Article Preview
TopIntroduction
The management of information generated in the field of medicine is a complex task. Some of the most important reasons are the variety of information sources, the amount of information and the diversity of types of data formats used to represent the information (Häyrinen et al., 2008). The main sources of information (McMullen et al., 2011) in this area are the data generated by the machines (readings from sensors, meters and other devices), biometric data (fingerprints, genetic data, retinal scanner, x-ray and other medical images, blood pressure, the pulse and pulse oximetry readings and other similar types of data), data generated by human beings (electronic medical records, health professionals' notes, electronic mails and paper documents and digital support), medical dissemination magazines or about clinical trials, or the information that is generated on the web in general, on health websites, social networks (Facebook, Twitter, Linkedin, etc.) or by smartphone applications related to health.
In the case of data formats, it is necessary to manage data that can be of several types (Walsh, 2004): structured, semi-structured or without structure. For example, structured data are personal data, data from medical tests such as blood and urine tests, etc., semi-structured data such as those that can be obtained from sensors such as electrocardiograms, blood pressure and other vital signs measurements, and data unstructured such as paper recipes, medical records, handwritten notes from doctors and nurses, voice recordings, x-rays, scanners, MRIs, CT scans, and other medical images.
Finally, the amount of medical information generated around each patient throughout his life has gigantic sizes (Poon et al., 2010). In general, this information is generated by different medical specialists and in different types of health centers (hospitals, local medical centers, etc.)
This work does not address the general problem of managing medical information, but focuses on a specific case: the management of clinical reports generated in neurology consultations. Neurology is a very large area with different specialties. This work focuses on neurology consultations that aim to diagnose and treat mental illness.
The operation of a neurology consultation is as follows (Van Der Meijden et al., 2003). The medical center assigns each patient to a specific specialist doctor. The doctor must create a clinical record of the patient. In the record, the doctor must keep the personal data of the patient and other auxiliary data of interest such as a history of neurological diseases in relatives, other diseases that he has had or has, medications to which the patient is allergic, surgical operations... These data are usually static since they will not change over time (except for the discovery of new drugs to which they are allergic, and other subsequent surgical operations). Each consultation with a patient usually consists of carrying out a set of medical tests with the aim of diagnosing a mental illness, assessing the improvement of the patient with respect to a treatment being followed or performing an active therapy with the doctor in the consultation. In either case, the doctor must perform two types of different reports (Fichman, 2011). A report that he performs at the same time that it is performed the tests to the patients. In this report, the doctor must take note of everything that happens during the consultation. For example, if the test consists in carrying out a series of questions to the patient, the questions asked, the answers given by the patient or any other event that occurs should be reflected in the report (patient's reactions to the questions, status psychic or physical of the patient, questions that the patient asks the doctor...). The doctor must do the second report once the consultation has finished and usually without the presence of the patient. In this report, the doctor leaves generic notes on how the consultation has developed, and makes an assessment of the results of the tests. Also, it is able to leave targeted annotations to take into account in the next consultation with the patient.