A new risk stratification score for patients with suspected cardiac chest pain in emergency departments, based on machine learning.
Ontology highlight
ABSTRACT:
SUBMITTER: Mao HF
PROVIDER: S-EPMC7147655 | biostudies-literature | 2020 Apr
REPOSITORIES: biostudies-literature
Publications
A new risk stratification score for patients with suspected cardiac chest pain in emergency departments, based on machine learning.
Mao Hai-Feng HF Chen Xiao-Hui XH Li Yun-Mei YM Zhang Si-Yuan SY Mo Jun-Rong JR Li Min M Lin Pei-Yi PY Rainer Timothy H TH Graham Colin A CA Jiang Hui-Lin HL
Chinese medical journal 20200401 7
Similar Datasets
Project description:We investigated patients presenting to emergency departments (EDs) with chest pain to identify factors that influence the use of invasive coronary angiography (ICA). Using linked ED, hospitalisations, death and cardiac biomarker data, we identified people aged 20 years and over who presented with chest pain to tertiary public hospital EDs in Western Australia from 1 January 2016 to 31 March 2017 (ED chest pain cohort). We report patient characteristics, ED discharge diagnosis, pathways to ICA, ICA within 90 days, troponin test results, and gender differences. Associations were examined with the Pearson Chi-squared test and multivariate logistic regression. There were 16,974 people in the ED chest pain cohort, with a mean age of 55.6 years and 50.7% males, accounting for 20,131 ED presentations. Acute coronary syndrome was the ED discharge diagnosis in 10.4% of presentations. ED pathways were: discharged home (57.5%); hospitalisation (41.7%); interhospital transfer (0.4%); and died in ED (0.03%)/inpatients (0.3%). There were 1546 (9.1%) ICAs performed within 90 days of the first ED chest pain visit, of which 59 visits (3.8%) had no troponin tests and 565 visits (36.6%) had normal troponin. ICAs were performed in more men than women (12.3% vs. 6.1%, p < 0.0001; adjusted OR 1.89, 95% CI 1.65, 2.18), and mostly within 7 days. Equal numbers of males and females present with chest pain to tertiary hospital EDs, but men are twice as likely to get ICA. Over one-third of ICAs occur in those with normal troponin levels, indicating that further investigation is required to determine risk profile, outcomes and cost effectiveness.
| S-EPMC7766965 | biostudies-literature
Project description:The emergency department assessment of chest pain score-accelerated diagnostic protocol (EDACS-ADP) are commonly used for risk stratification in undifferentiated patients with acute chest pain. This systematic review aimed to investigate EDACS-ADP for risk stratification of emergency department (ED) patients with chest pain. The PubMed, Web of Science, Medline, Cochrane Library, China National Knowledge Infrastructure, and Wanfang databases were searched for related studies without restrictions on the publication year. The Quality Assessment of Diagnostic Accuracy Studies 2 tool was used to assess the risk of bias, and Stata 16.0 was used to determine the combined sensitivity, specificity, positive diagnostic likelihood ratio (DLR), and negative DLR. Twelve studies comprising 14 290 patients were identified. Of these, 7537 (52.74%) patients were considered low risk, and 67 (0.89%) had major adverse cardiovascular events (MACE), including myocardial infarction, stroke, and cardiovascular death within 30 days of the patients' ED presentation. EDACS-ADP showed a combined sensitivity of 0.97 (95% confidence interval [CI]: 0.95-0.99); specificity, 0.58 (0.53-0.63); positive DLR, 2.34 (2.08-2.63); negative DLR, 0.04 (0.02-0.09); diagnostic odds ratio, 53.11 (26.45-106.63); and summary receiver operating characteristic area under the curve, 0.83 (0.79-0.86). Despite the large statistical heterogeneity of the results, EDACS-ADP identified a considerable number of low-risk patients for early discharge, with a specificity >50% and an incidence of MACE within 30-days of patients' ED presentation <1%. Thus, it is a useful tool with a potential for clinical application.
| S-EPMC10642332 | biostudies-literature
Project description:Diagnosing patients in the medical emergency department is complex and this is expected to increase in many countries due to an ageing population. In this study we investigate the feasibility of training machine learning algorithms to assist physicians handling the complex situation in the medical emergency departments. This is expected to reduce diagnostic errors and improve patient logistics and outcome. We included a total of 9,190 consecutive patient admissions diagnosed and treated in two hospitals in this cohort study. Patients had a biochemical workup including blood and urine analyses on clinical decision totaling 260 analyses. After adding nurse-registered data we trained 19 machine learning algorithms on a random 80% sample of the patients and validated the results on the remaining 20%. We trained algorithms for 19 different patient outcomes including the main outcomes death in 7 (Area under the Curve (AUC) 91.4%) and 30 days (AUC 91.3%) and safe-discharge(AUC 87.3%). The various algorithms obtained areas under the Receiver Operating Characteristics -curves in the range of 71.8-96.3% in the holdout cohort (68.3-98.2% in the training cohort). Performing this list of biochemical analyses at admission also reduced the number of subsequent venipunctures within 24 h from patient admittance by 22%. We have shown that it is possible to develop a list of machine-learning algorithms with high AUC for use in medical emergency departments. Moreover, the study showed that it is possible to reduce the number of venipunctures in this cohort.
| S-EPMC11297196 | biostudies-literature
Project description:BackgroundStroke misdiagnosis, associated with poor outcomes, is estimated to occur in 9% of all stroke patients.ObjectivesWe hypothesized that machine learning (ML) could assist in the diagnosis of ischemic stroke in emergency departments (EDs).DesignThe study was conducted and reported according to the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis guidelines. We performed model development and prospective temporal validation, using data from pre- and post-COVID periods; we also performed a case study on a small cohort of previously misdiagnosed stroke patients.MethodsWe used structured and unstructured electronic health records (EHRs) of 56,452 patient encounters from 13 hospitals in Pennsylvania, from September 2003 to January 2021. ML pipelines, including natural language processing, were created using pre-event clinical data and provider notes in the EDs.ResultsUsing pre-event information, our model's area under the receiver operating characteristics curve (AUROC) ranged from 0.88 to 0.92 with a similar range accuracy (0.87-0.90). Using provider notes, we identified five models that reached a balanced performance in terms of AUROC, sensitivity, and specificity. Model AUROC ranged from 0.93 to 0.99. Model sensitivity and specificity reached 0.90 and 0.99, respectively. Four of the top five performing models were based on the post-COVID provider notes; however, no performance difference between models tested on pre- and post-COVID was observed.ConclusionThis study leveraged pre-event and at-encounter level EHR for stroke prediction. The results indicate that available clinical information can be used for building EHR-based stroke prediction models and ED stroke alert systems.
| S-EPMC10989051 | biostudies-literature
Project description:BackgroundChest pain is amongst the most common reason for presentation to the emergency department (ED). There are many causes of chest pain, and it is important for the emergency physician to quickly and accurately diagnose life threatening causes such as acute myocardial infarction (AMI). Multiple clinical decision tools have been developed to assist clinicians in risk stratifying patients with chest. There is growing recognition that machine learning (ML) will have a significant impact on the practice of medicine in the near future and may assist with diagnosis and risk stratification. This systematic review aims to evaluate how ML has been applied to adults presenting to the ED with undifferentiated chest pain and assess if ML models show improved performance when compared to physicians or current risk stratification techniques.Methods and findingsWe conducted a systematic review of journal articles that applied a ML technique to an adult patient presenting to an emergency department with undifferentiated chest pain. Multiple databases were searched from inception through to November 2020. In total, 3361 articles were screened, and 23 articles were included. We did not conduct a metanalysis due to a high level of heterogeneity between studies in both their methods, and reporting. The most common primary outcomes assessed were diagnosis of acute myocardial infarction (AMI) (12 studies), and prognosis of major adverse cardiovascular event (MACE) (6 studies). There were 14 retrospective studies and 5 prospective studies. Four studies reported the development of a machine learning model retrospectively then tested it prospectively. The most common machine learning methods used were artificial neural networks (14 studies), random forest (6 studies), support vector machine (5 studies), and gradient boosting (2 studies). Multiple studies achieved high accuracy in both the diagnosis of AMI in the ED setting, and in predicting mortality and composite outcomes over various timeframes. ML outperformed existing risk stratification scores in all cases, and physicians in three out of four cases. The majority of studies were single centre, retrospective, and without prospective or external validation. There were only 3 studies that were considered low risk of bias and had low applicability concerns. Two studies reported integrating the ML model into clinical practice.ConclusionsResearch on applications of ML for undifferentiated chest pain in the ED has been ongoing for decades. ML has been reported to outperform emergency physicians and current risk stratification tools to diagnose AMI and predict MACE but has rarely been integrated into practice. Many studies assessing the use of ML in undifferentiated chest pain in the ED have a high risk of bias. It is important that future studies make use of recently developed standardised ML reporting guidelines, register their protocols, and share their datasets and code. Future work is required to assess the impact of ML model implementation on clinical decision making, patient orientated outcomes, and patient and physician acceptability.Trial registrationInternational Prospective Register of Systematic Reviews registration number: CRD42020184977.
| S-EPMC8384172 | biostudies-literature
Project description:Early risk stratification of out-of-hospital cardiac arrest (OHCA) patients with insufficient information in emergency departments (ED) is difficult but critical in improving intensive care resource allocation. This study aimed to develop a simple risk stratification score using initial information in the ED. Adult patients who had OHCA with medical etiology from 2016 to 2020 were enrolled from the Korean Cardiac Arrest Research Consortium (KoCARC) database. To develop a scoring system, a backward logistic regression analysis was conducted. The developed scoring system was validated in both external dataset and internal bootstrap resampling. A total of 8240 patients were analyzed, including 4712 in the development cohort and 3528 in the external validation cohort. An ED-PLANN score (range 0-5) was developed incorporating 1 point for each: P for serum pH ≤ 7.1, L for serum lactate ≥ 10 mmol/L, A for age ≥ 70 years old, N for non-shockable rhythm, and N for no-prehospital return of spontaneous circulation. The area under the receiver operating characteristics curve (AUROC) for favorable neurological outcome was 0.93 (95% CI, 0.92-0.94) in the development cohort, 0.94 (95% CI, 0.92-0.95) in the validation cohort. Hosmer-Lemeshow goodness-of-fit tests also indicated good agreement. The ED-PLANN score is a practical and easily applicable clinical scoring system for predicting favorable neurological outcomes of OHCA patients.
| S-EPMC8745643 | biostudies-literature
Project description: Not available
| S-EPMC8596494 | biostudies-literature
Project description:BACKGROUND:Hong Kong (HK) and Guangzhou (GZ) are cities in China with different healthcare systems. This study aimed to compare 30-day and 6-month mortality and characteristics of patients with suspected cardiac chest pain admitted to two emergency departments (ED) in HK and GZ. METHODS:A prospective observational study enrolled patients with suspected cardiac chest pain presenting to EDs in the Prince of Wales Hospital (PWH), HK and the Second Affiliated Hospital of Guangzhou Medical University (AHGZMU),GZ. The primary outcome was 30-day and 6-month mortality. RESULTS:In total, 996 patients were recruited, 407 cases from GZ and 589 cases from HK.The 30-day and 6-month mortality of chest patients were 3.7% and 4.7% in GZand 0.3% and 1.9% in HK, respectively. Serum creatinine level (Cr) was an independent factor for 30-day mortality whilst Cr and systolic blood pressure (SBP) were independent factors for 6-month mortality. In Cox regression analysis, unadjusted and adjusted hazard ratios for 30-day and 6-month mortality in GZ were significantly increased. CONCLUSION:The 30-day and 6-month mortality of patients with suspected cardiac chest pain in Guangzhou were higher than in Hong Kong due to due to different baseline clinical characteristics of patients and different distributions of diagnoses, which were associated with different healthcare systems. Serum creatinine and SBP were independent factors for 30-day and 6-month mortality.
| S-EPMC5956813 | biostudies-literature
Project description:For patients with chest pain who are deemed clinically to be low risk and discharged home from the emergency department (ED), it is unclear whether further laboratory tests can improve risk stratification. Here, we investigated the utility of a clinical chemistry score (CCS), which comprises plasma glucose, the estimated glomerular filtration rate, and high-sensitivity cardiac troponin (I or T) to generate a common score for risk stratification. In a cohort of 14,676 chest pain patients in the province of Ontario, Canada and who were discharged home from the ED (November 2012-February 2013 and April 2013-September 2015) we evaluated the CCS as a risk stratification tool for all-cause mortality, plus hospitalization for myocardial infarction or unstable angina (primary outcome) at 30, 90, and 365 days post-discharge using Cox proportional hazard models. At 30 days the primary outcome occurred in 0.3% of patients with a CCS < 2 (n = 6404), 0.9% of patients with a CCS = 2 (n = 4336), and 2.3% of patients with a CCS > 2 (n = 3936) (p < 0.001). At 90 days, patients with CCS < 2 (median age = 52y (IQR = 46-60), 59.4% female) had an adjusted HR = 0.51 (95% confidence interval (CI) = 0.32-0.82) for the composite outcome and patients with a CCS > 2 (median age = 74y (IQR = 64-82), 48.0% female) had an adjusted HR = 2.80 (95%CI = 1.98-3.97). At 365 days, 1.3%, 3.4%, and 11.1% of patients with a CCS < 2, 2, or >2 respectively, had the composite outcome (p < 0.001). In conclusion, the CCS can risk stratify chest pain patients discharged home from the ED and identifies both low- and high-risk patients who may warrant different medical care.
| S-EPMC7565964 | biostudies-literature
Project description:The sensitivity of triage systems in identifying acute cardiovascular events in patients presented to the emergency department with chest pain is not optimal. Recently, a clinical score, the Emergency Department Assessment of Chest Pain Score (EDACS), has been proposed for a rapid assessment without additional instruments. To evaluate whether the integration of EDACS into triage evaluation of patients with chest pain can improve the triage's predictive validity for an acute cardiovascular event, a single-center prospective observational study was conducted. This study involved all patients who needed a triage admission for chest pain between January 1, 2020, and December 31, 2020. All enrolled patients first underwent a standard triage assessment and then the EDACS was calculated. The primary outcome of the study was the presence of an acute cardiovascular event. The discriminatory ability of EDACS in triage compared with standard triage assessment was evaluated by comparing the areas under the receiver operating characteristic curve, decision curve analysis, and net reclassification improvement. The study involved 1,596 patients, of that 7.3% presented the study outcome. The discriminatory ability of triage presented an area under the receiver operating characteristic curve of 0.688 that increased to 0.818 after the application of EDACS in the triage assessment. EDACS improved the baseline assessment of priority assigned in triage, with a net reclassification improvement of 33.6% (p <0.001), and the decision curve analyses demonstrated that EDACS in triage resulted in a clear net clinical benefit. In conclusion, the results of the study suggest that EDACS has a good discriminatory capacity for acute cardiovascular events and that its implementation in routine triage may improve triage performance in patients with chest pain.
| S-EPMC9336201 | biostudies-literature