1
|
Nikolaj Blomberg S, Jensen TW, Porsborg Andersen M, Folke F, Kjær Ersbøll A, Torp-Petersen C, Lippert F, Collatz Christensen H. When the machine is wrong. Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machine-learning model. Resuscitation 2023; 183:109689. [PMID: 36634755 DOI: 10.1016/j.resuscitation.2023.109689] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/07/2022] [Revised: 12/29/2022] [Accepted: 01/02/2023] [Indexed: 01/10/2023]
Abstract
BACKGROUND A machine-learning model trained to recognize emergency calls regarding Out-of-Hospital Cardiac Arrest (OHCA) was tested in clinical practice at Copenhagen Emergency Medical Services (EMS) from September 2018 to December 2019. We aimed to investigate emergency call characteristics where the machine-learning model failed to recognize OHCA or misinterpreted a call as being OHCA. METHODS All emergency calls were linked to the dispatch database and verified OHCAs were identified by linkage to the Danish Cardiac Arrest Registry. Calls with either false negative or false positive predictions of OHCA were evaluated by trained auditors. Descriptive analyses were performed with absolute numbers and percentages reported. RESULTS The machine-learning model processed 169,236 calls to Copenhagen EMS and suspected 5,811 (3.4%) of the calls as OHCA, resulting in 84.5% sensitivity and 97.1% specificity. Among OHCAs not recognised by machine-learning model, a condition completely different from OHCA was presented by caller in 31% of the cases. In 28% of unrecognised calls, patient was reported breathing normally, and language barriers were identified in 23% of the cases. Among falsely suspected OHCA, the patient was reported unconscious in 28% of the cases, and in 13% of the false positive cases the machine-learning model interpreted calls regarding dead patients with irreversible signs of death as OHCA. CONCLUSION Continuous optimization of the language model is needed to improve the prediction of OHCA and thereby improve sensitivity and specificity of the machine-learning model on recognising OHCA in emergency telephone calls.
Collapse
Affiliation(s)
- Stig Nikolaj Blomberg
- Copenhagen Emergency Medical Services, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark.
| | - Theo W Jensen
- Copenhagen Emergency Medical Services, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark
| | | | - Fredrik Folke
- Copenhagen Emergency Medical Services, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark; Department of Cardiology, Herlev Gentofte University Hospital, Copenhagen, Denmark
| | - Annette Kjær Ersbøll
- Copenhagen Emergency Medical Services, Denmark; National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark
| | - Christian Torp-Petersen
- Department of Cardiology, Nordsjællands Hospital, Denmark; Department of Public Health, University of Copenhagen, Denmark
| | - Freddy Lippert
- Copenhagen Emergency Medical Services, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark; Falck, Denmark
| | - Helle Collatz Christensen
- Copenhagen Emergency Medical Services, Denmark; Department of Clinical Medicine, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark; Danish Clinical Quality Program (RKKP), National Clinical Registries, Denmark
| |
Collapse
|
2
|
|
3
|
Gnesin F, Møller AL, Mills EHA, Zylyftari N, Jensen B, Bøggild H, Ringgren KB, Blomberg SNF, Christensen HC, Kragholm K, Lippert F, Folke F, Torp-Pedersen C. Rapid dispatch for out-of-hospital cardiac arrest is associated with improved survival. Resuscitation 2021; 163:176-183. [PMID: 33775800 DOI: 10.1016/j.resuscitation.2021.03.015] [Citation(s) in RCA: 8] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/29/2020] [Revised: 03/02/2021] [Accepted: 03/16/2021] [Indexed: 11/25/2022]
Abstract
AIM As proxy for initiation of the first link in the Chain of Survival by the dispatcher, we aimed to investigate the effect of time to first dispatch on 30-day survival among patients with OHCA ultimately receiving the highest-level emergency medical response. METHODS We linked data on all OHCA unwitnessed by emergency medical services (EMS) treated by Copenhagen EMS from 2016 through 2018 to corresponding emergency call records. Among patients receiving highest priority emergency response, we calculated time to dispatch as time from start of call to time of first dispatch. RESULTS We included 3548 patients with OHCA. Of these, 94.1% received the highest priority response (median time to dispatch 0.84 min, 25th-75th percentile 0.58-1.24 min). Patients with time to dispatch within one minute compared to three or more minutes were more likely to receive bystander cardiopulmonary resuscitation (77.3 vs 54.2%), bystander defibrillation (11.5 vs 6.5%) and defibrillation by emergency medical services (24.1 vs 7.5%) and were 2.6-fold more likely to survive 30 days after the OHCA (P = 0.004). Results from multivariate logistic regression were similar: odds ratio (OR) of survival 0.83 per minute increase (95% confidence interval 0.70-1.00, P = 0.04). However, survival was similar between those who received highest priority response and those who did not: OR of survival 0.88 (95% confidence interval 0.53-1.46, P = 0.61). CONCLUSION Rapid time to dispatch among patients with highest priority response was significantly associated with a higher probability of 30-day survival following OHCA.
Collapse
Affiliation(s)
- Filip Gnesin
- Department of Cardiology, Nordsjællands Hospital, Dyrehavevej 29, 3400 Hillerød, Denmark.
| | | | | | - Nertila Zylyftari
- Department of Cardiology, Copenhagen University Hospital Herlev and Gentofte, Gentofte Hospitalsvej 6, 2900 Hellerup, Denmark
| | - Britta Jensen
- Public Health and Epidemiology, Department of Health Science and Technology, Aalborg University, Niels Jernes Vej 14, 9220 Aalborg E, Denmark
| | - Henrik Bøggild
- Public Health and Epidemiology, Department of Health Science and Technology, Aalborg University, Niels Jernes Vej 14, 9220 Aalborg E, Denmark
| | | | | | - Helle Collatz Christensen
- Copenhagen Emergency Medical Services, University of Copenhagen, Telegrafvej 5, 2750 Ballerup, Denmark; Danish Clinical Quality Program (RKKP), National Clinical Registries, Nordre Fasanvej 57, 2000 Frederiksberg, Denmark
| | - Kristian Kragholm
- Department of Cardiology, Aalborg University Hospital, Hobrovej 18-22, 9000 Aalborg, Denmark
| | - Freddy Lippert
- Copenhagen Emergency Medical Services, University of Copenhagen, Telegrafvej 5, 2750 Ballerup, Denmark; University of Copenhagen, Blegdamsvej 9, 2100 København Ø, Denmark
| | - Fredrik Folke
- Department of Cardiology, Copenhagen University Hospital Herlev and Gentofte, Gentofte Hospitalsvej 6, 2900 Hellerup, Denmark; Copenhagen Emergency Medical Services, University of Copenhagen, Telegrafvej 5, 2750 Ballerup, Denmark; University of Copenhagen, Blegdamsvej 9, 2100 København Ø, Denmark
| | - Christian Torp-Pedersen
- Department of Cardiology, Nordsjællands Hospital, Dyrehavevej 29, 3400 Hillerød, Denmark; Department of Cardiology, Aalborg University Hospital, Hobrovej 18-22, 9000 Aalborg, Denmark
| |
Collapse
|
4
|
Abstract
The European Resuscitation Council (ERC) has produced these Systems Saving Lives guidelines, which are based on the 2020 International Consensus on Cardiopulmonary Resuscitation Science with Treatment Recommendations. The topics covered include chain of survival, measuring performance of resuscitation, social media and smartphones apps for engaging community, European Restart a Heart Day, World Restart a Heart, KIDS SAVE LIVES campaign, lower-resource setting, European Resuscitation Academy and Global Resuscitation Alliance, early warning scores, rapid response systems, and medical emergency team, cardiac arrest centres and role of dispatcher.
Collapse
|
5
|
Diagnosis of out-of-hospital cardiac arrest by emergency medical dispatch: A diagnostic systematic review. Resuscitation 2020; 159:85-96. [PMID: 33253767 DOI: 10.1016/j.resuscitation.2020.11.025] [Citation(s) in RCA: 15] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/21/2020] [Revised: 11/01/2020] [Accepted: 11/18/2020] [Indexed: 12/25/2022]
Abstract
INTRODUCTION Cardiac arrest is a time-sensitive condition requiring urgent intervention. Prompt and accurate recognition of cardiac arrest by emergency medical dispatchers at the time of the emergency call is a critical early step in cardiac arrest management allowing for initiation of dispatcher-assisted bystander CPR and appropriate and timely emergency response. The overall accuracy of dispatchers in recognizing cardiac arrest is not known. It is also not known if there are specific call characteristics that impact the ability to recognize cardiac arrest. METHODS We performed a systematic review to examine dispatcher recognition of cardiac arrest as well as to identify call characteristics that may affect their ability to recognize cardiac arrest at the time of emergency call. We searched electronic databases for terms related to "emergency medical dispatcher", "cardiac arrest", and "diagnosis", among others, with a focus on studies that allowed for calculating diagnostic test characteristics (e.g. sensitivity and specificity). The review was consistent with Grading of Recommendations, Assessment, Development and Evaluation (GRADE) method for evidence evaluation. RESULTS We screened 2520 article titles, resulting in 47 studies included in this review. There was significant heterogeneity between studies with a high risk of bias in 18 of the 47 which precluded performing meta-analyses. The reported sensitivities for cardiac arrest recognition ranged from 0.46 to 0.98 whereas specificities ranged from 0.32 to 1.00. There were no obvious differences in diagnostic accuracy between different dispatching criteria/algorithms or with the level of education of dispatchers. CONCLUSION The sensitivity and specificity of cardiac arrest recognition at the time of emergency call varied across dispatch centres and did not appear to differ by dispatch algorithm/criteria used or education of the dispatcher, although comparisons were hampered by heterogeneity across studies. Future efforts should focus on ways to improve sensitivity of cardiac arrest recognition to optimize patient care and ensure appropriate and timely resource utilization.
Collapse
|
6
|
Adult Basic Life Support: International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations. Resuscitation 2020; 156:A35-A79. [PMID: 33098921 PMCID: PMC7576327 DOI: 10.1016/j.resuscitation.2020.09.010] [Citation(s) in RCA: 76] [Impact Index Per Article: 15.2] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/08/2023]
Abstract
This 2020 International Consensus on Cardiopulmonary Resuscitation (CPR) and Emergency Cardiovascular Care Science With Treatment Recommendations on basic life support summarizes evidence evaluations performed for 20 topics that were prioritized by the Basic Life Support Task Force of the International Liaison Committee on Resuscitation. The evidence reviews include 16 systematic reviews, 3 scoping reviews, and 1 evidence update. Per agreement within the International Liaison Committee on Resuscitation, new or revised treatment recommendations were only made after a systematic review. Systematic reviews were performed for the following topics: dispatch diagnosis of cardiac arrest, use of a firm surface for CPR, sequence for starting CPR (compressions-airway-breaths versus airway-breaths-compressions), CPR before calling for help, duration of CPR cycles, hand position during compressions, rhythm check timing, feedback for CPR quality, alternative techniques, public access automated external defibrillator programs, analysis of rhythm during chest compressions, CPR before defibrillation, removal of foreign-body airway obstruction, resuscitation care for suspected opioid-associated emergencies, drowning, and harm from CPR to victims not in cardiac arrest. The topics that resulted in the most extensive task force discussions included CPR during transport, CPR before calling for help, resuscitation care for suspected opioid-associated emergencies, feedback for CPR quality, and analysis of rhythm during chest compressions. After discussion of the scoping reviews and the evidence update, the task force prioritized several topics for new systematic reviews.
Collapse
|
7
|
Olasveengen TM, Mancini ME, Perkins GD, Avis S, Brooks S, Castrén M, Chung SP, Considine J, Couper K, Escalante R, Hatanaka T, Hung KK, Kudenchuk P, Lim SH, Nishiyama C, Ristagno G, Semeraro F, Smith CM, Smyth MA, Vaillancourt C, Nolan JP, Hazinski MF, Morley PT, Svavarsdóttir H, Raffay V, Kuzovlev A, Grasner JT, Dee R, Smith M, Rajendran K. Adult Basic Life Support: 2020 International Consensus on Cardiopulmonary Resuscitation and Emergency Cardiovascular Care Science With Treatment Recommendations. Circulation 2020; 142:S41-S91. [DOI: 10.1161/cir.0000000000000892] [Citation(s) in RCA: 49] [Impact Index Per Article: 9.8] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 01/18/2023]
Abstract
This2020 International Consensus on Cardiopulmonary Resuscitation(CPR)and Emergency Cardiovascular Care Science With Treatment Recommendationson basic life support summarizes evidence evaluations performed for 22 topics that were prioritized by the Basic Life Support Task Force of the International Liaison Committee on Resuscitation. The evidence reviews include 16 systematic reviews, 5 scoping reviews, and 1 evidence update. Per agreement within the International Liaison Committee on Resuscitation, new or revised treatment recommendations were only made after a systematic review.Systematic reviews were performed for the following topics: dispatch diagnosis of cardiac arrest, use of a firm surface for CPR, sequence for starting CPR (compressions-airway-breaths versus airway-breaths-compressions), CPR before calling for help, duration of CPR cycles, hand position during compressions, rhythm check timing, feedback for CPR quality, alternative techniques, public access automated external defibrillator programs, analysis of rhythm during chest compressions, CPR before defibrillation, removal of foreign-body airway obstruction, resuscitation care for suspected opioid-associated emergencies, drowning, and harm from CPR to victims not in cardiac arrest.The topics that resulted in the most extensive task force discussions included CPR during transport, CPR before calling for help, resuscitation care for suspected opioid-associated emergencies, feedback for CPR quality, and analysis of rhythm during chest compressions. After discussion of the scoping reviews and the evidence update, the task force prioritized several topics for new systematic reviews.
Collapse
|
8
|
Mathiesen WT, Birkenes TS, Lund H, Ushakova A, Søreide E, Bjørshol CA. Public knowledge and expectations about dispatcher assistance in out-of-hospital cardiac arrest. J Adv Nurs 2018; 75:783-792. [PMID: 30375018 DOI: 10.1111/jan.13886] [Citation(s) in RCA: 6] [Impact Index Per Article: 0.9] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/19/2018] [Revised: 08/22/2018] [Accepted: 10/02/2018] [Indexed: 11/30/2022]
Abstract
AIM To assess the factors associated with the knowledge and expectations among the general public about dispatcher assistance in out-of-hospital cardiac arrest incidents. BACKGROUND In medical dispatch centres, emergency calls are frequently operated by specially trained nurses as dispatchers. In cardiac arrest incidents, efficient communication between the dispatcher and the caller is vital for prompt recognition and treatment of the cardiac arrest. DESIGN A cross-sectional observational survey containing six questions and seven demographic items. METHOD From January-June 2017 we conducted standardized interviews among 500 members of the general public in Norway. In addition to explorative statistical methods, we used multivariate logistic analysis. RESULTS Most participants expected cardiopulmonary resuscitation instructions, while few expected "help in deciding what to do." More than half regarded the bystanders present to be responsible for the decision to initiate cardiopulmonary resuscitation. Most participants were able to give the correct emergency medical telephone number. The majority knew that the emergency call would not be terminated until the ambulance arrived at the scene. However, only one-third knew that the emergency telephone number operator was a trained nurse. CONCLUSION The public expect cardiopulmonary resuscitation instructions from the emergency medical dispatcher. However, the majority assume it is the responsibility of the bystanders to make the decision to initiate cardiopulmonary resuscitation or not. Based on these findings, cardiopulmonary resuscitation training initiatives and public campaigns should focus more on the role of the emergency medical dispatcher as the team leader of the first resuscitation team in cardiac arrest incidents.
Collapse
Affiliation(s)
- Wenche T Mathiesen
- Critical Care and Anaesthesiology Research Group, Stavanger University Hospital, Stavanger, Norway.,Department of Research and Development, Norwegian Air Ambulance Foundation, Drøbak, Norway
| | | | - Helene Lund
- Emergency Medical Communication Centre, Division of Prehospital Medicine, Stavanger University Hospital, Stavanger, Norway
| | - Anastasia Ushakova
- Department of Research, Stavanger University Hospital, Stavanger, Norway
| | - Eldar Søreide
- Critical Care and Anaesthesiology Research Group, Stavanger University Hospital, Stavanger, Norway.,Department of Clinical Medicine, University of Bergen, Bergen, Norway
| | - Conrad A Bjørshol
- Department of Clinical Medicine, University of Bergen, Bergen, Norway.,Department of Anesthesiology and Intensive Care, Stavanger University Hospital, Stavanger, Norway.,The Regional Centre for Emergency Medical Research and Development (RAKOS), Clinic of Prehospital Medicine, Stavanger University Hospital, Stavanger, Norway
| |
Collapse
|