Article Preview
TopIntroduction
The adoption of reference ontologies and their deployment for the personalization of multi-version resources has been recently proposed by several authors in the medical informatics domain (Grandi et al., 2012; Riaño et al., 2012; Tu et al., 2011, Wang et al. 2013) (but also in other application fields, e.g., e-Government (Grandi et al., 2009)). In this work, as resources we consider clinical guidelines (Field & Lohr, 1990), that is “best practices” encoding and standardizing health care procedures, either in textual or in executable format, and their personalization with respect to an ontology of diseases, patients or available hospital facilities they are applicable to. In practice, references to ontology classes are added to the computer encoding of resources (e.g., for which an XML (W3C, 2015a) format can conveniently be used) to introduce a sort of semantic indexing of contents representing their applicability, relevance or eligibility with respect to ontology classes. For instance, a given guideline (e.g., involving treatment of heart diseases) may contain different recommendations which are not uniformly applicable to the same classes of patients: one general therapy may be non-applicable to persons who suffer from some metabolic disorders (e.g., diabetes mellitus) or chronic diseases (e.g., kidney failure) or present some addictions (e.g., cocaine); one first-choice drug may not be administered to patients who are already under treatment with possibly interacting drugs (e.g., anticoagulants), or show genetic or acquired hypersensitivity or intolerance to some substances (e.g., patients with enzymatic defects or documented allergies), and so on. Hence, when dealing with a specific patient care case, a physician may be interested in retrieving a personalized version of a clinical guideline, that is a version tailored to his/her use needs by means of all the available personalization coordinates involving the patient's health state, anamnesis and characteristics (e.g., genetic, demographic or preferential) and local settings (including available hospital resources, diagnostic facilities and physicians’ skills). Therefore, the personalized version will only contain recommendations which are safely and effectively applicable by the user to the patient’s specific case. Furthermore, the emergence of patient-centered healthcare (Australian Commission on Safety and Quality in Health Care, 2010) and the development of patient-centered decision support systems (González-Ferrer et al., 2013; Sacchi et al., 2013), with the involvement of empowered patients as final users, requires the adoption of also non-strictly medical characteristics and individual preferences as further personalization coordinates (e.g., level of education, meal schedule and sleep habits).
To this purpose, we introduced in (Grandi et al., 2009; Grandi et al., 2012) a personalization query engine that, starting from a user-supplied list of ontology classes representing values of the semantic personalization coordinates, can exploit semantic indexing to retrieve the relevant contents only and produce a guideline version tailored to a specific use case. Notice that, coherently with ontology-based personalization solutions also proposed in other application fields (Callan et al., 2003; Cantador et al., 2008; Gauch et al., 2003; Middleton et al., 2004; Moreno et al., 2013; Pretschner, 1998; Riecken, 2000; Sieg et al., 2007), we use the term “personalized” as referred to the user of the computer system, that is either the medical care provider or the empowered patient who follows the guideline.