In medicine, and particularly in emergency situations, the survival of the patient sometimes depends on the quality and speed of the diagnosis. As such, important studies are being carried out into acquiring contextual information. The LERUDI (emergency services) project responds to these crucial challenges.
The electronic health record (DMP): expectations of emergency doctors.
"It would be extremely beneficial if doctors at French emergency medical service call centres (Samu Centre 15) and, more generally, emergency medical services structures were able to consult the DMP. Treatment, in both emergency and non-emergency situations, is safer, more effective and more efficient when a patient's documented medical history is available. By questioning a patient, the doctor is able to begin to understand the situation and establish a relationship of trust – a key stage. However, as everyone knows, information on patient history obtained in this way is imprecise and incomplete... Little by little, the DMP will become a visible reality and will establish itself as the universal method of accessing a patient's shared medical data."1
"…Broadly speaking, the time that the emergency doctor spends searching for information in the DMP must be kept within the following boundaries: one minute for the call centre doctor who has to decide upon the immediate response to a call, and fifteen minutes in other emergency situations… As such, an IT tool to help read the DMP is essential."1
It could not be any clearer. Emergency doctors are convinced that the DMP is worthwhile and they will use it. However, they have set their own standard requirement: a time constraint.
How to manage time constraints to improve the quality of care
The emergency doctor must be able to quickly access the relevant information contained in the patient's medical file, according to the context of the treatment. Within the framework of the emergency services project, the development of a text-based search engine prototype thus offers major potential:
Improving the quality of care means saving time and accurate diagnoses
The emergency doctor's diagnosis, whatever the emergency situation, is based on contextual information (circumstances, motives, etc) and the symptoms presented by the patient. Certain information about the patient's medical history is also available to the doctor, but the patient's account of it is rarely precise, usually incomplete, or even non-existent when patients are in shock or unconscious.
When the option is there, it is thought that an accident and emergency doctor consults the patient's hospital file in about 30 to 40% of emergency cases, but that this option would be useful in 60 to 70% of cases.2 The reasons for this gap are the absence of the patient's hospital file or lack of time to go through the existing file, whether it is a paper file (there is not enough time to "retrieve" it) or a computerized file (the time it takes to consult it in the hospital information system is considered too long). For French emergency call centres and the French mobile emergency and resuscitation service (SMUR), even when the patient has a file in the hospital information system the emergency doctor does not have time to find the file and consult it. Lastly, the search for documentation (access to reports, discharge notices, etc) is one of the main causes of "lost time" in accident and emergency departments3.
If it were possible to access the patient's file and find out useful or even decisive information that had been forgotten or hidden by the patient, the emergency doctor would be able to make a more reliable diagnosis, whatever the state of the patient. This would also optimize patient reception, triage and orientation processes within the hospital.
Using reliable semantic analysis tools means maximizing the potential of the DMP's added value
Accident and emergency departments in health institutions take in the highest proportion of "city" patients who do not have medical files in the institution. This is even more true in densely populated urban areas where there are multiple institutions (such as the Ile de France for example), and none of them can claim to serve the whole of their designated zones, admitting all patients requiring hospitalization in their catchment areas.
The DMP, a tool for coordinating care, will therefore be particularly useful for these services. When the patient arrives, it will often be the only reliable and available medical file.
By applying and perfecting techniques for the semantic analysis of text and medical data, the emergency services project must ensure the content of DMPs is usable, particularly when DMPs contain a substantial number of medical documents, and the time available to consult them is limited.
Ultimately, the volume of information contained in the DMP will likely become an impediment to its use, unless consulting the record is simplified through its organization and structure, and by tools that allow the relevant information in the patient's file to be displayed in specific situations.
The semantic approach, applied to emergency healthcare within the framework of this project, can be transposed to other medical specialties and will thus encourage use of the DMP.
Developing high-performance information systems means improving the quality of care
ASIP Santé’s objective is to organize healthcare information systems in order to improve the quality and coordination of care. It produces binding, structuring frames of reference/frameworks?, such as the national health identifier and the interoperability framework.
ASIP Santé is developing structured models for reports, standardized in conjunction with specialist organizations and representatives of healthcare professionals, with the ultimate objective of bringing them into widespread use. Structuring the data in this way will enable it to be processed by computer and, as a result, ensure it can be collected anonymously for epidemiological purposes. There is a long way to go before this target is reached.
The LERUDI (emergency services) project and all the concepts it involves (medical ontology, rules for indexing medical documents, and creating a human-machine interface [HMI] that will enable immediate consultation of a medical file and targeted access to the information sought), will allow tools to be developed to help doctors structure this data.
The LERUDI (emergency services) project, applying research and development methods, offers a new approach for evaluating information systems dedicated to improving the quality of care. This will undoubtedly inform future health information systems projects.
Finally, if the results meet expectations, the emergency services project should have a positive impact on the development and improvement of health information systems, which companies will be able to utilize without risk.
The general objective of the LERUDI (emergency services) project is to develop and evaluate a text-based search engine prototype capable of extracting information in a matter of seconds from electronic medical files that is useful or even decisive for the emergency doctor.
1 From a speech given by Marc Giroud – President of the French emergency medical services (Samu Urgences) – at the 2009 emergency services conference.
2 ASIP survey during the preparation phase of the Emergency Services project
3 2004 expert and hospital audit mission: CTN (National Technical Committees) no 6 of 09/06/04: Abstract of the accident and emergency services "lost time" study