Drug dose titration (DT) is the clinical process of progressively adjusting the dose of a medication for the maximum benefit of the patient. Several DT clinical models exist based on the elementary concepts of null, initial, and maximal doses, as well as, dose increments and decrements. These values depend on the target disease, the drug considered, and some parameters such as the patient’s age, gender, weight, and race. This paper describes the formalization of this knowledge as an ontology, and its use to detect chronic hypertension patient treatment deviations from standard DT models with regard to drug replacement (step-1 treatment) and drug supplementation (step-2 treatment).

In health-care, there is a need to quantify medical errors. Among these errors, we observe wrong dose prescriptions. Drug dose titration (DT) is the process by which dosage is progressively adjusted to the patient till a steady dose is reached. Depending on the clinical disease, drug, and patient condition, dose titration can follow different procedures. Once modeled, these procedures can serve for clinical homogenization, standardization, decision support and retrospective analysis. Here, we propose a language to model dose titration procedures. The language was used to formalize one- and two-drug titration of chronic and acute cases, and to perform retrospective analysis of the drug titration processes on 253 patients diagnosed of diabetes mellitus type 2 and treated with metformin, 321 patients treated of chonic heart failure with furosemide, 155 patients with hyperuricemia treated with allopurinol as initial drug and febuxostat as alternative drug, and 187 hyperuricemia patients with primary drug allopurinol and supplementary drug probenecid, in order to identify different types of drug titration deviations from standard DT methods.

In health-care, medical errors are quantified. Among them, wrong dose prescriptions occur. Drug dose titration (DT) is the process by which dosage is progressively adjusted to the patient till a steady dose is reached. Depending on the clinical disease, drug, and patient, dose titration can follow different procedures. Once modeled, these procedures can serve for clinical homogenization, standardization, decision support and retrospective analysis. Here, we propose a language to model dose titration procedures. The language was used to formalize single-drug titration of chronic and acute cases, and perform retrospective analysis of the drug titration processes on 1,000 cases treated with Bisoprolol and 2,430 cases treated with Ramipril, in order to identify different types of drug titration deviations from standard DT methods.

Patients in intensive care units are heterogeneous and the daily prediction of their days to discharge (DTD) a complex task that practitioners and computers are not always able to solve satisfactorily. In order to make more precise DTD predictors, it is necessary to have tools for the analysis of the heterogeneity of the patients. Unfortunately, the number of publications in this field is almost non-existent. In order to alleviate this lack of tools, we propose four methods and their corresponding measures to quantify the heterogeneity of intensive patients in the process of determining the DTD. These new methods and measures have been tested with patients admitted over four years to a tertiary hospital in Spain. The results deepen the understanding of the intensive patient and can serve as a basis for the construction of better DTD predictors.

Abstract Natural language processing (NLP) has been used to process text pertaining to patient records and narratives. However, most of the methods used were developed for specific systems, so new research is necessary to assess whether such methods can be easily retargeted for new applications and goals, with the same performance. In this paper, open‐source tools are reused as building blocks on which a new system is built. The aim of our work is to evaluate the applicability of the current NLP technology to a new domain: automatic knowledge acquisition of diagnostic and therapeutic procedures from clinical practice guideline free‐text documents. In order to do this, two publicly available syntactic parsers, several terminology resources and a tool oriented to identify semantic predications were tailored to increase the performance of each tool individually. We apply this new approach to 171 sentences selected by the experts from a clinical guideline, and compare the results with those of the tools applied with no tailoring. The results of this paper show that with some adaptation, open‐source NLP tools can be retargeted for new tasks, providing an accuracy that is equivalent to the methods designed for specific tasks.

Clinical Algorithms (CAs) are obtained from Clinical Practice Guidelines (CPGs) using all the healthcare knowledge available to assist patients that suffer from one or several diseases. For some diseases CAs can be explicitly given, while for others we have to use other mechanisms for their generation. Explicitly given CAs use to be atemporal. To provide the time dimension of CAs it is necessary to obtain temporal knowledge from physicians or by some other mechanism. Often, physicians have difficulties in providing temporal knowledge or the knowledge engineering mechanism and tools used are difficult to apply or extremely time consuming. However, as data saved in clinical databases are time dependent, they can be used to implicitly obtain temporal constraints for CAs. We have identified two sorts of temporal constraints (micro- and macro-temporality) and have proposed an approach to generate micro-temporalities that complements previous works. We have decided to use the SDA (state-decision-action) formalism for CA representation. The generated micro-temporality constraints are introduced in the SDA representation of a particular CA. As the final CAs have a time dimension, they are no longer atemporal, which helps physicians making temporal predictions in healthcare procedures.

Proposing a treatment to patients is one of the physicians' most common tasks. There are different elements that influence the decision of a physician to propose an appropriate treatment. Formal intervention plans (FIPs) are formal structures representing health care procedures to assist patients suffering from particular ailments or diseases. The introduction of temporal constraints in FIPs is a difficult task that physicians are not used to. This difficulty can be overcome with mechanisms to generate temporal constraints directly from the existing data on patient treatments. We have chosen the SDA formalism to represent FIPs. Here, our objective is to approximate time constraints from patient state transition sequences and as a generalization of the times assigned to each transition (or patient evolution). This approximation is used to construct the time dimension of FIPs.

There are several domains, such as health-care, in which the decision process usually has a background knowledge that must be considered. We need to maximize the accuracy of the models, but we also need them to be meaningful. Otherwise it will lead to the problem that the expert finds the obtained models incomprehensible. We propose a way for representing the knowledge of the experts in order to modify the C 4.5 algorithm to produce decision trees which are more comprehensible to medical doctors without losing accuracy.

Clinical practice guidelines are important in the work of physicians. These guidelines are manually created by experts using their knowledge and experience. This work gives an approach to automatically develop the clinical guideline charts with the SDA representation model. In addition, this paper details an example of application of the methodology proposed with the treatment of hypertension.

Procedural knowledge in medicine is embedded in Clinical Practice Guidelines whose textual condition makes it difficult to share and to reuse. Several languages for formal definition of clinical practice guidelines have been proposed to overcome these difficulties. In order to deal with the huge amount of medical situations, these languages use to be extensive and complex in such a way that they, and the knowledge they are used to represent, are arduous to understand and to manage by non-trained general practitioners. The SDA* model is introduced as an alternative language that promotes representation capability and simplicity in such a way that not only computers, but also health care professionals are able to understand and manage easily without any sort of training. Here, a description of this model from a set theory perspective is provided.

Clinical practice guidelines (CPG) are used to represent the clinical procedures that must be followed for patient assistance. The utility of these guidelines is counteracted by the costs of making and updating them. Simultaneously, the minimum basic data set (MBDS) is defined as a health-care standard to represent the patients as a reduced number of fourteen data that can be used to share or compare different hospitals or services within the same hospital. Here, an algorithm is proposed that uses the MBDS to automatically generate CPG. The algorithm is tested with the MBDS patient episodes of the Hospital Joan XXIII (Spain) for four pediatric diseases.

The huge amount of medical data in the Internet requires intelligent tools to retrieve and process medical information. In this paper, we introduce GINY, a computer system that integrates several procedures to deal with the structured, semi-structured and non-structured documents that can be found in the Internet. From an ontology that describes a medical concept and its properties, the system retrieves related web pages in the web, analyzes the contents of the pages, and organizes the extracted data in a relational data base (RDB) or a resource description framework (RDF). The stages of the process can be made partially or totally automatic.

The analysis of hospital costs in public health-care systems is a difficult task that the use of diagnosis-related groups (DRGs) has tried to simplify. After more than two decades, DRGs are still used, though they have had important criticisms. In this paper, COSYS (COsting SYStem), a new computer system that incorporates some AI procedures, is introduced as a tool to study hospital costs and also to validate whether DRGs are close to the reality of a particular hospital. Seven representative DRGs are studied and the conclusions from the study are set out.

Palliative care units are hospital services that assist advanced-state terminal patients. These patients are geographically distributed and some hospital tasks such as the organization of visits, the management of medicines, and the analysis of patients are difficult to manage. Here, a multi-layer computer system is proposed. The communication, information system, and multi-agent layers are described and a prototype of the system is used to deal with the Palliative Care Unit at the Hospital de la Santa Creu i Sant Pau in Barcelona.