Predicting dry weight change in Hemodialysis patients using machine learning.

BACKGROUND: Machine Learning has been increasingly used in the medical field, including managing patients undergoing hemodialysis. The random forest classifier is a Machine Learning method that can generate high accuracy and interpretability in the data analysis of various diseases. We attempted to apply Machine Learning to adjust dry weight, the appropriate volume status of patients undergoing hemodialysis, which requires a complex decision-making process considering multiple indicators and the patient's physical conditions. METHODS: All medical data and 69,375 dialysis records of 314 Asian patients undergoing hemodialysis at a single dialysis center in Japan between July 2018 and April 2020 were collected from the electronic medical record system. Using the random forest classifier, we developed models to predict the probabilities of adjusting the dry weight at each dialysis session. RESULTS: The areas under the receiver-operating-characteristic curves of the models for adjusting the dry weight upward and downward were 0.70 and 0.74, respectively. The average probability of upward adjustment of the dry weight had sharp a peak around the actual change over time, while the average probability of downward adjustment of the dry weight formed a gradual peak. Feature importance analysis revealed that median blood pressure decline was a strong predictor for adjusting the dry weight upward. In contrast, elevated serum levels of C-reactive protein and hypoalbuminemia were important indicators for adjusting the dry weight downward. CONCLUSIONS: The random forest classifier should provide a helpful guide to predict the optimal changes to the dry weight with relative accuracy and may be useful in clinical practice.

Identification of factors affecting meropenem pharmacokinetics in critically ill patients: Impact of inflammation on clearance.

INTRODUCTION: The purpose of this study was to explore factors influencing meropenem pharmacokinetics (PKs) in critically ill patients by developing a population PK model and to determine the optimal dosing strategy. METHODS: This prospective observational study involved 12 critically ill patients admitted to the intensive care unit and treated with meropenem 1 g infused over 1 h every 8 h. Blood samples were collected on days 1, 2, and 5 immediately prior to dosing, and at 1, 2, 4, and 6 h after the start of infusion. Population PK parameters were estimated using nonlinear mixed-effects model software. RESULTS: Meropenem PK was adequately described using a two-compartment model. Typical values of total and inter-compartmental clearance were 9.30 L/h and 9.70 L/h, respectively, and the central and peripheral compartment volumes of distribution were 12.61 L and 7.80 L, respectively. C-reactive protein (CRP) was identified as significant covariate affecting total meropenem clearance. The probability of target attainment (PTA) predicted by Monte Carlo simulations varied according to the patients' CRP. The PTA of 100% time above the minimum inhibitory concentration ≤2 mg/L for bacteria was achieved after a dose of 1 and 2 g infused over 4 h every 8 h in patients with CRP of 30 and 5 mg/dL, respectively. CONCLUSION: The findings of this study suggest that CRP might be helpful in managing meropenem dosing in critically ill patients. Higher doses and extended infusion may be required to achieve optimal pharmacodynamic targets.

Interhospital transportation of a COVID-19 patient undergoing veno-venous extracorporeal membrane oxygenation by helicopter.

Some coronavirus disease 2019 (COVID-19) patients develop rapidly progressive acute respiratory distress syndrome and require veno-venous extracorporeal membrane oxygenation (V-V ECMO). A previous study recommended the transfer of ECMO patients to ECMO centers. However, because of the pandemic, a limited number of ECMO centers are available for patient transfer. The safe long-distance interhospital transport of these patients is a concern. To minimize transportation time, helicopter use is a suitable choice. We report the first case of a COVID-19 patient on V-V ECMO, transferred to our ECMO center by helicopter. A 45-year-old man with rheumatoid arthritis history, treated with immunosuppressants, presented with fever and sore throat. He was diagnosed with COVID-19 following a positive severe acute respiratory syndrome coronavirus 2 polymerase chain reaction test result and was subsequently prescribed favipiravir. However, his respiratory failure progressively worsened. On day 10 of hospitalization at the previous hospital, he was intubated, and we received a request for ECMO transport on the next day. The ECMO team, who wore personal protective equipment (N95 respirators, gloves, gowns, and face shields), initiated V-V ECMO in the referring hospital and safely transported the patient by helicopter. The flight time was 7 min. He was admitted to the intensive care unit of our hospital and received tocilizumab. He was discharged on hospital day 31 with no significant sequelae. In this case report, we discuss important factors for the safe and appropriate interhospital transportation of COVID-19 patients on ECMO as well as staff and patient safety during helicopter transportation.

Population Pharmacokinetic Analysis of Meropenem in Critically Ill Patients With Acute Kidney Injury Treated With Continuous Hemodiafiltration.

BACKGROUND: The aim of this study was to conduct a population pharmacokinetic (PK) analysis of meropenem and to explore the optimal dosing strategy for meropenem in critically ill patients with acute kidney injury receiving treatment with continuous hemodiafiltration (CHDF). METHODS: Blood samples were obtained on days 1, 2, and 5 after the start of meropenem administration, immediately before dosing, and at 1, 2, 6, and 8 hours after dosing. Population PK model analysis was performed and concentration-time profiles were simulated using the Nonlinear Mixed Effects Model software. RESULTS: Twenty-one patients receiving CHDF in our intensive care unit were enrolled and 350 serum concentration-time data points were obtained. The PKs of meropenem were best described using a 2-compartment model. Typical total and intercompartmental clearance values were 4.22 L/h and 7.84 L/h, respectively, whereas the central and peripheral compartment volumes of distribution were 14.82 L and 11.75 L, respectively. Estimated glomerular filtration rate was identified as a significant covariate of meropenem total clearance. In simulations of patients with renal failure receiving CHDF, the dose was affected by estimated glomerular filtration rate; a dose of 0.5 g every 8 hours or 1 g every 12 hours showed the probability of target attainment of achieving 100% time above the minimum inhibitory concentration for bacteria with a minimum inhibitory concentration ≤2 mg/L. CONCLUSIONS: A population PK model was developed for meropenem in critically ill patients with acute kidney injury receiving CHDF. Our results indicated that a meropenem dosage of 0.5 g every 8 hours or 1 g every 12 hours was suitable in this population and for susceptible bacteria.

Population Pharmacokinetics of Vancomycin Under Continuous Renal Replacement Therapy Using a Polymethylmethacrylate Hemofilter.

BACKGROUND: Although continuous hemodiafiltration (CHDF) is often performed in critically ill patients during sepsis treatment, the pharmacokinetics of vancomycin (VCM) during CHDF with a polymethylmethacrylate hemofilter (PMMA-CHDF) have not been revealed. In this study, the authors aimed to describe the population pharmacokinetics of VCM in critically ill patients undergoing PMMA-CHDF and clarify its hemofilter clearance (CLhemofilter). METHODS: This single-center, retrospective study enrolled patients who underwent intravenous VCM therapy during PMMA-CHDF at the intensive care unit of Chiba University Hospital between 2008 and 2016. A population analysis was performed, and CLhemofilter was assessed. RESULTS: Twenty-five patients were enrolled. Median body weight (BW) and Sequential Organ Failure Assessment (SOFA) score were 63 kg and 15, respectively. Mean conditions for CHDF were 107.5 ± 18.3 mL/min for blood flow rate and 26.3 ± 6.3 mL/kg/h for effluent flow rate. The mean parameter estimates were distribution volume of the central compartment (V1), 59.1 L; clearance of the central compartment (CL1), 1.35 L/h; distribution volume of the peripheral compartment (V2), 56.1 L; and clearance of the peripheral compartment (CL2), 3.65 L/h. BW and SOFA score were significantly associated with V1 (P < 0.05) and CL1 (P < 0.05), respectively, and were thus selected as covariates in the final model. The estimated dosage of VCM to achieve a target area under the concentration-time curve/minimum inhibitory concentration ≥400 was 27.1 mg/kg for loading and 9.7 mg/kg every 24 hours for maintenance; these dosages were affected by BW and SOFA score. Mean CLhemofilter obtained from 8 patients was 1.35 L/h, which was similar to CL1. CONCLUSIONS: The authors clarified the pharmacokinetics and CLhemofilter of VCM in PMMA-CHDF patients. The PK of VCM in patients undergoing CHDF appeared to vary not only with the CHDF setting and BW but also with SOFA score.

Changes in acute blood purification therapy in critical care: republication of the article published in the Japanese Journal of Artificial Organs.

Acute blood purification therapy is an essential artificial organ in critical care. In the review article, looking back on the history, we describe our present knowledge and techniques of acute blood purification therapy in critical care. The topics include continuous hemodiafiltration (CHDF), online HDF as an artificial liver support, blood purification therapy aiming to remove pathogenic substances of sepsis, a procedure for connecting a CRRT device into an extra-corporeal membrane oxygenation circuit, and replacement fluid for CHDF. We also raise remaining issues and clarify the future direction of acute blood purification therapy in critical care. This review was created based on a translation of the Japanese review written in the Japanese Journal of Artificial Organs in 2017 (Vol. 46, No. 1, pp. 67-70), with adding some references.

Clinical course of a critically ill patient with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).

Although several studies have reported on the clinical and epidemiological characteristics of the patient with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), clinical course of the most severe cases requiring treatment in ICU have been insufficiently reported. A 73-year-old man traveling on a cruise ship with history of hypertension and dyslipidemia developed high fever, dyspnea and cough after 7 days of steroid treatment for sudden sensorineural hearing loss, and tested positive for SARS-CoV-2 in sputa polymerase chain reaction (PCR) examination. His respiratory function deteriorated despite treatments with lopinavir/ritonavir, oseltamivir, azithromycin and meropenem at a regional hospital. He was intubated and transferred to the ICU in the tertiary university hospital on day 10 (ICU day 1). Interferon beta-1b subcutaneous injection was initiated immediately to enhance anti-viral therapy, and favipiravir on ICU day 10 upon availability. Progression of organ dysfunctions necessitated inhalation of nitrogen oxide for respiratory dysfunction, noradrenaline for cardiovascular dysfunction and continuous renal replacement therapy for renal dysfunction. His blood samples PCR also tested positive for SARS-CoV-2, indicating viremia, concomitantly with elevated IL-6 levels. VV-ECMO was initiated after sudden exacerbation of respiratory dysfunction on ICU day 7 to maintain oxygenation. The sustained excessive inflammatory cytokines in the present case might have led to the exacerbation of the disease, requiring vigorous organ support therapies to allow for survival and recovery from the rapid progression of multiple organ dysfunctions and severe respiratory failure.

Interleukin-6 as a diagnostic marker for infection in critically ill patients: A systematic review and meta-analysis.

BACKGROUND: The ability of blood levels of interleukin (IL)-6 to differentiate between infection and non-infection in critically ill patients with suspected infection is unclear. We assessed the diagnostic accuracy of serum IL-6 levels for the diagnosis of infection in critically ill patients. METHODS: We systematically searched the PubMed, MEDLINE, Cochrane Resister of Controlled Trials, Cochrane Database of Systematic Reviews, CINAHL, and Igaku Chuo Zasshi databases for studies published from 1986 to August 2016 that evaluated the accuracy of IL-6 levels for the diagnosis of infection. We constructed 2 × 2 tables and calculated summary estimates of sensitivity and specificity using a bivariate random-effects model. RESULTS: The literature search identified 775 articles, six of which with a total of 527 patients were included according to the predefined criteria. The pooled sensitivity, specificity, and diagnostic odds ratio were 0.73 (95% confidence interval [CI], 0.61-0.82), 0.76 (95% CI, 0.61-0.87), and 2.31 (95% CI, 1.20-3.48), respectively. The area under the curve (AUC) of the summary receiver operator characteristic (SROC) curve was 0.81 (95% CI, 0.78-0.85). In the secondary analysis of two studies with a total of 263 adult critically ill patients with organ dysfunction, the pooled sensitivity, specificity, and diagnostic odds ratio were 0.81 (95% CI, 0.75-0.86), 0.77 (95% CI, 0.67-0.84), and 2.87 (95% CI 2.15-3.60), respectively. CONCLUSIONS: Blood levels of IL-6 have a moderate diagnostic value and a potential clinical utility to differentiate infection in critically ill patients with suspected infection.

An extremely high bioavailability of orally administered vancomycin in a patient with severe colitis and renal insufficiency.

Because there is little absorption of orally administered vancomycin hydrochloride (VCM) through the normal intestinal microvillus membrane, the pharmacokinetics of VCM absorbed from the digestive tract are mostly unknown. Here we report a case of severe colitis and renal insufficiency in which the serum concentration of VCM reached the supratherapeutic range after oral administration. A 54-year-old man receiving outpatient chemotherapy for rectal cancer was admitted to our hospital for severe sepsis and acute renal failure. Multimodal therapy including continuous renal replacement therapy (CRRT) and mechanical ventilation was initiated, and oral VCM administration (0.5 g every 6 h) was begun for suspected severe pseudomembranous colitis with large amounts of watery stool. Despite continued CRRT, the serum VCM concentration increased to 30.6 µg/mL after 4 days. Based on pharmacokinetic analysis, the bioavailability of VCM was estimated to be over 54.5%. Colonoscopy showed that the mucosa was severely damaged throughout the large intestine, resulting in considerable exudation of plasma and blood. This case indicates the need for careful and early monitoring during high-dose oral VCM administration to patients with severe mucosal injury and renal insufficiency.

A safe procedure for connecting a continuous renal replacement therapy device into an extracorporeal membrane oxygenation circuit.

Patients receiving extracorporeal membrane oxygenation (ECMO) often require continuous renal replacement therapy (CRRT). The intra-circuit pressure of adult ECMO usually deviates from the physiological range. We investigated the use of CRRT connected to an ECMO circuit with physiological intra-circuit pressures (0-150 mmHg, defined as the "safety range") using an in vitro experiment involving a water-filled ECMO circuit. The intra-circuit pressure pre-pump, post-pump, and post-oxygenator were measured while varying the height of the pump or ECMO flow. The bypass conduit pressure and distance from the post-oxygenator port were measured to find the "safety point", where the bypass pressure remained within the safety range. Both drainage and return limbs of the CRRT machine were connected to the safety point and the inlet and outlet pressures of the hemofilter were recorded while varying the ECMO and CRRT flow. The pre-pump pressure only remained within the safety range for heights >75 cm (ECMO flow = 4 L/min) or ECMO flow <3.5 L min (height = 50 cm). The post-pump and post-oxygenator pressure was generally outside of the safety range. The bypass pressure decreased according to the distance from the post-oxygenator port and the safety point was found at 60 or 75 cm (in a 90-cm length conduit) regardless of ECMO flow. The hemofilter inlet and outlet pressures remained within the safety range for all conditions of ECMO and CRRT flow, findings validated in clinical cases. The bypass conduit within an ECMO circuit can be connected to a CRRT machine safely under physiological pressures in adult patients receiving ECMO.

Chest CT findings in severe acute respiratory distress syndrome requiring V-V ECMO: J-CARVE registry.

BACKGROUND: Chest computed tomography findings are helpful for understanding the pathophysiology of severe acute respiratory distress syndrome (ARDS). However, there is no large, multicenter, chest computed tomography registry for patients requiring veno-venous extracorporeal membrane oxygenation (V-V ECMO). The aim of this study was to describe chest computed tomography findings at V-V ECMO initiation and to evaluate the association between the findings and outcomes in severe ARDS. METHODS: This multicenter, retrospective cohort study enrolled patients with severe ARDS on V-V ECMO, who were admitted to the intensive care units of 24 hospitals in Japan between January 1, 2012, and December 31, 2022. RESULTS: The primary outcome was 90-day in-hospital mortality. The secondary outcomes were the successful liberation from V-V ECMO and the values of static lung compliance. Among the 697 registry patients, of the 582 patients who underwent chest computed tomography at V-V ECMO initiation, 394 survived and 188 died. Multivariate Cox regression showed that traction bronchiectasis and subcutaneous emphysema increased the risk of 90-day in-hospital mortality (hazard ratio [95% confidence interval] 1.77 [1.19-2.63], p = 0.005 and 1.97 [1.02-3.79], p = 0.044, respectively). The presence of traction bronchiectasis was also associated with decreased successful liberation from V-V ECMO (odds ratio: 0.27 [0.14-0.52], p < 0.001). Lower static lung compliance was associated with some chest computed tomography findings related to changes outside of pulmonary opacity, but not with the findings related to pulmonary opacity. CONCLUSIONS: Traction bronchiectasis and subcutaneous emphysema increased the risk of 90-day in-hospital mortality in patients with severe ARDS who required V-V ECMO.

Median arcuate ligament syndrome presenting as hemorrhagic shock.

The major symptoms of median arcuate ligament syndrome, celiac axis stenosis, or occlusion compressed by the median arcuate ligament include eating-associated abdominal pain and weight loss. Because celiac stenosis increases retrograde collateral blood flow from the superior mesenteric artery to the celiac artery via the pancreaticoduodenal arcade, a pancreaticoduodenal artery aneurysm could occur at a low incidence rate. Rupture of the pancreaticoduodenal artery aneurysm and hemorrhagic shock are rare. In this report, we present 3 cases of patients who had been well with no abdominal symptoms until the day of admission, when they experienced sudden-onset intra-abdominal hemorrhage and shock. These 3 patients were admitted to the emergency department, and contrast-enhanced computed tomography and radiographic selective catheter angiography revealed intra-abdominal hemorrhage, stenosis of the celiac arteries, and dilated pancreaticoduodenal arcade. Case 1 demonstrated severe hemorrhagic shock, whereas case 2 demonstrated moderate shock. We treated ruptured pancreaticoduodenal artery aneurysms with coil embolization. Case 3 demonstrated complete celiac occlusion and moderate hemorrhagic shock, and no aneurysm was detected.

Pharmacokinetics of vancomycin and meropenem during continuous online hemodiafiltration: A case report.

OBJECTIVES: The pharmacokinetics of vancomycin and meropenem in patients treated with continuous online hemodiafiltration (OL-HDF) are not well understood. CASE: We evaluated dialytic clearance and serum concentrations of vancomycin and meropenem by OL-HDF in a critically ill patient with soft tissue infection. The mean clearance of OL-HDF and mean serum concentrations during continuous OL-HDF were 155.2 mL/min and 23.1 µg/mL for vancomycin and 145.6 mL/min and 22.7 µg/mL for meropenem. CONCLUSION: Vancomycin and meropenem showed high clearance rates during continuous OL-HDF. However, continuous infusion of these agents at high doses maintained therapeutic serum concentrations.

Prehospital stroke-scale machine-learning model predicts the need for surgical intervention.

While the development of prehospital diagnosis scales has been reported in various regions, we have also developed a scale to predict stroke type using machine learning. In the present study, we aimed to assess for the first time a scale that predicts the need for surgical intervention across stroke types, including subarachnoid haemorrhage and intracerebral haemorrhage. A multicentre retrospective study was conducted within a secondary medical care area. Twenty-three items, including vitals and neurological symptoms, were analysed in adult patients suspected of having a stroke by paramedics. The primary outcome was a binary classification model for predicting surgical intervention based on eXtreme Gradient Boosting (XGBoost). Of the 1143 patients enrolled, 765 (70%) were used as the training cohort, and 378 (30%) were used as the test cohort. The XGBoost model predicted stroke requiring surgical intervention with high accuracy in the test cohort, with an area under the receiver operating characteristic curve of 0.802 (sensitivity 0.748, specificity 0.853). We found that simple survey items, such as the level of consciousness, vital signs, sudden headache, and speech abnormalities were the most significant variables for accurate prediction. This algorithm can be useful for prehospital stroke management, which is crucial for better patient outcomes.

Urinary neutrophil gelatinase-associated lipocalin and plasma IL-6 in discontinuation of continuous venovenous hemodiafiltration for severe acute kidney injury: a multicenter prospective observational study.

BACKGROUND: Patients with severe acute kidney injury (AKI) who require continuous venovenous hemodiafiltration (CVVHDF) in intensive care unit (ICU) are at high mortality risk. Little is known about clinical biomarkers for risk prediction, optimal initiation, and optimal discontinuation of CVVHDF. METHODS: This prospective observational study was conducted in seven university-affiliated ICUs. For urinary neutrophil gelatinase-associated lipocalin (NGAL) and plasma IL-6 measurements, samples were collected at initiation, 24 h, 48 h after, and CVVHDF discontinuation in adult patients with severe AKI. The outcomes were deaths during CVVHDF and CVVHDF dependence. RESULTS: A total number of 133 patients were included. Twenty-eight patients died without CVVHDF discontinuation (CVVHDF nonsurvivors). Urinary NGAL and plasma IL-6 at the CVVHDF initiation were significantly higher in CVVHDF nonsurvivors than in survivors. Among 105 CVVHDF survivors, 70 patients were free from renal replacement therapy (RRT) or death in the next 7 days after discontinuation (success group), whereas 35 patients died or needed RRT again (failure group). Urinary NGAL at CVVHDF discontinuation was significantly lower in the success group (93.8 ng/ml vs. 999 ng/ml, p < 0.01), whereas no significant difference was observed in plasma IL-6 between the groups. Temporal elevations of urinary NGAL levels during the first 48 h since CVVHDF initiation were observed in CVVHDF nonsurvivors and those who failed in CVVHDF discontinuation. CONCLUSIONS: Urinary NGAL at CVVHDF initiation and discontinuation was associated with mortality and RRT dependence, respectively. The serial changes of urinary NGAL might also help predict the prognosis of patients with AKI on CVVHDF.

High versus low positive end-expiratory pressure setting in patients receiving veno-venous extracorporeal membrane oxygenation support for severe acute respiratory distress syndrome: study protocol for the multicentre, randomised ExPress SAVER Trial.

INTRODUCTION: While limiting the tidal volume to 6 mL/kg during veno-venous extracorporeal membrane oxygenation (V-V ECMO) to ameliorate lung injury in patients with acute respiratory distress syndrome (ARDS) is widely accepted, the best setting for positive end-expiratory pressure (PEEP) is still controversial. This study is being conducted to investigate whether a higher PEEP setting (15 cmH2O) during V-V ECMO can decrease the duration of ECMO support needed in patients with severe ARDS, as compared with a lower PEEP setting. METHODS AND ANALYSIS: The study is an investigator-initiated, multicentre, open-label, two-arm, randomised controlled trial conducted with the participation of 20 intensive care units (ICUs) at academic as well as non-academic hospitals in Japan. The subjects of the study are patients with severe ARDS who require V-V ECMO support. Eligible patients will be randomised equally to the high PEEP group or low PEEP group. Recruitment to the study will continue until a total of 210 patients with ARDS requiring V-V ECMO support have been randomised. In the high PEEP group, PEEP will be set at 15 cmH2O from the start of V-V ECMO until the trials for liberation from V-V ECMO (or until day 28 after the allocation), while in the low PEEP group, the PEEP will be set at 5 cmH2O. Other treatments will be the same in the two groups. The primary endpoint of the study is the number of ECMO-free days until day 28, defined as the length of time (in days) from successful libration from V-V ECMO to day 28. The secondary endpoints are mortality on day 28, in-hospital mortality on day 60, ventilator-free days during the first 60 days and length of ICU stay. ETHICS AND DISSEMINATION: Ethics approval for the trial at all the participating hospitals was obtained on 27 September 2022, by central ethics approval (IRB at Hiroshima University Hospital, C2022-0006). The results of this study will be presented at domestic and international medical congresses, and also published in scientific journals. TRIAL REGISTRATION NUMBER: The Japan Registry of Clinical Trials jRCT1062220062. Registered on 28 September 2022. PROTOCOL VERSION: 28 March 2023, version 4.0.

Successful weight reduction of over 70 kg under hemodialysis and renal function recovery after 87 days of anuria in a patient with dilated cardiomyopathy.

Cardiorenal syndrome is the term that describes the interaction between the heart and kidney that leads to diuretic resistance and worsening renal function. Prolonged anuria generally represents irreversible renal failure, and recovery of renal function after prolonged anuria in patients with heart failure has rarely been reported. Moreover, increased central venous pressure including heart failure is associated with impaired renal function. We herein report a rare case of a 46-year-old man with dilated cardiomyopathy who presented with dyspnea and generalized edema. His body weight increased from 90 kg to 128 kg in one year and he was hospitalized. Central venous pressure (CVP) on admission was 33 mmHg. Intravenous catecholamines were not effective; thus, he was started on continuous hemodiafiltration. Anuria occurred after hemodiafiltration due to heart failure, sepsis, and antibiotics use. However, he experienced weight reduction of over 70 kg, under hemodialysis guided by central venous pressure measurement, and renal function recovery after 87 days of anuria. His CVP had improved to 5 mmHg at discharge. This case showed continuous trial to reduce the CVP and raise cardiac output could result in the recovery of impaired renal function even in the presence of prolonged anuria. .

Prehospital diagnostic algorithm for acute coronary syndrome using machine learning: a prospective observational study.

Rapid and precise prehospital recognition of acute coronary syndrome (ACS) is key to improving clinical outcomes. The aim of this study was to investigate a predictive power for predicting ACS using the machine learning-based prehospital algorithm. We conducted a multicenter observational prospective study that included 10 participating facilities in an urban area of Japan. The data from consecutive adult patients, identified by emergency medical service personnel with suspected ACS, were analyzed. In this study, we used nested cross-validation to evaluate the predictive performance of the model. The primary outcomes were binary classification models for ACS prediction based on the nine machine learning algorithms. The voting classifier model for ACS using 43 features had the highest area under the receiver operating curve (AUC) (0.861 [95% CI 0.775-0.832]) in the test score. After validating the accuracy of the model using the external cohort, we repeated the analysis with a limited number of selected features. The performance of the algorithms using 17 features remained high AUC (voting classifier, 0.864 [95% CI 0.830-0.898], support vector machine (radial basis function), 0.864 [95% CI 0.829-0.887]) in the test score. We found that the machine learning-based prehospital algorithms showed a high predictive power for predicting ACS.