The Experts below are selected from a list of 61080 Experts worldwide ranked by ideXlab platform
Harold S Kaplan - One of the best experts on this subject based on the ideXlab platform.
-
experience with the Medical Event reporting system for transfusion medicine mers tm at three hospitals
Transfusion and Apheresis Science, 2004Co-Authors: Jeannie Callum, Lisa L. Merkley, Ahmed S Coovadia, Ana Lima, Harold S KaplanAbstract:Abstract Background : The MERS-TM assists hospital transfusion services to identify, analyze, and correct system Events relating to the delivery of blood to patients. Methods : The MERS-TM system was used from February of 1999 to December 2002. All reported near-miss and actual Events were recorded and analyzed. Results : During these 47 months, 4670 Events were reported by the transfusion service. Of these Events, 94% were classified as a near-miss Event and 93% were detected before the blood product was administered. No ABO-incompatible transfusions were detected despite transfusion of 50,137 units of red blood cells. High severity Events with the potential for patient harm accounted for 241 (5%) of the 4670 Events. Nursing related Events accounted for 188 (78%) of the high severity Events. In one out of 4430 (0.023%) samples tested, a high severity sample-testing Event was detected. In one out of 1550 (0.06%) samples collected, a high severity sample-collection Event was detected. Conclusion : An Event reporting system is essential if one is to determine where and how often Events are occurring within the transfusion process.
-
The Medical Event reporting system for transfusion medicine: Will it help get the right blood to the right patient?
Transfusion medicine reviews, 2002Co-Authors: Harold S Kaplan, Jeannie Callum, Barbara Rabin Fastman, Lisa L. MerkleyAbstract:The Medical Event Reporting System for Transfusion Medicine (MERS-TM) collects, classifies, and analyzes Events that potentially could compromise the safety of transfused blood to facilitate system improvement. This system is designed to collect data on near misses as well as actual Events. Near-miss Events are a valuable source of data because they occur more frequently than, but share many characteristics and causes of, actual Events. Further, although most current reporting efforts describe only what has occurred with little attention to what caused the Event, MERS-TM includes a standardized method of causal analysis. The standardization provided by MERS allows users to compare their experivided by MERS allows users to compare their experience with that of other organizations, which speeds learning across the entire transfusion medicine community. Important features of the MERS-TM system are that it is able to capture threats, hazards, near misses, injuries, and deaths; characterizes failures and recoveries systematically; identifies and provides causal codes for the entire range of system defects including technical, organizational, cultural, and human factors; raises staff awareness about error management; is easily integrated with existing quality assurance programs; has a consistent and straightforward classification method; enables compliance with mandatory Food and Drug Administration reporting and accreditation requirements; has features to deal with a high volume of reports; supplies Web-based training, data entry, and analysis; and provides comparative benchmarks from comparable institutions.
-
the attributes of Medical Event reporting systems experience with a prototype Medical Event reporting system for transfusion medicine
Archives of Pathology & Laboratory Medicine, 1998Co-Authors: J B Battles, Harold S Kaplan, T W Van Der Schaaf, C E SheaAbstract:. Objective.- To design, develop, and implement a prototype Medical Event-reporting system for use in transfusion medicine to improve transfusion safety by studying incidents and errors. Methods.-The IDEALS concept of design was used to identify specifications for the Event-reporting system, and a Delphi and subsequent nominal group technique meetings were used to reach consensus on the development of the system. An interdisciplinary panel of experts from aviation safety, nuclear power, cognitive psychology, artificial intelligence, and education and representatives of major transfusion medicine organizations participated in the development process. Setting.-Three blood centers and three hospital transfusion services implemented the reporting system. Results.-A working prototype Event-reporting system was recommended and implemented. The system has seven components: detection, selection, description, classification, computation, interpretation, and local evaluation. Its unique features include no-fault reporting initiated by the individual discovering the Event, who submits a report that is investigated by local quality assurance personnel and forwarded to a nonregulatory central system for computation and interpretation. Conclusions.-An Event-reporting system incorporated into present quality assurance and risk management efforts can help organizations address system structural and procedural weakness where the potential for errors can adversely affect health care outcomes. Input from the end users of the system as well as from external experts should enable this reporting system to serve as a useful model for others who may develop Event-reporting systems in other Medical domains.
Thomas A Lasko - One of the best experts on this subject based on the ideXlab platform.
-
efficient inference of gaussian process modulated renewal processes with application to Medical Event data
Uncertainty in Artificial Intelligence, 2014Co-Authors: Thomas A LaskoAbstract:The episodic, irregular and asynchronous nature of Medical data render them difficult substrates for standard machine learning algorithms. We would like to abstract away this difficulty for the class of time-stamped categorical variables (or Events) by modeling them as a renewal process and inferring a probability density over non-parametric longitudinal intensity functions that modulate the process. Several methods exist for inferring such a density over intensity functions, but either their constraints prEvent their use with our potentially bursty Event streams, or their time complexity renders their use intractable on our long-duration observations of high-resolution Events, or both. In this paper we present a new efficient and flexible inference method that uses direct numeric integration and smooth interpolation over Gaussian processes. We demonstrate that our direct method is up to twice as accurate and two orders of magnitude more efficient than the best existing method (thinning). Importantly, our direct method can infer intensity functions over the full range of bursty to memoryless to regular Events, which thinning and many other methods cannot do. Finally, we apply the method to clinical Event data and demonstrate a simple example application facilitated by the abstraction.
-
efficient inference of gaussian process modulated renewal processes with application to Medical Event data
arXiv: Machine Learning, 2014Co-Authors: Thomas A LaskoAbstract:The episodic, irregular and asynchronous nature of Medical data render them difficult substrates for standard machine learning algorithms. We would like to abstract away this difficulty for the class of time-stamped categorical variables (or Events) by modeling them as a renewal process and inferring a probability density over continuous, longitudinal, nonparametric intensity functions modulating that process. Several methods exist for inferring such a density over intensity functions, but either their constraints and assumptions prEvent their use with our potentially bursty Event streams, or their time complexity renders their use intractable on our long-duration observations of high-resolution Events, or both. In this paper we present a new and efficient method for inferring a distribution over intensity functions that uses direct numeric integration and smooth interpolation over Gaussian processes. We demonstrate that our direct method is up to twice as accurate and two orders of magnitude more efficient than the best existing method (thinning). Importantly, the direct method can infer intensity functions over the full range of bursty to memoryless to regular Events, which thinning and many other methods cannot. Finally, we apply the method to clinical Event data and demonstrate the face-validity of the abstraction, which is now amenable to standard learning algorithms.
Lisa L. Merkley - One of the best experts on this subject based on the ideXlab platform.
-
experience with the Medical Event reporting system for transfusion medicine mers tm at three hospitals
Transfusion and Apheresis Science, 2004Co-Authors: Jeannie Callum, Lisa L. Merkley, Ahmed S Coovadia, Ana Lima, Harold S KaplanAbstract:Abstract Background : The MERS-TM assists hospital transfusion services to identify, analyze, and correct system Events relating to the delivery of blood to patients. Methods : The MERS-TM system was used from February of 1999 to December 2002. All reported near-miss and actual Events were recorded and analyzed. Results : During these 47 months, 4670 Events were reported by the transfusion service. Of these Events, 94% were classified as a near-miss Event and 93% were detected before the blood product was administered. No ABO-incompatible transfusions were detected despite transfusion of 50,137 units of red blood cells. High severity Events with the potential for patient harm accounted for 241 (5%) of the 4670 Events. Nursing related Events accounted for 188 (78%) of the high severity Events. In one out of 4430 (0.023%) samples tested, a high severity sample-testing Event was detected. In one out of 1550 (0.06%) samples collected, a high severity sample-collection Event was detected. Conclusion : An Event reporting system is essential if one is to determine where and how often Events are occurring within the transfusion process.
-
The Medical Event reporting system for transfusion medicine: Will it help get the right blood to the right patient?
Transfusion medicine reviews, 2002Co-Authors: Harold S Kaplan, Jeannie Callum, Barbara Rabin Fastman, Lisa L. MerkleyAbstract:The Medical Event Reporting System for Transfusion Medicine (MERS-TM) collects, classifies, and analyzes Events that potentially could compromise the safety of transfused blood to facilitate system improvement. This system is designed to collect data on near misses as well as actual Events. Near-miss Events are a valuable source of data because they occur more frequently than, but share many characteristics and causes of, actual Events. Further, although most current reporting efforts describe only what has occurred with little attention to what caused the Event, MERS-TM includes a standardized method of causal analysis. The standardization provided by MERS allows users to compare their experivided by MERS allows users to compare their experience with that of other organizations, which speeds learning across the entire transfusion medicine community. Important features of the MERS-TM system are that it is able to capture threats, hazards, near misses, injuries, and deaths; characterizes failures and recoveries systematically; identifies and provides causal codes for the entire range of system defects including technical, organizational, cultural, and human factors; raises staff awareness about error management; is easily integrated with existing quality assurance programs; has a consistent and straightforward classification method; enables compliance with mandatory Food and Drug Administration reporting and accreditation requirements; has features to deal with a high volume of reports; supplies Web-based training, data entry, and analysis; and provides comparative benchmarks from comparable institutions.
Abhyuday N. Jagannatha - One of the best experts on this subject based on the ideXlab platform.
-
Bidirectional Recurrent Neural Networks for Medical Event Detection in Electronic Health Records.
arXiv: Computation and Language, 2016Co-Authors: Abhyuday N. JagannathaAbstract:Sequence labeling for extraction of Medical Events and their attributes from unstructured text in Electronic Health Record (EHR) notes is a key step towards semantic understanding of EHRs. It has important applications in health informatics including pharmacovigilance and drug surveillance. The state of the art supervised machine learning models in this domain are based on Conditional Random Fields (CRFs) with features calculated from fixed context windows. In this application, we explored various recurrent neural network frameworks and show that they significantly outperformed the CRF models.
-
Bidirectional RNN for Medical Event detection in electronic health records
2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies NAACL HLT 2016 - Proceedin, 2016Co-Authors: Abhyuday N. Jagannatha, Hong YuAbstract:Sequence labeling for extraction of Medical Events and their attributes from unstructured text in Electronic Health Record (EHR) notes is a key step towards semantic understanding of EHRs. It has important applications in health informatics including pharmacovigilance and drug surveillance. The state of the art supervised machine learning models in this domain are based on Conditional Random Fields (CRFs) with features calculated from fixed context windows. In this application, we explored recurrent neural network frameworks and show that they significantly out-performed the CRF models.
-
HLT-NAACL - Bidirectional RNN for Medical Event Detection in Electronic Health Records
Proceedings of the conference. Association for Computational Linguistics. North American Chapter. Meeting, 2016Co-Authors: Abhyuday N. JagannathaAbstract:Sequence labeling for extraction of Medical Events and their attributes from unstructured text in Electronic Health Record (EHR) notes is a key step towards semantic understanding of EHRs. It has important applications in health informatics including pharmacovigilance and drug surveillance. The state of the art supervised machine learning models in this domain are based on Conditional Random Fields (CRFs) with features calculated from fixed context windows. In this application, we explored recurrent neural network frameworks and show that they significantly out-performed the CRF models.
Hong Yu - One of the best experts on this subject based on the ideXlab platform.
-
Bidirectional RNN for Medical Event detection in electronic health records
2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies NAACL HLT 2016 - Proceedin, 2016Co-Authors: Abhyuday N. Jagannatha, Hong YuAbstract:Sequence labeling for extraction of Medical Events and their attributes from unstructured text in Electronic Health Record (EHR) notes is a key step towards semantic understanding of EHRs. It has important applications in health informatics including pharmacovigilance and drug surveillance. The state of the art supervised machine learning models in this domain are based on Conditional Random Fields (CRFs) with features calculated from fixed context windows. In this application, we explored recurrent neural network frameworks and show that they significantly out-performed the CRF models.