Evidence-based Clinical Decision Support Systems for the prediction and detection of three disease states in critical care: A systematic literature review

Background: Clinical decision support (CDS) systems have emerged as tools providing intelligent decision making to address challenges of critical care. CDS systems can be based on existing guidelines or best practices; and can also utilize machine learning to provide a diagnosis, recommendation, or therapy course. Methods: This research aimed to identify evidence-based study designs and outcome measures to determine the clinical effectiveness of clinical decision support systems in the detection and prediction of hemodynamic instability, respiratory distress, and infection within critical care settings. PubMed, ClinicalTrials.gov and Cochrane Database of Systematic Reviews were systematically searched to identify primary research published in English between 2013 and 2018. Studies conducted in the USA, Canada, UK, Germany and France with more than 10 participants per arm were included. Results: In studies on hemodynamic instability, the prediction and management of septic shock were the most researched topics followed by the early prediction of heart failure. For respiratory distress, the most popular topics were pneumonia detection and prediction followed by pulmonary embolisms. Given the importance of imaging and clinical notes, this area combined Machine Learning with image analysis and natural language processing. In studies on infection, the most researched areas were the detection, prediction, and management of sepsis, surgical site infections, as well as acute kidney injury. Overall, a variety of Machine Learning algorithms were utilized frequently, particularly support vector machines, boosting techniques, random forest classifiers and neural networks. Sensitivity, specificity, and ROC AUC were the most frequently reported performance measures. Conclusion: This review showed an increasing use of Machine Learning for CDS in all three areas. Large datasets are required for training these algorithms; making it imperative to appropriately address, challenges such as class imbalance, correct labelling of data and missing data. Recommendations are formulated for the development and successful adoption of CDS systems.

[1]  Lisa J. Martin,et al.  Genetic risk signatures of opioid-induced respiratory depression following pediatric tonsillectomy. , 2014, Pharmacogenomics.

[2]  K. Kashani,et al.  Development and validation of electronic surveillance tool for acute kidney injury: A retrospective analysis. , 2015, Journal of critical care.

[3]  Kei-Hoi Cheung,et al.  Predicting urinary tract infections in the emergency department with machine learning , 2018, PloS one.

[4]  P. Pronovost,et al.  A targeted real-time early warning score (TREWScore) for septic shock , 2015, Science Translational Medicine.

[5]  Keshini Madara Marasinghe.  Computerised clinical decision support systems to improve medication safety in long-term care homes: a systematic review , 2015, BMJ Open.

[6]  Heather Duncan,et al.  The Pediatric Early Warning System score: a severity of illness score to predict urgent medical need in hospitalized children. , 2006, Journal of critical care.

[7]  D. Angus,et al.  Epidemiology of severe sepsis , 2013, Virulence.

[8]  B. Bucher,et al.  Enhanced neonatal surgical site infection prediction model utilizing statistically and clinically significant variables in combination with a machine learning algorithm. , 2018, American journal of surgery.

[9]  Atsuto Maki,et al.  A systematic study of the class imbalance problem in convolutional neural networks , 2017, Neural Networks.

[10]  A. Garg,et al.  PILOTING A RENAL DRUG ALERT SYSTEM FOR PRESCRIBING TO RESIDENTS IN LONG‐TERM CARE , 2011, Journal of the American Geriatrics Society.

[11]  Hamid Mohamadlou,et al.  High-performance detection and early prediction of septic shock for alcohol-use disorder patients , 2016, Annals of medicine and surgery.

[12]  Paul Landais,et al.  Patient healthcare trajectory. An essential monitoring tool: a systematic review , 2017, Health Information Science and Systems.

[13]  David L. Buckeridge,et al.  The effectiveness of a new generation of computerized drug alerts in reducing the risk of injury from drug side effects: a cluster randomized trial , 2012, J. Am. Medical Informatics Assoc..

[14]  K. Soman,et al.  Improved Detection of Invasive Pulmonary Aspergillosis Arising during Leukemia Treatment Using a Panel of Host Response Proteins and Fungal Antigens , 2015, PloS one.

[15]  Munish Goyal,et al.  Impact of time to antibiotics on survival in patients with severe sepsis or septic shock in whom early goal-directed therapy was initiated in the emergency department* , 2010, Critical care medicine.

[16]  S. Asirvatham,et al.  Performance of an Automatic Arrhythmia Classification Algorithm: Comparison to the ALTITUDE Electrophysiologist Panel Adjudications , 2014, Pacing and clinical electrophysiology : PACE.

[17]  James C. Cox,et al.  Higher Quality and Lower Cost from Improving Hospital Discharge Decision Making. , 2016, Journal of economic behavior & organization.

[18]  Marco D Huesch,et al.  Evaluating Report Text Variation and Informativeness: Natural Language Processing of CT Chest Imaging for Pulmonary Embolism. , 2018, Journal of the American College of Radiology : JACR.

[19]  Constantin F. Aliferis,et al.  Medical decision support using machine learning for early detection of late-onset neonatal sepsis , 2014, J. Am. Medical Informatics Assoc..

[20]  Lana Dbeibo,et al.  Assessing patient risk of central line‐associated bacteremia via machine learning , 2018, American journal of infection control.

[21]  Peter Nürnberg,et al.  Classification of patients with sepsis according to blood genomic endotype: a prospective cohort study. , 2017, The Lancet. Respiratory medicine.

[22]  Qiao Li,et al.  Ventricular Fibrillation and Tachycardia Classification Using a Machine Learning Approach , 2014, IEEE Transactions on Biomedical Engineering.

[23]  Ella S. Franklin,et al.  Learning Data-Driven Patient Risk Stratification Models for Clostridium difficile , 2014, Open forum infectious diseases.

[24]  Bin Yao,et al.  A prototype of knowledge-based patient safety event reporting and learning system , 2018, BMC Medical Informatics and Decision Making.

[25]  Tewodros Eguale,et al.  Automated Extraction of VTE Events From Narrative Radiology Reports in Electronic Health Records , 2015, Medical care.

[26]  Linda Gavendo,et al.  Computerized clinical decision support during medication ordering for long-term care residents with renal insufficiency. , 2009, Journal of the American Medical Informatics Association : JAMIA.

[27]  N. Shah,et al.  Implementing Machine Learning in Health Care - Addressing Ethical Challenges. , 2018, The New England journal of medicine.

[28]  Patrick C. Sanger,et al.  A Prognostic Model of Surgical Site Infection Using Daily Clinical Wound Assessment. , 2016, Journal of the American College of Surgeons.

[29]  D. Lynch,et al.  Identification of usual interstitial pneumonia pattern using RNA-Seq and machine learning: challenges and solutions , 2018, BMC Genomics.

[30]  S. Jaber,et al.  Combined Thoracic Ultrasound Assessment during a Successful Weaning Trial Predicts Postextubation Distress , 2017, Anesthesiology.

[31]  Shamim Nemati,et al.  An Interpretable Machine Learning Model for Accurate Prediction of Sepsis in the ICU , 2017, Critical care medicine.

[32]  Matthew M. Churpek,et al.  The Development of a Machine Learning Inpatient Acute Kidney Injury Prediction Model* , 2018, Critical care medicine.

[33]  Atul Nakhasi,et al.  Diagnostic errors in the intensive care unit: a systematic review of autopsy studies , 2012, BMJ quality & safety.

[34]  Raúl San José Estépar,et al.  Disease Staging and Prognosis in Smokers Using Deep Learning in Chest Computed Tomography , 2018, American journal of respiratory and critical care medicine.

[35]  Uli K. Chettipally,et al.  Multicentre validation of a sepsis prediction algorithm using only vital sign data in the emergency department, general ward and ICU , 2018, BMJ Open.

[36]  Uli K. Chettipally,et al.  Prediction of Acute Kidney Injury With a Machine Learning Algorithm Using Electronic Health Record Data , 2017, bioRxiv.

[37]  Christofer Toumazou,et al.  Supervised learning for infection risk inference using pathology data , 2017, BMC Medical Informatics and Decision Making.

[38]  J. Vincent,et al.  Sepsis in Intensive Care Unit Patients: Worldwide Data From the Intensive Care over Nations Audit , 2018, Open forum infectious diseases.

[39]  Bryan R. Conroy,et al.  A clinical prediction model to identify patients at high risk of hemodynamic instability in the pediatric intensive care unit , 2017, Critical Care.

[40]  C. Pagel,et al.  ‘The Score Matters’: wide variations in predictive performance of 18 paediatric track and trigger systems , 2017, Archives of Disease in Childhood.

[41]  Romain Pirracchio,et al.  Incidence, risk factors and prediction of post-operative acute kidney injury following cardiac surgery for active infective endocarditis: an observational study , 2013, Critical Care.

[42]  R P Radecki,et al.  The accuracy of an electronic Pulmonary Embolism Severity Index auto-populated from the electronic health record , 2015, Applied Clinical Informatics.

[43]  D. Sontag,et al.  Comparison of Approaches for Heart Failure Case Identification From Electronic Health Record Data. , 2016, JAMA cardiology.

[44]  Eric A Hoffman,et al.  Registration-based lung mechanical analysis of chronic obstructive pulmonary disease (COPD) using a supervised machine learning framework. , 2013, Academic radiology.

[45]  J. Beyene,et al.  Multicentre validation of the bedside paediatric early warning system score: a severity of illness score to detect evolving critical illness in hospitalised children , 2011, Critical care.

[46]  A. Edwards,et al.  Reaching a consensus on research priorities for supporting women with autoimmune rheumatic diseases during pre-conception, pregnancy and early parenting: A Nominal Group Technique exercise with lay and professional stakeholders , 2018, Wellcome open research.

[47]  Marianne Huebner,et al.  Leveraging electronic health records for predictive modeling of post-surgical complications , 2018, Statistical methods in medical research.

[48]  A V Gundlapalli,et al.  Development and Validation of a Natural Language Processing Tool to Identify Patients Treated for Pneumonia across VA Emergency Departments , 2018, Applied Clinical Informatics.

[49]  David L. Buckeridge,et al.  A novel method of adverse event detection can accurately identify venous thromboembolisms (VTEs) from narrative electronic health record data , 2014, J. Am. Medical Informatics Assoc..

[50]  J. Pickard,et al.  Forewarning of hypotensive events using a Bayesian artificial neural network in neurocritical care , 2019, Journal of Clinical Monitoring and Computing.

[51]  E. Seow,et al.  Outcomes of direct and indirect medical intensive care unit admissions from the emergency department of an acute care hospital: a retrospective cohort study , 2014, BMJ Open.

[52]  Kayvan Najarian,et al.  Accounting for Label Uncertainty in Machine Learning for Detection of Acute Respiratory Distress Syndrome , 2019, IEEE Journal of Biomedical and Health Informatics.

[53]  Sheng Yu,et al.  Classification of CT pulmonary angiography reports by presence, chronicity, and location of pulmonary embolism with natural language processing , 2014, J. Biomed. Informatics.

[54]  Marilyn Hravnak,et al.  Cardiorespiratory instability before and after implementing an integrated monitoring system* , 2011, Critical care medicine.

[55]  Lee M. Christensen,et al.  Natural Language Processing to identify pneumonia from radiology reports , 2013, Pharmacoepidemiology and drug safety.

[56]  Eduardo Iturrate,et al.  Creation of a simple natural language processing tool to support an imaging utilization quality dashboard , 2017, Int. J. Medical Informatics.

[57]  R. Bashir,et al.  Combining Biomarkers with EMR Data to Identify Patients in Different Phases of Sepsis , 2017, Scientific Reports.

[58]  Jing Zhang,et al.  Prediction of Adverse Events in Patients Undergoing Major Cardiovascular Procedures , 2017, IEEE Journal of Biomedical and Health Informatics.

[59]  Yitzhak Mendelson,et al.  Using support vector machines on photoplethysmographic signals to discriminate between hypovolemia and euvolemia , 2018, PloS one.

[60]  Kotagiri Ramamohanarao,et al.  Septic shock prediction for ICU patients via coupled HMM walking on sequential contrast patterns , 2017, J. Biomed. Informatics.

[61]  R. Rattan,et al.  Predicting central line‐associated bloodstream infections and mortality using supervised machine learning , 2018, Journal of critical care.

[62]  Ulysses Balis,et al.  Assessing risk of fibrosis progression and liver-related clinical outcomes among patients with both early stage and advanced chronic hepatitis C , 2017, PloS one.

[63]  E. Coiera,et al.  The Changing Nature of Clinical Decision Support Systems: a Focus on Consumers, Genomics, Public Health and Decision Safety , 2009, Yearbook of Medical Informatics.

[64]  D F Sittig,et al.  Clinical Decision Support: a 25 Year Retrospective and a 25 Year Vision , 2016, Yearbook of Medical Informatics.

[65]  J. Rumsfeld,et al.  Big data analytics to improve cardiovascular care: promise and challenges , 2016, Nature Reviews Cardiology.

[66]  Hongfang Liu,et al.  Detection of clinically important colorectal surgical site infection using Bayesian network. , 2017, The Journal of surgical research.

[67]  T. Claypole,et al.  Short-Term Intra-Subject Variation in Exhaled Volatile Organic Compounds (VOCs) in COPD Patients and Healthy Controls and Its Effect on Disease Classification , 2014, Metabolites.

[68]  Vincent Liu,et al.  Automated identification of pneumonia in chest radiograph reports in critically ill patients , 2013, BMC Medical Informatics and Decision Making.

[69]  Majid Sarrafzadeh,et al.  A flexible data-driven comorbidity feature extraction framework , 2016, Comput. Biol. Medicine.

[70]  C. Quinn,et al.  WellDoc mobile diabetes management randomized controlled trial: change in clinical and behavioral outcomes and patient and physician satisfaction. , 2008, Diabetes technology & therapeutics.

[71]  D. Sontag,et al.  Early Identification of Patients With Acute Decompensated Heart Failure. , 2017, Journal of Cardiac Failure.

[72]  Yong Hu,et al.  Causal risk factor discovery for severe acute kidney injury using electronic health records , 2018, BMC Medical Informatics and Decision Making.

[73]  Loris Nanni,et al.  Coupling different methods for overcoming the class imbalance problem , 2015, Neurocomputing.

[74]  L. Celi,et al.  Intensive Care Unit Telemedicine in the Era of Big Data, Artificial Intelligence, and Computer Clinical Decision Support Systems. , 2019, Critical care clinics.

[75]  Nils Strodthoff,et al.  Detecting and interpreting myocardial infarction using fully convolutional neural networks , 2018, Physiological measurement.

[76]  Jeffrey P. Ferraro,et al.  An ontology-driven, diagnostic modeling system. , 2013, Journal of the American Medical Informatics Association : JAMIA.

[77]  Uli K. Chettipally,et al.  Prediction of Sepsis in the Intensive Care Unit With Minimal Electronic Health Record Data: A Machine Learning Approach , 2016, JMIR medical informatics.

[78]  J. Vincent,et al.  Improving detection of patient deterioration in the general hospital ward environment , 2018, European journal of anaesthesiology.

[79]  Gloria P. Lipori,et al.  MySurgeryRisk: Development and Validation of a Machine-learning Risk Algorithm for Major Complications and Death After Surgery , 2019, Annals of surgery.

[80]  Elias Ebrahimzadeh,et al.  Prediction of paroxysmal Atrial Fibrillation: A machine learning based approach using combined feature vector and mixture of expert classification on HRV signal , 2018, Comput. Methods Programs Biomed..

[81]  Raúl San José Estépar,et al.  Implementation and Performance of Automated Software for Computing Right-to-Left Ventricular Diameter Ratio From Computed Tomography Pulmonary Angiography Images , 2016, Journal of computer assisted tomography.

[82]  D. Bates,et al.  The Critical Care Safety Study: The incidence and nature of adverse events and serious medical errors in intensive care* , 2005, Critical care medicine.

[83]  Joann L. Baril,et al.  EFFECT OF CLINICAL DECISION SUPPORT ON PSYCHOTROPIC MEDICATION PRESCRIBING IN THE LONG‐TERM CARE SETTING , 2010, Journal of the American Geriatrics Society.

[84]  I. Nikiforov,et al.  A prediction model for Clostridium difficile recurrence , 2015, Journal of community hospital internal medicine perspectives.

[85]  Thomas Lavergne,et al.  Natural language processing of radiology reports for the detection of thromboembolic diseases and clinically relevant incidental findings , 2014, BMC Bioinformatics.

[86]  Pavel S. Roshanov,et al.  Development of a Tailored Intervention With Computerized Clinical Decision Support to Improve Quality of Care for Patients With Knee Osteoarthritis: Multi-Method Study , 2018, JMIR research protocols.

[87]  A. Kirk,et al.  Towards precision medicine: Accurate predictive modeling of infectious complications in combat casualties , 2017, The journal of trauma and acute care surgery.