ARAndrew T. Reisner

Andrew T. Reisner

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Andrew T. Reisner has a total of 43 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Atypical symptoms in emergency department patients with urosepsis challenge current urinary tract infection management guidelines

    by Brett Biebelberg, Iain E. Kehoe, Hui Zheng, Abigail O'COnnell, Michael R. Filbin, Thomas Heldt, Andrew T. Reisner


  • Open Access

    Diagnostic suspicion bias and machine learning: Breaking the awareness deadlock for sepsis detection

    by Varesh Prasad, Baturay Aydemir, Iain E. Kehoe, Chaya Kotturesh, Abigail O’cOnnell, Brett Biebelberg, Yang Wang, James C. Lynch, Jeremy A. Pepino, Michael R. Filbin, et al.

    Many early warning algorithms are downstream of clinical evaluation and diagnostic testing, which means that they may not be useful when clinicians fail to suspect illness and fail to order appropriate tests. Depending on how such algorithms handle missing data, they could even indicate “low risk” simply because the testing data were never ordered. We considered predictive methodologies to identify sepsis at triage, before diagnostic tests are ordered, in a busy Emergency Department (ED). One algorithm used “bland clinical data” (data available at triage for nearly every patient). The second algorithm added three yes/no questions to be answered after the triage interview. Retrospectively, we studied adult patients from a single ED between 2014–16, separated into training (70%) and testing (30%) cohorts, and a final validation cohort of patients from four EDs between 2016–2018. Sepsis was defined per the Rhee criteria. Investigational predictors were demographics and triage vital signs (downloaded from the hospital EMR); past medical history; and the auxiliary queries (answered by chart reviewers who were blinded to all data except the triage note and initial HPI). We developed L2-regularized logistic regression models using a greedy forward feature selection. There were 1164, 499, and 784 patients in the training, testing, and validation cohorts, respectively. The bland clinical data model yielded ROC AUC’s 0.78 (0.76–0.81) and 0.77 (0.73–0.81), for training and testing, respectively, and ranged from 0.74–0.79 in four hospital validation. The second model which included auxiliary queries yielded 0.84 (0.82–0.87) and 0.83 (0.79–0.86), and ranged from 0.78–0.83 in four hospital validation. The first algorithm did not require clinician input but yielded middling performance. The second showed a trend towards superior performance, though required additional user effort. These methods are alternatives to predictive algorithms downstream of clinical evaluation and diagnostic testing. For hospital early warning algorithms, consideration should be given to bias and usability of various methods.


  • Open Access

    APPRAISE-HRI: AN ARTIFICIAL INTELLIGENCE ALGORITHM FOR TRIAGE OF HEMORRHAGE CASUALTIES

    by Jonathan D. Stallings, Srinivas Laxminarayan, Chenggang Yu, Adam Kapela, Andrew Frock, Andrew P. Cap, Andrew T. Reisner, Jaques Reifman

    Background: Hemorrhage remains the leading cause of death on the battlefield. This study aims to assess the ability of an artificial intelligence triage algorithm to automatically analyze vital-sign data and stratify hemorrhage risk in trauma patients. Methods: Here, we developed the APPRAISE–Hemorrhage Risk Index (HRI) algorithm, which uses three routinely measured vital signs (heart rate and diastolic and systolic blood pressures) to identify trauma patients at greatest risk of hemorrhage. The algorithm preprocesses the vital signs to discard unreliable data, analyzes reliable data using an artificial intelligence–based linear regression model, and stratifies hemorrhage risk into low (HRI:I), average (HRI:II), and high (HRI:III). Results: To train and test the algorithm, we used 540 h of continuous vital-sign data collected from 1,659 trauma patients in prehospital and hospital (i.e., emergency department) settings. We defined hemorrhage cases (n = 198) as those patients who received ≥1 unit of packed red blood cells within 24 h of hospital admission and had documented hemorrhagic injuries. The APPRAISE-HRI stratification yielded a hemorrhage likelihood ratio (95% confidence interval) of 0.28 (0.13–0.43) for HRI:I, 1.00 (0.85–1.15) for HRI:II, and 5.75 (3.57–7.93) for HRI:III, suggesting that patients categorized in the low-risk (high-risk) category were at least 3-fold less (more) likely to have hemorrhage than those in the average trauma population. We obtained similar results in a cross-validation analysis. Conclusions: The APPRAISE-HRI algorithm provides a new capability to evaluate routine vital signs and alert medics to specific casualties who have the highest risk of hemorrhage, to optimize decision-making for triage, treatment, and evacuation.


  • Open Access

    Hypotension During Vasopressor Infusion Occurs in Predictable Clusters: A Multicenter Analysis

    by Daisuke Horiguchi, Sungtae Shin, Jeremy A. Pepino, Jeffrey T. Peterson, Iain E. Kehoe, Joshua N. Goldstein, Mary E. McAuley, Jarone Lee, Laura R. Lux, Brian K. Kwon, et al.

    Rationale Published evidence indicates that mean arterial pressure (MAP) below a goal range (hypotension) is associated with worse outcomes, though MAP management failures are common. Objectives Characterize hypotension occurrences in ICUs and consider the implications for MAP management. Methods Observational study of three retrospective and one prospective cohorts of adult ICU patients during continuous vasopressor infusion. Three cohorts of general, mixed ICU patients; one cohort of acute spinal cord injury patients. Hypotension clusters were defined wherever ≥50% of MAP measurements were hypotensive with ≥10 minutes of total hypotension. Trend analysis was performed (predicting future MAP using 14 min of preceding MAP data) to understand which hypotension clusters could likely have been predicted by clinician awareness of MAP trends. Measurements and Main Results In cohorts of 155, 66, 16, and 10 subjects, respectively, hypotension occurred 12.9%, 5.5%, 21.5%, and 3.9% of the time. Hypotension occurred overwhelmingly within temporal clusters. MAP trend analysis predicted most substantial hypotension clusters (range: 66.6% to 80.0% sensitivity) prior to continuous hypotension occurring, usually before the occurrence of any hypotension at all. When MAP trend analysis predicted hypotension, positive predictive value ranged from 56% to 77% (probability of subsequent temporal cluster of hypotension or vasopressor up-titration). Conclusions Hypotension usually occurred in temporal clusters that can typically be predicted by extrapolation from trends. This implies that clinicians may be slow to identify incipient hypotension. In Cohort 4, analysis was performed in real-time, establishing a proof-of-principle for a computational tool with possible clinical utility. Statement of Clinical Significance This analysis found that hypotension was relatively common in the bottom quartiles in each of four ICU cohorts, occurring in temporal clusters that could usually be predicted by trend extrapolation. This implies that clinicians may be slow to identify incipient hypotension in ICU patients. In one cohort, hypotension prediction was performed in real-time, establishing a proof-of-principle for future clinical usability.

  • Open Access

    Modeling of Usual Care: Vasopressor Initiation for Sepsis With Hypotension

    by Varesh Prasad, Andrew T. Reisner, James C. Lynch, Michael R. Filbin, Thomas Heldt

    Usual care regarding vasopressor initiation is ill-defined. We aimed to develop a quantitative “dynamic practice” model for usual care in the emergency department (ED) regarding the timing of vasopressor initiation in sepsis. In a retrospective study of 589 septic patients with hypotension in an urban tertiary care center ED, we developed a multi-variable model that distinguishes between patients who did and did not subsequently receive sustained (>24 h) vasopressor therapy. Candidate predictors were vital signs, intravenous fluid (IVF) volumes, laboratory measurements, and elapsed time from triage computed at timepoints leading up to the final decision timepoint of either vasopressor initiation or ED hypotension resolution without vasopressors. A model with six independently significant covariates (respiratory rate, Glasgow Coma Scale score, SBP, SpO2, administered IVF, and elapsed time) achieved a C-statistic of 0.78 in a held-out test set at the final decision timepoint, demonstrating the ability to reliably model usual care for vasopressor initiation for hypotensive septic patients. The included variables measured depth of hypotension, extent of disease severity and organ dysfunction. At an operating point of 90% specificity, the model identified a minority of patients (39%) more than an hour before actual vasopressor initiation, during which time a median of 2,250 (IQR 1,200–3,300) mL of IVF was administered. This single-center analysis shows the feasibility of a quantitative, objective tool for describing usual care. Dynamic practice models may help assess when management was atypical; such tools may also be useful for designing and interpreting clinical trials.


  • Is novel coronavirus 2019 reinfection possible? Interpreting dynamic SARS-CoV-2 test results

    by Nicole M. Duggan, Stephanie M. Ludy, Bryant C. Shannon, Andrew T. Reisner, Susan R. Wilcox


  • Open Access

    Closed-Loop Fluid Resuscitation Control Via Blood Volume Estimation

    by Ramin Bighamian, Chang-Sei Kim, Andrew T. Reisner, Jin-Oh Hahn

    This paper presents a closed-loop control of fluid resuscitation to overcome hypovolemia based on model-based estimation of relative changes in blood volume (BV). In this approach, the control system consists of a model-based relative BV (RBV) estimator and a feedback controller. The former predicts relative changes in the BV response to augmented fluid by analyzing an arterial blood pressure (BP) waveform and the electrocardiogram (ECG). Then, the latter determines the amount of fluid to be augmented by comparing target versus predicted relative changes in BV. In this way, unlike many previous methods for fluid resuscitation based on controlled variable(s) nonlinearly correlated with the changes in BV, fluid resuscitation can be guided by a controlled variable linearly correlated with the changes in BV. This paper reports initial design of the closed-loop fluid resuscitation system and its in silico evaluation in a wide range of hypovolemic scenarios. The results suggest that closed-loop fluid resuscitation guided by a controlled variable linearly correlated with the changes in BV can be effective in overcoming hypovolemia: across 100 randomly produced hypovolemia cases, it resulted in the BV regulation error of 7.98 ± 171.6 ml, amounting to 0.18 ± 3.04% of the underlying BV. When guided by pulse pressure (PP), a classical controlled variable nonlinearly correlated with the changes in BV; the same closed-loop fluid resuscitation system resulted in persistent under-resuscitation with the BV regulation error of −779.1 ± 147.4 ml, amounting to −13.9 ± 2.65% of the underlying BV.


  • Identification of Multichannel Cardiovascular Dynamics Using Dual Laguerre Basis Functions for Noninvasive Cardiovascular Monitoring

    by Jin-Oh Hahn, Devin B. McCombie, Andrew T. Reisner, Horacio M. Hojman, H. Harry Asada

    This paper presents a novel method to identify the cardiovascular (CV) system using two distinct peripheral blood pressure (BP) signals. The method can characterize the distinct arterial path dynamics that shape each of the BP signals and recover the common central-flow signal fed to them. A Laguerre series data-compression technique is used to obtain a compact representation of the CV system, whose coefficients are identified using the multichannel blind system identification. A Laguerre model deconvolution algorithm is developed to stably recover the central-flow signal. Persistent excitation, model identifiability, and asymptotic variance are analyzed to quantify the method's validity and reliability, without using any direct measurement of central-flow input signal. Experimental results based on 7000 data segments obtained from nine swine subjects show that, for all the swine subjects under diverse physiologic conditions, the CV dynamics can be identified very reliably and the waveform of the central flow can be recovered stably from peripheral BP signals.


  • Identification of Multi-Channel Cardiovascular System Using Dual-Pole Laguerre Basis Functions for Assessment of Aortic Flow and TPR

    by Jin-Oh Hahn, Andrew T. Reisner, H. Harry Asada

    This paper presents a novel method to identify the multichannel cardiovascular system using two distinct peripheral blood pressure signals. The method can characterize the distinct arterial path dynamics that shape each of the blood pressure signals, and recover the common aortic flow signal fed to them. A Laguerre series data compression technique is used to obtain a compact representation of the cardiovascular system, whose coefficients are identified using the multi-channel blind system identification. A Laguerre model de-convolution algorithm is developed to stably recover the aortic flow signal. Persistent excitation, model identifiability, and asymptotic variance are analyzed to quantify the method’s validity and reliability. From the identified Laguerre series coefficients of the cardiovascular dynamics, mean aortic flow and total peripheral resistance are estimated. Experimental results based on 7,000 data segments obtained from 9 swine models show that the waveform of the aortic flow is stably recovered from peripheral blood pressure signals and that the cardiovascular dynamics can be identified very reliably for all the swine models under diverse physiologic conditions. In addition, the use of the identified cardiovascular dynamics results in the improvement in estimating the mean aortic flow and total peripheral resistance by 60% and 45% in terms of the R2 value, compared to their standard counterparts.