting Early Recanalization and Reperfusion With IV Alteplase and Other Treatments Using Serial CT Angiography; n=614). In the derivation cohort, we selected 2 significant predictors of mRS ≤2 (National Institutes of Health Stroke Scale score at 24 hours and age [β-coefficient, 0.34 and 0.06]) and derived the HERMES-24 score: age (years)/10+National Institutes of Health Stroke Scale score at 24 hours. The HERMES-24 score was highly predictive for mRS ≤2 (c-statistic 0.907 [95% CI, 0.879-0.935]) in the derivation cohort. In the validation cohort and the control arm, the HERMES-24 score predicts mRS ≤2 (c-statistic, 0.914 [95% CI, 0.886-0.944] and 0.909 [95% CI, 0.887-0.930]). Observed provability of mRS ≤2 ranged between 3.1% and 3.4% when HERMES-24 score ≥25, while it ranged between 90.6% and 93.0% when HERMES-24 score <10 in the derivation cohort, validation cohort, and control arm. The HERMES-24 score also showed c-statistics of 0.894 and 0.889 for mRS ≤2 in the ESCAPE-NA1 and INTERRSeCT populations. The post-treatment HERMES-24 score is a simple validated score that predicts a 3-month outcome after anterior circulation large vessel occlusion stroke regardless of intervention, which helps prognostic discussion with families on day 2.

Novel Predictive Scoring System for Bone Union Rate After Conservative Management of Lumbar Spondylolysis.

A single-center retrospective cohort study. To develop a predictive scoring system for bone union after conservative treatment of lumbar spondylolysis and assess its internal validity. Lumbar spondylolysis, a common stress fracture in young athletes, is typically treated conservatively. Predicting bone union rates remains a challenge. This study included patients aged ≤18 years with lumbar spondylolysis undergoing conservative treatment. A multivariable logistic regression analysis was used to develop a scoring system containing six factors: sex, age, lesion level, main side stage of the lesion, contralateral side stage of the lesion, and spina bifida occulta. The predictive scoring system was internally validated from the receiver operating characteristic (ROC) curve using bootstrap methods. The final analysis included 301 patients with 416 lesions, with an overall bone union rate of 80%. On multivariable analysis, the main and contralateral stages were identified as factors associated with bone union. The predictive scoring system was developed from the main side stage score (prelysis, early=0, progressive stage=1) and the contralateral side stage score (none=0, prelysis, early, progressive stage=1, terminal stage=3). The area under the curve was 0.855 (95% confidence interval: 0.811-0.896) for the ROC curve, showing good internal validity. The predicted bone union rates were generally consistent with the actual rates. A simple predictive scoring system was developed for bone union after conservative treatment of lumbar spondylolysis, based on the stage of the lesion on the main and contralateral sides. The predicted bone union rate was approximately 90% for a total score of 0-1 and ≤30% for a score of 3-4. This system demonstrated good internal validity, suggesting its potential as a useful tool in clinical decision making for the management of spondylolysis.

Development of a Simple Scoring System for Predicting Discharge Safety from the Medical ICU to Low-Acuity Wards: The Role of the Sequential Organ Failure Assessment Score, Albumin, and Red Blood Cell Distribution Width.

Despite advancements in artificial intelligence-based decision-making, transitioning patients from intensive care units (ICUs) to low-acuity wards is challenging, especially in resource-limited settings. This study aimed to develop a simple scoring system to predict ICU discharge safety. We retrospectively analyzed patients admitted to a tertiary hospital's medical ICU (MICU) between July 2016 and December 2021. This period was divided into two phases for model development and validation. We identified risk factors associated with unexpected death within 14 days of MICU discharge and developed a predictive scoring system that incorporated these factors. We verified the system's performance using validation data. In the development cohort, 522 patients were discharged from the MICU, and 42 (8.04%) died unexpectedly. In multivariate analysis, the Sequential Organ Failure Assessment (SOFA) score (odds ratio [OR] 1.26, 95% confidence interval [CI] 1.13-1.41), red blood cell distribution width (RDW) (OR 1.20, 95% CI 1.07-1.36), and albumin (OR 0.37, 95% CI 0.16-0.84) were predictors of unexpected death. Each variable was assigned a weighted point in the scoring system, and the area under the curve (AUC) was 0.788 (95% CI 0.714-0.855). The scoring system was performed using an AUC of 0.738 (95% CI 0.653-0.822) in the validation cohort of 343 patients with 9.62% of unexpected deaths. When a cut-off of 0.032 was applied, a sensitivity and a specificity of 81.8% and 55.2%, respectively, were achieved. This simple bedside predictive score for ICU discharge uses the SOFA score, albumin level, and RDW to aid in timely decision-making and optimize critical care facility allocation in resource-limited settings.

Estimating Efficacy of Conversion Therapy on Patients with Initially Unresectable Colorectal Cancer Liver Metastases by using MRI: Development of a Predictive Score

The conversion success rate (CSR) has crucial implication for clinical outcomes of initially unresectable colorectal liver metastases (CRLM) following conversion therapy. This study aimed to develop a simple predictive scoring model for identifying CSR according to baseline magnetic resonance imaging (MRI) features, and confirm its performance and prognostic significance in a validation cohort. A total of 155 consecutive patients with initially unresectable CRLM were retrospectively reviewed in the study. A simple MRI-based predictive scoring model for identifying CSR was developed in the development cohort (n=104) by using multivariable logistic regression analyzes. The diagnostic performance was evaluated for the predictive score. Thereafter, patients in the validation cohort (n=51) were stratified into groups with predicted high CSR or low CSR according to the score. The progression-free survival (PFS) and overall survival (OS) were compared between two groups using the log-rank test. The predictive score of CSR, named mrNISE, incorporated the number of CRLM ≥10, the largest size ≥50mm, poorly defined tumor-liver interface, and peritumoral enhancement. The AUC of the mrNISE score was 0.845 for the development cohort and 0.776 for the validation cohort. According to the score, patients with predicted high CSR had better PFS and OS than those with low CSR in both development and validation cohorts. The predictive score demonstrated great performance for identifying CSR of initially unresectable CRLM. Stratifying patients by the score, personalized treatment goals can be formulated before conversion therapy to improve clinical prognosis and reduce adverse events caused by ineffective treatment.

A Simple Score for Predicting Paroxysmal Atrial Fibrillation in Patients with Embolic Stroke of Undetermined Source in a Tunisian Cohort Study.

The annualized recurrent stroke rate in patients with Embolic Stroke of Undetermined Source (ESUS) under antiplatelet therapy is around 4.5%. Only a fraction of these patients will develop atrial fibrillation (FA), to which a stroke can be attributed retrospectively. The challenge is to identify patients at risk of occult AF during follow-up. This work aims to determine clinical factors and electrocardiographic and ultrasound parameters that can predict occult AF in patients with ESUS and build a simple predictive score applicable worldwide. This is a single-center, registry-based retrospective study conducted at the stroke unit of Sahloul University Hospital, Sousse, Tunisia, between January 2016 and December 2020. Consecutive patients meeting ESUS criteria were monitored for a minimum of one year, with a standardized follow-up consisting of outpatient visits, including ECG every three months and a new 24-hour Holter monitoring in case of palpitations. We performed multivariate stepwise regression to identify predictors of new paroxysmal AF among initial clinical, electrocardiographic (ECG and 24-hour Holter monitoring) and echocardiographic parameters. The coefficient of each independent covariate of the fitted multivariable model was used to generate an integerbased point-scoring system. Three hundred patients met the criteria for ESUS. Among them, 42 (14%) patients showed at least one episode of paroxysmal AF during a median follow-up of two years. In univariate analysis, age, gender, coronary artery disease, history of ischemic stroke, higher NIHSS at admission and lower NIHSS at discharge, abnormal P-wave axis, prolonged P-wave duration, premature atrial contractions (PAC) frequency of more than 500/24 hours, and left atrial (LA) mean area of more than 20 cm2 were associated with the risk of occurrence of paroxysmal AF. We proposed an AF predictive score based on (1.771 x NIHSS score at admission) + (10.015 x P-wave dispersion; coded 1 if yes and 0 if no) + (9.841x PAC class; coded 1 if ≥500 and 0 if no) + (9.828x LA class surface; coded 1 if ≥20 and 0 if no) + (0.548xNIHSS score at discharge) + 0.004. A score of ≥33 had a sensitivity of 76% and a specificity of 93%. In this cohort of patients with ESUS, NIHSS at both admission and discharge, Pwave dispersion, PAC≥500/24h on a 24-hour Holter monitoring, and LA surface area≥20 cm2 provide a simple AF predictive score with very reasonable sensitivity and specificity and is applicable almost worldwide. An external validation of this score is ongoing.

Abstract TP234: A Predictive Model for Early Neurologic Deterioration in Small Subcortical Infarcts

Introduction: Roughly a quarter of patients with an acute small subcortical infarct may develop early neurologic deterioration (END) hallmarked by progressive or stuttering neurologic deficits. Importantly, END results in worse short- and long-term functional outcomes. However, despite its importance, there is an inability to identify those at risk for END, thereby creating a barrier to designing future treatment and preventative trials. We aimed to generate a simple predictive scoring model for END in patients with an acute SSI. Methods: We performed a retrospective chart review of all adult patients admitted to Rhode Island Hospital with an acute SSI between 2015 and 2022. Clinical and demographic data were extracted. Radiographic variables included: 1)infarct signal intensity ratio ≤ 1.15 (analogous to the MR-WITNESS trial), 2) infarct morphology, and 3) infarct size. The primary outcome was END, defined as a ≥ 2-point increase in NIHSS or progressive disabling stroke-related deficits while hospitalized attributed to the index acute SSI. Standard descriptive statistical techniques were used. We then employed a backward stepwise logistic regression with a 0.05 threshold to discern factors associated with END, with each factor contributing one point to the model. An external validation cohort was ascertained from Columbia University. The area under the curve (AUC) was used to assess the utility of the predictive model. Results: A total of 252 patients met inclusion criteria. Demographic and clinical histories did not differ amongst the primary or validation cohorts. A signal intensity ratio ≤ 1.15, “finger-like” SSI morphology, and the presence of SSI on &gt; 3 axial diffusion weighted imaging slices were predictive of END. Allocating one point for each domain (minimum 0, maximum 3) yielded an AUC of 0.88. A score of 2 points captured 91.6% of all END cases. Applying this model to the external validation cohort (n=43) yielded an AUC of 0.68. Conclusion: In sum, we found that a radiographic ordinal model including signal intensity ratio ≤ 1.15, “finger-like” SSI morphology, and infarct presence on &gt; 3 axial diffusion weight imaging slices was helpful in predicting END. Future studies are needed to expand the external cohort size to test this model’s validity.

Abstract WP248: Developing and Validating Post-Treatment HERMES Score to Predict Outcome From Anterior Circulation Large Vessel Occlusion Stroke: A Meta-Analysis of Individual Data From 7 Randomized Clinical Trials

Introduction: Clinicians need simple and highly predictive prognostic scores to assist practical decision-making and family discussion. We aimed to develop and validate a simple prediction score applied at 24 hours to assist prognostication in patients with anterior circulation ischemic stroke due to large vessel occlusion. Methods: Using the HERMES collaboration dataset (n = 1764), patients in the endovascular therapy (EVT) arm were divided randomly into a derivation cohort (n = 430) and a validation cohort (n = 441). From a set of candidate predictors, forward selection using c-statistics was employed to select a model which was both parsimonious and highly predictive for modified Rankin Scale (mRS) ≤2 at 90 days. The score was validated in both the EVT validation cohort and in the control arm (n = 893) for mRS ≤2 and ≤3. Results: In the derivation cohort, two significant predictors of mRS ≤2 (National Institutes of Health Stroke Scale [NIHSS] score at 24 h and age [β-coefficient 0.34 and 0.06]) were selected. Incorporating other variables did not much improve model performance. Among models with different weights, we derived the HERMES score: age (years)/10 + NIHSS score at 24 h, based on model performance and simplicity. The HERMES score was highly predictive for mRS ≤2 in the derivation cohort, validation cohort-EVT, and control arm (c-statistics 0.907, 0.914, and 0.909, respectively). Evaluation of the score against mRS ≤3 as an alternative outcome yielded similar results (c-statistics 0.911, 0.903, and 0.885). Among 435 subjects (24.7%) with HERMES score ≥25, the observed probability was 3.1-3.4% for mRS ≤2 and 9.4-16.7% for mRS ≤3 in the derivation cohort, validation cohort-EVT, and control arm (Figure). Conclusions: The HERMES score is a simple validated score to predict outcomes in patients with anterior circulation large vessel occlusion ischemic stroke regardless of intervention. HERMES score should be helpful in prognostic discussion with families on day two.

A Simple Bacteremia Score for Predicting Bacteremia in Patients with Suspected Infection in the Emergency Department: A Cohort Study.

Bacteremia is a life-threatening condition that has increased in prevalence over the past two decades. Prompt recognition of bacteremia is important; however, identification of bacteremia requires 1 to 2 days. This retrospective cohort study, conducted from 10 November 2014 to November 2019, among patients with suspected infection who visited the emergency department (ED), aimed to develop and validate a simple tool for predicting bacteremia. The study population was randomly divided into derivation and development cohorts. Predictors of bacteremia based on the literature and logistic regression were assessed. A weighted value was assigned to predictors to develop a prediction model for bacteremia using the derivation cohort; discrimination was then assessed using the area under the receiver operating characteristic curve (AUC). Among the 22,519 patients enrolled, 18,015 were assigned to the derivation group and 4504 to the validation group. Sixteen candidate variables were selected, and all sixteen were used as significant predictors of bacteremia (model 1). Among the sixteen variables, the top five with higher odds ratio, including procalcitonin, neutrophil-lymphocyte ratio (NLR), lactate level, platelet count, and body temperature, were used for the simple bacteremia score (model 2). The proportion of bacteremia increased according to the simple bacteremia score in both cohorts. The AUC for model 1 was 0.805 (95% confidence interval [CI] 0.785-0.824) and model 2 was 0.791 (95% CI 0.772-0.810). The simple bacteremia prediction score using only five variables demonstrated a comparable performance with the model including sixteen variables using all laboratory results and vital signs. This simple score is useful for predicting bacteremia-assisted clinical decisions.

Reverse Shock Index multiplied by Glasgow Coma Scale score as a point-of-care severity assessment for initial trauma management: a nationwide cohort study

BackgroundSevere trauma patients often require emergent interventions, such as massive transfusion, resuscitative procedures, and surgical procedures, and consume considerable human and medical resources. However, few practical indices can be easily used for emergent interventions. In recent years, it has become clear that rSIG (Reverse Shock Index multiplied by Glasgow Coma Scale [GCS] score), which can be easily calculated from vital signs, is a promising predictor of mortality. However, it is unclear whether rSIG is useful for emergent interventions. MethodsData collected by the Japan Trauma Data Bank for adult patients admitted directly from the scene of trauma between April 2019 and December 2020 were analysed. The outcomes were massive transfusion, resuscitative procedures, surgical procedures and emergent interventions. Emergent interventions were defined as the composite outcome of massive transfusion, resuscitative procedures, and surgical procedures. The ability of rSIG to predict massive transfusion was compared with that of the ABC score and FASILA score by receiver-operating characteristic curve analysis. The ability of rSIG to predict resuscitative and surgical procedures was compared with that of the Shock Index (SI), GCS, Triage Revised Trauma score (T-RTS), and Previous Simple Prediction (PSP) score. The ability of rSIG to predict emergent interventions was compared with that of T-RTS, PSP, ABC, and FASILA. In addition to rSIG, rSIM (Reverse Shock Index multiplied by best motor response score) was also analysed as a supplement. ResultsThe study included 32,201 patients, 6,371 of whom required emergent interventions. The area under the receiver-operating characteristic curve (AUROC) for massive transfusion was highest for rSIG (0.846 [95% confidence interval 0.832–0.859]) and significantly higher for rSIG than for rSIM, ABC and FASILA (all p<0.0001). AUROCs for resuscitative and surgical procedures were highest for rSIG (0.777 [0.769–0.785] and 0.731 [0.720–0.741], respectively) and significantly higher than those for rSIM, SI, GCS, T-RTS, and PSP (all p<0.0001). The AUROC for emergent interventions was highest for rSIG (0.760 [0.753–0.768]) and significantly higher for rSIG than for rSIM, T-RTS, PSP, ABC, or FASILA (all p<0.0001). ConclusionsrSIG is a simple and effective point-of-care predictor of emergent interventions during initial management of trauma.

A simple prediction score for postoperative mortality after decortication.

Decortication of the lung, either by video-thoracoscopy or thoracotomy is potentially a morbid procedure and has significantly higher mortality compared with other major thoracic procedures. Much of this difference can be attributed to other significant comorbidities and to the non-elective nature of the surgery. Our primary goal was to recognize the preoperative unique characteristics of patients who had postoperative mortality within the first 30 days. Our secondary goal was to build a score system to calculate the odds of death after decortication. Patients who had undergone either partial or total pulmonary decortication were retrospectively identified from the 2015-2017 databases of the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) and were employed for this analysis. Multivariable regression models were used to evaluate the possible association of multiple risk factors with postoperative death. Factors that remained significant in the multivariable regression analysis were used to develop the Decortication Prognostic Score (DPS). The final study population consisted of 2,315 patients. The overall observed mortality rate was 5.6%. The greatest independent risk factor for increased 30-day mortality in multivariable logistic regression analysis was disseminated cancer, followed by age ≥65 years, ventilator dependence, active hemodialysis, open wound or wound infection, partially or totally dependent preoperative functional status, preoperative systemic inflammatory response syndrome (SIRS), sepsis or septic shock, congestive heart failure (CHF), preoperative need for blood transfusion, dyspnea, and chronic obstructive pulmonary disease (COPD). Afterwards, we developed a prognostic score for calculating the odds of postoperative death. The total score was associated with a stepwise higher risk of postoperative death after decortication. Patients with a score of 1 had an associated mortality of 1.1% [odds ratio (OR): 2, 95% confidence interval (CI): 0.43-9.32, P=0.375], patients with scores 2-3 had an associated mortality of 6.6% (OR: 12.5, 95% CI: 3.04-51.36, P<0.001), and patients with scores ≥4 had an associated mortality of 27.1% (OR: 65.8, 95% CI: 15.86-273.2, P<0.001). Preoperative factors can predict postoperative mortality after decortication. DPS may help guide surgeons with bedside decision making and heighten awareness to patients most likely to be at risk for 30-day re-intubation, failure to wean from ventilator, surgical site infections, prolong length of stay and higher mortality after decortication.

Abstract 245: A Validated Score to Predict Poor Neurological Outcome After 72 Hours of Refractory VF/VT in Out of Hospital Cardiac Arrest Patients Requiring Extracorporeal Cardiopulmonary Resuscitation

Introduction: The challenge with neuroprognostication is finding a multimodal prediction model that yields the lowest false positive rate for a poor outcome. Objectives: To develop and validate a score that accurately predicts poor neurological outcome after 72 hours of ECPR initiation. Methods: Single-center, retrospective cohort study included 361 OHCA patients who underwent ECPR at the University of Minnesota between December 2015-January 2023. In a 50% random modeling sample, we obtained a derivation (n=180) and validation sample (n=181). The primary outcome was poor neurological outcome (defined as CPC 3-5) at hospital discharge. Neurological predictors were neuro specific enolase (NSE)&gt;3ULN on day 1, myoclonus status epilepticus on day 1-3, neurological pupillary index (NPI)&lt;2 on day 3-6, Glasgow motor=3 on day 3-6, highly malignant pattern on EEG (hmEEG) on day 3-6, CT with anoxic brain injury on day 3, and absent of brainstem reflexes on day 3-6. Univariate and multivariate analysis were performed on each predictor. A simple prediction score tool was constructed and evaluated through ROC analysis. A youden index was used for optimal cut-off. Results: Of 361 patients, 170 (47%) survived to hospital discharge. Of these, 85% (145/170) survived with favorable neurological outcome. The final multivariate logistic regression model in the derivation sample retained eight predictors: Absence of witnessed status and bystander CPR; downtime≥ 7min, lactate≥ 9mmol, low flow≥ 55min, NSE&gt;3ULN, hmEEG, CT with anoxic brain injury and absence of brainstem reflexes, which were combined to develop an 8-point risk scoring model Minnesota OHCA using ECPR (MECPR). Model discrimination had an AUC of 0.87 [95%CI 0.82-0.92] in the derivation sample and 0.86 [95% CI 0.85-0.92] in the validation sample, and calibration (goodness of fit p=0.04). MECPR≥6 predicted poor neurological outcome with a sensitivity ~21%, specificity 100% and a FPR 0%. A MECPR ≥4 and ≥6 had an 89% and 97% probability of poor neurological outcome, respectively. Conclusions: A simple, accurate prediction scoring tool (MECPR) based on a multimodal approach after 72hrs of refractory cardiac arrest provides estimates of probability of poor neurological outcome among patients treated with ECPR.

Risk Factors and Predictive Model for Mortality of Hospitalized COVID-19 Elderly Patients from a Tertiary Care Hospital in Thailand.

Background: Early detection of elderly patients with COVID-19 who are at high risk of mortality is vital for appropriate clinical decisions. We aimed to evaluate the risk factors associated with all-cause in-hospital mortality among elderly patients with COVID-19. Methods: In this retrospective study, the medical records of elderly patients aged over 60 who were hospitalized with COVID-19 at Thammasat University Hospital from 1 July to 30 September 2021 were reviewed. Multivariate logistic regression was used to identify independent predictors of mortality. The sum of weighted integers was used as a total risk score for each patient. Results: In total, 138 medical records of patients were reviewed. Four identified variables based on the odds ratio (age, respiratory rate, glomerular filtration rate and history of stroke) were assigned a weighted integer and were developed to predict mortality risk in hospitalized elderly patients. The AUROC of the scoring system were 0.9415 (95% confidence interval, 0.9033-0.9716). The optimized scoring system was developed and a risk score over 213 was considered a cut-off point for high mortality risk. Conclusions: A simple predictive risk score provides an initial assessment of mortality risk at the time of admission with a high degree of accuracy among hospitalized elderly patients with COVID-19.

Pancreaticogastrostomy versus Pancreaticojejunostomy and the Proposal of a New Postoperative Pancreatic Fistula Risk Score.

Despite the substantial decrease in mortality rates following a pancreaticoduodenectomy to less than 5%, morbidity rates remain significant, reaching even 73%. Postoperative pancreatic fistula is one of the most frequent major complications and is significantly associated with other complications, including patient death. Currently, there is no consensus regarding the ideal type of pancreatic anastomosis, as the question of the choice between a pancreaticogastrostomy and pancreaticojejunostomy is still open. Furthermore, worldwide implementation of an ideal pancreatic fistula risk prediction score is missing. Our study found several significant predictive factors for the postoperative occurrence of fistulas, such as the soft consistency of the pancreas, non-dilated Wirsung duct, important intraoperative blood loss, other perioperative complications, preoperative patient hypoalbuminemia, and patient weight loss. Our study also revealed that for patients who exhibit fistula risk factors, pancreaticogastrostomy demonstrates a significantly lower pancreatic fistula rate than pancreaticojejunostomy. The occurrence of pancreatic fistulas has been significantly associated with the development of other postoperative major complications, and patient death. As the current pancreatic fistula risk scores proposed by various authors have not been consensually validated, we propose a simple, easy-to-use, and sensitive score for the risk prediction of postoperative pancreatic fistula occurrence based on important predictors from statistical analyses that have also been found to be significant by most of the reported studies. The new pancreatic fistula risk score proposed by us could be extremely useful for improved therapeutic management of cephalic pancreaticoduodenectomy patients.

Short-term mortality prediction using a combination of clinical and CT features: Refining the prognosis of critically ill patients in shock

PurposeTo assess the predictive value of combining CT and clinical findings for predicting 10-day mortality in critically ill patients in shock. Materials and methodsFrom January 1, 2018, to December 31, 2021, 289 consecutives critically ill patients in shock who underwent a contrast enhanced CT were included. Variables at the time of the CT were retrospectively extracted from medical charts. CT examinations were blindly analyzed by two independent radiologists. Multivariable analysis was performed, combining clinical and CT features. A simple survival score for 10-day mortality prediction was built and validated in a further independent external cohort of 70 patients. Results10-day mortality rate was 135/289 (47%) in the study sample. At multivariate analysis, catecholamine infusion (OR = 2.11; 95%CI [1.21–4.18], P = 0.011), lactates level > 5 mmol/l (OR = 3.54; 95%CI [1.94–6.54], P < 0.001); total bilirubin > 50 mg/l (OR = 1.79 CI 95% [1.03–3.13], P = 0.039); small bowel dilation (OR = 1.82; 95%CI [1.01–3.32], P = 0.047); diffuse kidney infarction (OR = 2.76; 95%CI [1.26–6.37], P = 0.013) and superior mesentery artery < 5 mm (OR = 1.96; 95%CI [1.10–3.49], P = 0.021) were associated with 10-days mortality. The AUC of the combined model was 0.79; 95%CI [0.74–0.85] in the study sample and 0.87; 95%CI [0.71–0.91] in the validation cohort. ConclusionThe combination of CT imaging features and clinical data should emerge as a novel approach to predict short-term mortality in critically ill patients in shock.

A predictive score for severity in patients with confirmed dengue fever in a tertiary care hospital in Kerala, India.

The study aimed to identify predictors of severe dengue during the 2017 epidemic and to develop and validate a simple predictive score for severity. A retrospective analytical study was conducted using clinical and laboratory data from adult dengue patients with a confirmed microbiological diagnosis. The study included patients who presented to a tertiary care centre in Kerala, India, during the febrile phase (≤4d) between June 2017 and February 2019. Using appropriate statistical tests, we derived predictors of severe disease and computed a risk score model. Of the 153 patients (mean age 50±17y; 64% males), 31 (20%) had severe dengue and 4 (3%) died. Petechial lesions, hypoalbuminemia (<3.5 g/dl), elevated alanine aminotransferase (>40IU/l) and urea >40IU/l were significant predictors. Our scoring system (cut-off: 2) showed excellent performance, with an area under the receiver operating characteristics curve of 0.9741, sensitivity of 100%, specificity of 96% and accuracy of 98%. The risk score was secondarily validated on 48 patients hospitalized from March 2019 to June 2019. Our scoring system is easy to implement and will help primary healthcare practitioners in promptly identifying severe dengue cases upon hospital presentation.