Discovering HIV related information by means of association rules and machine learning.

Acquired immunodeficiency syndrome (AIDS) is still one of the main health problems worldwide. It is therefore essential to keep making progress in improving the prognosis and quality of life of affected patients. One way to advance along this pathway is to uncover connections between other disorders associated with HIV/AIDS-so that they can be anticipated and possibly mitigated. We propose to achieve this by using Association Rules (ARs). They allow us to represent the dependencies between a number of diseases and other specific diseases. However, classical techniques systematically generate every AR meeting some minimal conditions on data frequency, hence generating a vast amount of uninteresting ARs, which need to be filtered out. The lack of manually annotated ARs has favored unsupervised filtering, even though they produce limited results. In this paper, we propose a semi-supervised system, able to identify relevant ARs among HIV-related diseases with a minimal amount of annotated training data. Our system has been able to extract a good number of relationships between HIV-related diseases that have been previously detected in the literature but are scattered and are often little known. Furthermore, a number of plausible new relationships have shown up which deserve further investigation by qualified medical experts.

Semi-supervised incremental learning with few examples for discovering medical association rules.

BACKGROUND: Association Rules are one of the main ways to represent structural patterns underlying raw data. They represent dependencies between sets of observations contained in the data. The associations established by these rules are very useful in the medical domain, for example in the predictive health field. Classic algorithms for association rule mining give rise to huge amounts of possible rules that should be filtered in order to select those most likely to be true. Most of the proposed techniques for these tasks are unsupervised. However, the accuracy provided by unsupervised systems is limited. Conversely, resorting to annotated data for training supervised systems is expensive and time-consuming. The purpose of this research is to design a new semi-supervised algorithm that performs like supervised algorithms but uses an affordable amount of training data. METHODS: In this work we propose a new semi-supervised data mining model that combines unsupervised techniques (Fisher's exact test) with limited supervision. Starting with a small seed of annotated data, the model improves results (F-measure) obtained, using a fully supervised system (standard supervised ML algorithms). The idea is based on utilising the agreement between the predictions of the supervised system and those of the unsupervised techniques in a series of iterative steps. RESULTS: The new semi-supervised ML algorithm improves the results of supervised algorithms computed using the F-measure in the task of mining medical association rules, but training with an affordable amount of manually annotated data. CONCLUSIONS: Using a small amount of annotated data (which is easily achievable) leads to results similar to those of a supervised system. The proposal may be an important step for the practical development of techniques for mining association rules and generating new valuable scientific medical knowledge.

An Archetype Query Language interpreter into MongoDB: Managing NoSQL standardized Electronic Health Record extracts systems.

The fast development of today's healthcare and the need to extract new medical knowledge from exponentially-growing volumes of standardized Electronic Health Records data, as required by studies in Precision Medicine, brings up a challenge that may probably only be addressed using NoSQL DBMSs, due to the non-optimal performance of traditional relational DBMSs on standardized data; and these database systems operated by semantic archetype-based query languages, because of the expected generalized extension of standardized EHR systems. An AQL into MongoDB interpreter has been developed to its first version. It translates system-independent AQL queries posed on ISO/EN 13606 standardized EHR extracts into the NoSQL MongoDB query language. The new interpreter has the advantages of both the archetype-based system-independent AQL queries and the dual-model-based standardized EHR extracts stored on document-centric NoSQL DBMSs, such as MongoDB. AQL queries are independent of applications, programming languages and system environments due to the use of the dual model, but EHR extracts featuring this model are best persisted on document-based NoSQL databases. Consequently, the interpreter allows us to query standardized EHR extracts semantically, and also affording optimal performance.

Effectiveness and cost-effectiveness of a multicomponent intervention to implement a clinical practice guideline for systemic lupus erythematosus: protocol for a cluster-randomized controlled trial.

BACKGROUND: Systemic lupus erythematosus (SLE) is a heterogeneous autoimmune disease with significant potential morbidity and mortality. Substantial gaps have been documented between the development and dissemination of clinical practice guidelines (CPG) and their implementation in practice. The aim of this study is to assess the effectiveness and cost-effectiveness of a multi-component knowledge transfer intervention to implement a CPG for the management of SLE (CPG-SLE). METHODS: The study is an open, multicentre, controlled trial with random allocation by clusters to intervention or control. Clusters are four public university hospitals of the Canary Islands Health Service where rheumatologists are invited to participate. Patients diagnosed with SLE at least one year prior to recruitment are selected. Rheumatologists in intervention group receive a short educational group programme to both update their knowledge about SLE management according to CPG-SLE recommendations and to acquire knowledge and training on use of the patient-centred approach, a decision support tool embedded in the electronic clinical record and a quarterly feedback report containing information on management of SLE patients. Primary endpoint is change in self-perceived disease activity. Secondary endpoints are adherence of professionals to CPG-SLE recommendations, health-related quality of life, patient perception of their participation in decision making, attitudes of professionals towards shared decision making, knowledge of professionals about SLE and use of healthcare resources. Calculated sample size is 412 patients. Data will be collected from questionnaires and clinical records. Length of follow-up will be 18 months. Multilevel mixed models with repeated time measurements will be used to analyze changes in outcomes over time. Cost-effectiveness, from both social and healthcare services perspectives, will be analyzed by measuring effectiveness in terms of quality-adjusted life years gained. Deterministic and probabilistic sensitivity analyses are planned. DISCUSSION: Impact of CPGs in clinical practice could be improved by applying proven value interventions to implement them. The results of this ongoing trial are expected to generate important scientifically valid and reproducible information not only on clinical effectiveness but also on cost-effectiveness of a multi-component intervention for implementation of a CPG based on communication technologies for chronic patients in the hospital setting. TRIAL REGISTRATION: ClinicalTrial.gov NCT03537638 . Registered on 25 May 2018.

Executing Complexity-Increasing Queries in Relational (MySQL) and NoSQL (MongoDB and EXist) Size-Growing ISO/EN 13606 Standardized EHR Databases.

This research shows a protocol to assess the computational complexity of querying relational and non-relational (NoSQL (not only Structured Query Language)) standardized electronic health record (EHR) medical information database systems (DBMS). It uses a set of three doubling-sized databases, i.e. databases storing 5000, 10,000 and 20,000 realistic standardized EHR extracts, in three different database management systems (DBMS): relational MySQL object-relational mapping (ORM), document-based NoSQL MongoDB, and native extensible markup language (XML) NoSQL eXist. The average response times to six complexity-increasing queries were computed, and the results showed a linear behavior in the NoSQL cases. In the NoSQL field, MongoDB presents a much flatter linear slope than eXist. NoSQL systems may also be more appropriate to maintain standardized medical information systems due to the special nature of the updating policies of medical information, which should not affect the consistency and efficiency of the data stored in NoSQL databases. One limitation of this protocol is the lack of direct results of improved relational systems such as archetype relational mapping (ARM) with the same data. However, the interpolation of doubling-size database results to those presented in the literature and other published results suggests that NoSQL systems might be more appropriate in many specific scenarios and problems to be solved. For example, NoSQL may be appropriate for document-based tasks such as EHR extracts used in clinical practice, or edition and visualization, or situations where the aim is not only to query medical information, but also to restore the EHR in exactly its original form.

Examining database persistence of ISO/EN 13606 standardized electronic health record extracts: relational vs. NoSQL approaches.

BACKGROUND: The objective of this research is to compare the relational and non-relational (NoSQL) database systems approaches in order to store, recover, query and persist standardized medical information in the form of ISO/EN 13606 normalized Electronic Health Record XML extracts, both in isolation and concurrently. NoSQL database systems have recently attracted much attention, but few studies in the literature address their direct comparison with relational databases when applied to build the persistence layer of a standardized medical information system. METHODS: One relational and two NoSQL databases (one document-based and one native XML database) of three different sizes have been created in order to evaluate and compare the response times (algorithmic complexity) of six different complexity growing queries, which have been performed on them. Similar appropriate results available in the literature have also been considered. RESULTS: Relational and non-relational NoSQL database systems show almost linear algorithmic complexity query execution. However, they show very different linear slopes, the former being much steeper than the two latter. Document-based NoSQL databases perform better in concurrency than in isolation, and also better than relational databases in concurrency. CONCLUSION: Non-relational NoSQL databases seem to be more appropriate than standard relational SQL databases when database size is extremely high (secondary use, research applications). Document-based NoSQL databases perform in general better than native XML NoSQL databases. EHR extracts visualization and edition are also document-based tasks more appropriate to NoSQL database systems. However, the appropriate database solution much depends on each particular situation and specific problem.

Development and validation of a risk stratification model for prediction of disability and hospitalisation in patients with heart failure: a study protocol.

BACKGROUND: Chronic heart failure (CHF) reduces quality of life and causes hospitalisation and death. Identifying predictive factors of such events may help change the natural history of this condition. AIM: To develop and validate a stratification system for classifying patients with CHF, according to their degree of disability and need for hospitalisation due to any unscheduled cause, over a period of 1 year. METHODS AND ANALYSIS: Prospective, concurrent, cohort-type study in two towns in the Madrid autonomous region having a combined population of 1 32 851. The study will include patients aged over 18 years who meet the following diagnostic criteria: symptoms and typical signs of CHF (Framingham criteria) and left ventricular ejection fraction (EF)<50% or structural cardiac lesion and/or diastolic dysfunction in the presence of preserved EF (EF>50%).Outcome variables will be(a) Disability, as measured by the WHO Disability Assessment Schedule V.2.0 Questionnaire, and (b) unscheduled hospitalisations. The estimated sample size is 557 patients, 371 for predictive model development (development cohort) and 186 for validation purposes (validation cohort). Predictive models of disability or hospitalisation will be constructed using logistic regression techniques. The resulting model(s) will be validated by estimating the probability of outcomes of interest for each individual included in the validation cohort. ETHICS AND DISSEMINATION: The study protocol has been approved by the Clinical Research Ethics Committee of La Princesa University Teaching Hospital (PI-705). All results will be published in a peer-reviewed journal and shared with the medical community at conferences and scientific meetings.

Normalized medical information visualization.

A new mark-up programming language is introduced in order to facilitate and improve the visualization of ISO/EN 13606 dual model-based normalized medical information. This is the first time that visualization of normalized medical information is addressed and the programming language is intended to be used by medical non-IT professionals.

Service for the Pseudonymization of Electronic Healthcare Records Based on ISO/EN 13606 for the Secondary Use of Information.

The availability of electronic health data favors scientific advance through the creation of repositories for secondary use. Data anonymization is a mandatory step to comply with current legislation. A service for the pseudonymization of electronic healthcare record (EHR) extracts aimed at facilitating the exchange of clinical information for secondary use in compliance with legislation on data protection is presented. According to ISO/TS 25237, pseudonymization is a particular type of anonymization. This tool performs the anonymizations by maintaining three quasi-identifiers (gender, date of birth, and place of residence) with a degree of specification selected by the user. The developed system is based on the ISO/EN 13606 norm using its characteristics specifically favorable for anonymization. The service is made up of two independent modules: the demographic server and the pseudonymizing module. The demographic server supports the permanent storage of the demographic entities and the management of the identifiers. The pseudonymizing module anonymizes the ISO/EN 13606 extracts. The pseudonymizing process consists of four phases: the storage of the demographic information included in the extract, the substitution of the identifiers, the elimination of the demographic information of the extract, and the elimination of key data in free-text fields. The described pseudonymizing system was used in three telemedicine research projects with satisfactory results. A problem was detected with the type of data in a demographic data field and a proposal for modification was prepared for the group in charge of the drawing up and revision of the ISO/EN 13606 norm.

ccML, a new mark-up language to improve ISO/EN 13606-based electronic health record extracts practical edition.

OBJECTIVE: The objective of this paper is to introduce a new language called ccML, designed to provide convenient pragmatic information to applications using the ISO/EN13606 reference model (RM), such as electronic health record (EHR) extracts editors. EHR extracts are presently built using the syntactic and semantic information provided in the RM and constrained by archetypes. The ccML extra information enables the automation of the medico-legal context information edition, which is over 70% of the total in an extract, without modifying the RM information. MATERIALS AND METHODS: ccML is defined using a W3C XML schema file. Valid ccML files complement the RM with additional pragmatics information. The ccML language grammar is defined using formal language theory as a single-type tree grammar. The new language is tested using an EHR extracts editor application as proof-of-concept system. RESULTS: Seven ccML PVCodes (predefined value codes) are introduced in this grammar to cope with different realistic EHR edition situations. These seven PVCodes have different interpretation strategies, from direct look up in the ccML file itself, to more complex searches in archetypes or system precomputation. DISCUSSION: The possibility to declare generic types in ccML gives rise to ambiguity during interpretation. The criterion used to overcome ambiguity is that specificity should prevail over generality. The opposite would make the individual specific element declarations useless. CONCLUSION: A new mark-up language ccML is introduced that opens up the possibility of providing applications using the ISO/EN13606 RM with the necessary pragmatics information to be practical and realistic.