Big data and machine learning in critical care: Opportunities for collaborative research.

The introduction of clinical information systems (CIS) in Intensive Care Units (ICUs) offers the possibility of storing a huge amount of machine-ready clinical data that can be used to improve patient outcomes and the allocation of resources, as well as suggest topics for randomized clinical trials. Clinicians, however, usually lack the necessary training for the analysis of large databases. In addition, there are issues referred to patient privacy and consent, and data quality. Multidisciplinary collaboration among clinicians, data engineers, machine-learning experts, statisticians, epidemiologists and other information scientists may overcome these problems. A multidisciplinary event (Critical Care Datathon) was held in Madrid (Spain) from 1 to 3 December 2017. Under the auspices of the Spanish Critical Care Society (SEMICYUC), the event was organized by the Massachusetts Institute of Technology (MIT) Critical Data Group (Cambridge, MA, USA), the Innovation Unit and Critical Care Department of San Carlos Clinic Hospital, and the Life Supporting Technologies group of Madrid Polytechnic University. After presentations referred to big data in the critical care environment, clinicians, data scientists and other health data science enthusiasts and lawyers worked in collaboration using an anonymized database (MIMIC III). Eight groups were formed to answer different clinical research questions elaborated prior to the meeting. The event produced analyses for the questions posed and outlined several future clinical research opportunities. Foundations were laid to enable future use of ICU databases in Spain, and a timeline was established for future meetings, as an example of how big data analysis tools have tremendous potential in our field.

Development of a computer-interpretable clinical guideline model for decision support in the differential diagnosis of hyponatremia.

INTRODUCTION: Hyponatremia is the most common type of electrolyte imbalance, occurring when serum sodium is below threshold levels, typically 135mmol/L. Electrolyte balance has been identified as one of the most challenging subjects for medical students, but also as one of the most relevant areas to learn about according to physicians and researchers. We present a computer-interpretable guideline (CIG) model that will be used for medical training to learn how to improve the diagnosis of hyponatremia applying an expert consensus document (ECDs). METHODS: We used the PROForma set of tools to develop the model, using an iterative process involving two knowledge engineers (a computer science Ph.D. and a preventive medicine specialist) and two expert endocrinologists. We also carried out an initial validation of the model and a qualitative post-analysis from the results of a retrospective study (N=65 patients), comparing the consensus diagnosis of two experts with the output of the tool. RESULTS: The model includes over two-hundred "for", "against" and "neutral" arguments that are selectively triggered depending on the input value of more than forty patient-state variables. We share the methodology followed for the development process and the initial validation results, that achieved a high ratio of 61/65 agreements with the consensus diagnosis, having a kappa value of K=0.86 for overall agreement and K=0.80 for first-ranked agreement. CONCLUSION: Hospital care professionals involved in the project showed high expectations of using this tool for training, but the process to follow for a successful diagnosis and application is not trivial, as reported in this manuscript. Secondary benefits of using these tools are associated to improving research knowledge and existing clinical practice guidelines (CPGs) or ECDs. Beyond point-of-care clinical decision support, knowledge-based decision support systems are very attractive as a training tool, to help selected professionals to better understand difficult diseases that are underdiagnosed and/or incorrectly managed.