Accuracy of Referral and Phone-Triage Diagnoses in an Eye Emergency Department.
Ontology highlight
ABSTRACT:
SUBMITTER: Deaner JD
PROVIDER: S-EPMC7384427 | biostudies-literature | 2020 Jul
REPOSITORIES: biostudies-literature
Publications
Accuracy of Referral and Phone-Triage Diagnoses in an Eye Emergency Department.
Deaner Jordan D JD Amarasekera Dilru C DC Ozzello Daniel J DJ Swaminathan Vishal V Bonafede Lucas L Meeker Austin R AR Zhang Qiang Q Haller Julia A JA
Ophthalmology 20200727 3
Similar Datasets
Project description:ObjectiveTo predict hospital admission at the time of ED triage using patient history in addition to information collected at triage.MethodsThis retrospective study included all adult ED visits between March 2014 and July 2017 from one academic and two community emergency rooms that resulted in either admission or discharge. A total of 972 variables were extracted per patient visit. Samples were randomly partitioned into training (80%), validation (10%), and test (10%) sets. We trained a series of nine binary classifiers using logistic regression (LR), gradient boosting (XGBoost), and deep neural networks (DNN) on three dataset types: one using only triage information, one using only patient history, and one using the full set of variables. Next, we tested the potential benefit of additional training samples by training models on increasing fractions of our data. Lastly, variables of importance were identified using information gain as a metric to create a low-dimensional model.ResultsA total of 560,486 patient visits were included in the study, with an overall admission risk of 29.7%. Models trained on triage information yielded a test AUC of 0.87 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.88) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on patient history yielded an AUC of 0.86 for LR (95% CI 0.86-0.87), 0.87 for XGBoost (95% CI 0.87-0.87) and 0.87 for DNN (95% CI 0.87-0.88). Models trained on the full set of variables yielded an AUC of 0.91 for LR (95% CI 0.91-0.91), 0.92 for XGBoost (95% CI 0.92-0.93) and 0.92 for DNN (95% CI 0.92-0.92). All algorithms reached maximum performance at 50% of the training set or less. A low-dimensional XGBoost model built on ESI level, outpatient medication counts, demographics, and hospital usage statistics yielded an AUC of 0.91 (95% CI 0.91-0.91).ConclusionMachine learning can robustly predict hospital admission using triage information and patient history. The addition of historical information improves predictive performance significantly compared to using triage information alone, highlighting the need to incorporate these variables into prediction models.
| S-EPMC6054406 | biostudies-literature
Project description:BackgroundSyndromic surveillance in emergency departments (EDs) may be used to deliver early warnings of increases in disease activity, to provide situational awareness during events of public health significance, to supplement other information on trends in acute disease and injury, and to support the development and monitoring of prevention or response strategies. Changes in mental health related ED presentations may be relevant to these goals, provided they can be identified accurately and efficiently. This study aimed to measure the accuracy of using diagnostic codes in electronic ED presentation records to identify mental health-related visits.MethodsWe selected a random sample of 500 records from a total of 1,815,588 ED electronic presentation records from 59 NSW public hospitals during 2010. ED diagnoses were recorded using any of ICD-9, ICD-10 or SNOMED CT classifications. Three clinicians, blinded to the automatically generated syndromic grouping and each other's classification, reviewed the triage notes and classified each of the 500 visits as mental health-related or not. A "mental health problem presentation" for the purposes of this study was defined as any ED presentation where either a mental disorder or a mental health problem was the reason for the ED visit. The combined clinicians' assessment of the records was used as reference standard to measure the sensitivity, specificity, and positive and negative predictive values of the automatic classification of coded emergency department diagnoses. Agreement between the reference standard and the automated coded classification was estimated using the Kappa statistic.ResultsAgreement between clinician's classification and automated coded classification was substantial (Kappa = 0.73. 95% CI: 0.58 - 0.87). The automatic syndromic grouping of coded ED diagnoses for mental health-related visits was found to be moderately sensitive (68% 95% CI: 46%-84%) and highly specific at 99% (95% CI: 98%-99.7%) when compared with the reference standard in identifying mental health related ED visits. Positive predictive value was 81% (95% CI: 0.57 - 0.94) and negative predictive value was 98% (95% CI: 0.97-0.99).ConclusionsMental health presentations identified using diagnoses coded with various classifications in electronic ED presentation records offers sufficient accuracy for application in near real-time syndromic surveillance.
| S-EPMC4177714 | biostudies-literature
Project description:ObjectiveIn the emergency department, hyperglycemic patients are screened for diabetic ketoacidosis (DKA) via a urine dipstick. In this prospective study, we compared the test characteristics of point-of-care ?-hydroxybutyrate (?-OHB) analysis with the urine dipstick.Research design and methodsEmergency-department patients with blood glucose ?250 mg/dL had urine dipstick, chemistry panel, venous blood gas, and capillary ?-OHB measurements. DKA was diagnosed according to American Diabetes Association criteria.ResultsOf 516 hyperglycemic subjects, 54 had DKA. The urine dipstick had a sensitivity of 98.1% (95% CI 90.1-100), a specificity of 35.1% (30.7-39.6), a positive predictive value of 15% (11.5-19.2), and a negative predictive value of 99.4% (96.6-100) for DKA. Using the manufacturer-suggested cutoff of >1.5 mmol/L, ?-OHB had a sensitivity of 98.1% (90.1-100), a specificity of 78.6% (74.5-82.2), a positive predictive value of 34.9% (27.3-43), and a negative predictive value of 99.7% (98.5-100) for DKA.ConclusionsPoint-of-care ?-OHB and the urine dipstick are equally sensitive for detecting DKA (98.1%). However, ?-OHB is more specific (78.6 vs. 35.1%), offering the potential to significantly reduce unnecessary DKA work-ups among hyperglycemic patients in the emergency department.
| S-EPMC3064039 | biostudies-literature
Project description: Not available
| S-EPMC7436489 | biostudies-literature
Project description:BackgroundDevelopment of emergency department (ED) triage systems that accurately differentiate and prioritize critically ill from stable patients remains challenging. We used machine learning models to predict clinical outcomes, and then compared their performance with that of a conventional approach-the Emergency Severity Index (ESI).MethodsUsing National Hospital and Ambulatory Medical Care Survey (NHAMCS) ED data, from 2007 through 2015, we identified all adult patients (aged ??18?years). In the randomly sampled training set (70%), using routinely available triage data as predictors (e.g., demographics, triage vital signs, chief complaints, comorbidities), we developed four machine learning models: Lasso regression, random forest, gradient boosted decision tree, and deep neural network. As the reference model, we constructed a logistic regression model using the five-level ESI data. The clinical outcomes were critical care (admission to intensive care unit or in-hospital death) and hospitalization (direct hospital admission or transfer). In the test set (the remaining 30%), we measured the predictive performance, including area under the receiver-operating-characteristics curve (AUC) and net benefit (decision curves) for each model.ResultsOf 135,470 eligible ED visits, 2.1% had critical care outcome and 16.2% had hospitalization outcome. In the critical care outcome prediction, all four machine learning models outperformed the reference model (e.g., AUC, 0.86 [95%CI 0.85-0.87] in the deep neural network vs 0.74 [95%CI 0.72-0.75] in the reference model), with less under-triaged patients in ESI triage levels 3 to 5 (urgent to non-urgent). Likewise, in the hospitalization outcome prediction, all machine learning models outperformed the reference model (e.g., AUC, 0.82 [95%CI 0.82-0.83] in the deep neural network vs 0.69 [95%CI 0.68-0.69] in the reference model) with less over-triages in ESI triage levels 1 to 3 (immediate to urgent). In the decision curve analysis, all machine learning models consistently achieved a greater net benefit-a larger number of appropriate triages considering a trade-off with over-triages-across the range of clinical thresholds.ConclusionsCompared to the conventional approach, the machine learning models demonstrated a superior performance to predict critical care and hospitalization outcomes. The application of modern machine learning models may enhance clinicians' triage decision making, thereby achieving better clinical care and optimal resource utilization.
| S-EPMC6387562 | biostudies-literature
Project description:BackgroundEmergency department (ED) crowding is associated with negative health outcomes, patient dissatisfaction, and longer length of stay (LOS). The addition of advanced diagnostic imaging (ADI), namely CT, ultrasound (U/S), and MRI to ED encounter work up is a predictor of longer length of stay. Earlier and improved prediction of patients' need for advanced imaging may improve overall ED efficiency. The aim of the study was to detect the association between ADI utilization and the structured and unstructured information immediately available during ED triage, and to develop and validate models to predict utilization of ADI during an ED encounter.MethodsWe used the United States National Hospital Ambulatory Medical Care Survey data from 2009 to 2014 to examine which sociodemographic and clinical factors immediately available at ED triage were associated with the utilization of CT, U/S, MRI, and multiple ADI during a patient's ED stay. We used natural language processing (NLP) topic modeling to incorporate free-text reason for visit data available at time of ED triage in addition to other structured patient data to predict the use of ADI using multivariable logistic regression models.ResultsAmong the 139,150 adult ED visits from a national probability sample of hospitals across the U.S, 21.9% resulted in ADI use, including 16.8% who had a CT, 3.6% who had an ultrasound, 0.4% who had an MRI, and 1.2% of the population who had multiple types of ADI. The c-statistic of the predictive models was greater than or equal to 0.78 for all imaging outcomes, and the addition of text-based reason for visit information improved the accuracy of all predictive models.ConclusionsPatient information immediately available during ED triage can accurately predict the eventual use of advanced diagnostic imaging during an ED visit. Such models have the potential to be incorporated into the ED triage workflow in order to more rapidly identify patients who may require advanced imaging during their ED stay and assist with medical decision-making.
| S-EPMC6456195 | biostudies-literature
Project description:IntroductionMany patients who are discharged from the emergency department (ED) with a symptom-based discharge diagnosis (SBD) have post-discharge challenges related to lack of a definitive discharge diagnosis and follow-up plan. There is no well-defined method for identifying patients with a SBD without individual chart review. We describe a method for automated identification of SBDs from ICD-10 codes using the Unified Medical Language System (UMLS) Metathesaurus.MethodsWe mapped discharge diagnosis, with use of ICD-10 codes from a one-month period of ED discharges at an urban, academic ED to UMLS concepts and semantic types. Two physician reviewers independently manually identified all discharge diagnoses consistent with SBDs. We calculated inter-rater reliability for manual review and the sensitivity and specificity for our automated process for identifying SBDs against this "gold standard."ResultsWe identified 3642 ED discharges with 1382 unique discharge diagnoses that corresponded to 875 unique ICD-10 codes and 10 UMLS semantic types. Over one third (37.5%, n = 1367) of ED discharges were assigned codes that mapped to the "Sign or Symptom" semantic type. Inter-rater reliability for manual review of SBDs was very good (0.87). Sensitivity and specificity of our automated process for identifying encounters with SBDs were 84.7% and 96.3%, respectively.ConclusionUse of our automated process to identify ICD-10 codes that classify into the UMLS "Sign or Symptom" semantic type identified the majority of patients with a SBD. While this method needs refinement to increase sensitivity of capture, it has potential to automate an otherwise highly time-consuming process. This novel use of informatics methods can facilitate future research specific to patients with SBDs.
| S-EPMC6860381 | biostudies-literature
Project description:ObjectiveTo investigate compliance with telephone helpline advice to attend an emergency department (ED) and the acuity of patients who presented to ED following a call.Data sources/collection methodsIn New South Wales (NSW), Australia, 2009-2012, all (1.04 million) calls to a telephone triage service, ED presentations, hospital admissions and death registrations, linked using probabilistic data linkage.Study designPopulation-based, observational cohort study measuring ED presentations within 24 hours of a call in patients (1) with dispositions to attend ED (compliance) and (2) low-urgency dispositions (self-referral), triage categories on ED presentation.Principal findingsA total of 66.5 percent of patients were compliant with dispositions to attend an ED. A total of 6.2 percent of patients with low-urgency dispositions self-referred to the ED within 24 hours. After age adjustment, healthdirect compliant patients were significantly less likely (7.8 percent) to receive the least urgent ED triage category compared to the general NSW ED population (16.9 percent).ConclusionsThis large population-based data linkage study provides precise estimates of ED attendance following calls to a telephone triage service and details the predictors of ED attendance. Patients who attend an ED compliant with a healthdirect helpline disposition are significantly less likely than the general ED population to receive the lowest urgency triage category on arrival.
| S-EPMC5867179 | biostudies-literature
Project description:ObjectiveTriage is a tool developed to identify patients who need immediate care and those who can safely wait. The aim of this study was to assess the validity and interrater reliability of a modified version of the pediatric South African triage scale (pSATS) in a single-center tertiary pediatric emergency department in Norway.MethodsThis prospective, observational study included all patients with medical conditions, referred to the pediatric emergency department of a tertiary hospital in Norway from September 1, 2015, to November 17, 2015. Their assigned triage priority was compared with rate of hospitalization and resource utilization. Validity parameters were sensitivity, specificity, positive and negative predictive value, and percentage of over- and undertriage. Interrater agreement and accuracy of the triage ratings were calculated from triage performed by nurses on written case scenarios.ResultsDuring the study period, 1171 patients arrived at the hospital for emergency assessment. A total of 790 patients (67 %) were triaged and included in the study. The percentage of hospital admission increased with increasing level of urgency, from 30 % of the patients triaged to priority green to 81 % of those triaged to priority red. The sensitivity was 74 %, the specificity was 48 %, the positive predictive value was 52 %, and the negative predictive value was 70 % for predicting hospitalization. The level of over- and undertriage was 52 % and 26 %, respectively. Resource utilization correlated with higher triage priority. The interrater agreement had an intraclass correlation coefficient of 0.99 by Cronbach's alpha, and the accuracy was 92 %.ConclusionsThe modified pSATS had a moderate sensitivity and specificity but showed good correlation with resource utilization. The nurses demonstrated excellent interrater agreement and accuracy when triaging written case scenarios.
| S-EPMC6205310 | biostudies-literature
Project description:Objectives:The aim of this study was to develop machine learning (ML) and initial nursing assessment (INA)-based emergency department (ED) triage to predict adverse clinical outcome. Methods:The retrospective study included ED visits between January 2016 and December 2017 that resulted in either intensive care unit admission or emergency room death. We trained four classifiers using logistic regression and a deep learning model on INA and low dimensional (LD) INA, logistic regression on the Korea Triage and acuity scale (KTAS) and Sequential Related Organ Failure Assessment (SOFA). We varied the outcome ratio for external validation. Finally, variables of importance were identified using the random forest model's information gain. The four most influential variables were used for LD modeling for efficiency. Results:A total of 86,304 patient visits were included, with an overall outcome rate of 3.5%. The area under the curve (AUC) values for the KTAS model were 76.8 (74.9-78.6) with logistic regression and 74.0 (72.1-75.9) for the SOFA model, while the AUC values of the INA model were 87.2 (85.9-88.6) and 87.6 (86.3-88.9) with logistic regression and deep learning, suggesting that the ML and INA-based triage system result more accurately predicted the outcomes. The AUC values for the LD model were 81.2 (79.4-82.9) and 80.7 (78.9-82.5) for logistic regression and deep learning, respectively. Conclusions:We developed an ML and INA-based triage system for EDs. The novel system was able to predict clinical outcomes more accurately than existing triage systems, KTAS and SOFA.
| S-EPMC7010940 | biostudies-literature