Characterizing database granularity using SNOMED-CT hierarchy.

Multi-center observational studies require recognition and reconciliation of differences in patient representations arising from underlying populations, disparate coding practices and specifics of data capture. This leads to different granularity or detail of concepts representing the clinical facts. For researchers studying certain populations of interest, it is important to ensure that concepts at the right level are used for the definition of these populations. We studied the granularity of concepts within 22 data sources in the OHDSI network and calculated a composite granularity score for each dataset. Three alternative SNOMED-based approaches for such score showed consistency in classifying data sources into three levels of granularity (low, moderate and high), which correlated with the provenance of data and country of origin. However, they performed unsatisfactorily in ordering data sources within these groups and showed inconsistency for small data sources. Further studies on examining approaches to data source granularity are needed.

Measuring multimorbidity in older adults: comparing different data sources.

BACKGROUND: Multimorbidity is a global health issue, particularly for older adults in the primary care setting. An adequate portrayal of its epidemiology is essential to properly identify and understand the health care needs of this population. This study aimed to compare the differences in the prevalence of selected chronic conditions and multimorbidity, including its associated characteristics, using health survey/self-reported (SR) information only, administrative (Adm) data only and the combined (either) sources. METHODS: This was a secondary analysis of survey data from the first cycle of the Longitudinal Survey on Senior's Health and Health Services Use linked to health-Adm data. The analytical sample consisted of 1625 community-dwelling older adults (≥65 years) recruited in the waiting rooms of primary health clinics in a selected administrative region of the province of Quebec. Seventeen chronic conditions were assessed according to two different data sources. We examined the differences in the observed prevalence of chronic conditions and multimorbidity and the agreement between data sources. RESULTS: The prevalence of each of the 17 chronic conditions ranged from 1.2 to 68.7% depending on the data source. The agreement between different data sources was highly variable, with kappa coefficients (κ) ranging from 0.05 to 0.73. Multimorbidity was very high in this population, with an estimated prevalence of up to 95.9%. In addition, we found that the association between sociodemographic and behavioural factors and the presence of multimorbidity varied according to the different data sources and thresholds. CONCLUSIONS: This is the first study to simultaneously investigate chronic conditions and multimorbidity prevalence among primary care older adults using combined SR and health-Adm data. Our results call attention to (1) the possibility of underestimating cases when using a single data source and (2) the potential benefits of integrating information from different data sources to increase case identification. This is an important aspect of characterizing the health care needs of this fast-growing population.

The Role of Data Type and Recipient in Individuals' Perspectives on Sharing Passively Collected Smartphone Data for Mental Health: Cross-Sectional Questionnaire Study.

BACKGROUND: The growing field of personal sensing harnesses sensor data collected from individuals' smartphones to understand their behaviors and experiences. Such data could be a powerful tool within mental health care. However, it is important to note that the nature of these data differs from the information usually available to, or discussed with, health care professionals. To design digital mental health tools that are acceptable to users, understanding how personal sensing data can be used and shared is critical. OBJECTIVE: This study aimed to investigate individuals' perspectives about sharing different types of sensor data beyond the research context, specifically with doctors, electronic health record (EHR) systems, and family members. METHODS: A questionnaire assessed participants' comfort with sharing six types of sensed data: physical activity, mood, sleep, communication logs, location, and social activity. Participants were asked about their comfort with sharing these data with three different recipients: doctors, EHR systems, and family members. A series of principal component analyses (one for each data recipient) was performed to identify clusters of sensor data types according to participants' comfort with sharing them. Relationships between recipients and sensor clusters were then explored using generalized estimating equation logistic regression models. RESULTS: A total of 211 participants completed the questionnaire. The majority were female (171/211, 81.0%), and the mean age was 38 years (SD 10.32). Principal component analyses consistently identified two clusters of sensed data across the three data recipients: "health information," including sleep, mood, and physical activity, and "personal data," including communication logs, location, and social activity. Overall, participants were significantly more comfortable sharing any type of sensed data with their doctor than with the EHR system or family members (P<.001) and more comfortable sharing "health information" than "personal data" (P<.001). Participant characteristics such as age or presence of depression or anxiety did not influence participants' comfort with sharing sensed data. CONCLUSIONS: The comfort level in sharing sensed data was dependent on both data type and recipient, but not individual characteristics. Given the identified differences in comfort with sensed data sharing, contextual factors of data type and recipient appear to be critically important as we design systems that harness sensor data for mental health treatment and support.

A Clinical Decision Support Framework for Heterogeneous Data Sources.

To keep pace with the developments in medical informatics, health medical data is being collected continually. But, owing to the diversity of its categories and sources, medical data has become so complicated in many hospitals that it now needs a clinical decision support (CDS) system for its management. To effectively utilize the accumulating health data, we propose a CDS framework that can integrate heterogeneous health data from different sources such as laboratory test results, basic information of patients, and health records into a consolidated representation of features of all patients. Using the electronic health medical data so created, multilabel classification was employed to recommend a list of diseases and thus assist physicians in diagnosing or treating their patients' health issues more efficiently. Once the physician diagnoses the disease of a patient, the next step is to consider the likely complications of that disease, which can lead to more diseases. Previous studies reveal that correlations do exist among some diseases. Considering these correlations, a k-nearest neighbors algorithm is improved for multilabel learning by using correlations among labels (CML-kNN). The CML- kNN algorithm first exploits the dependence between every two labels to update the origin label matrix and then performs multilabel learning to estimate the probabilities of labels by using the integrated features. Finally, it recommends the top N diseases to the physicians. Experimental results on real health medical data establish the effectiveness and practicability of the proposed CDS framework.

Resource Classification for Medical Questions.

We present an approach for manually and automatically classifying the resource type of medical questions. Three types of resources are considered: patient-specific, general knowledge, and research. Using this approach, an automatic question answering system could select the best type of resource from which to consider answers. We first describe our methodology for manually annotating resource type on four different question corpora totaling over 5,000 questions. We then describe our approach for automatically identifying the appropriate type of resource. A supervised machine learning approach is used with lexical, syntactic, semantic, and topic-based feature types. This approach is able to achieve accuracies in the range of 80.9% to 92.8% across four datasets. Finally, we discuss the difficulties encountered in both manual and automatic classification of this challenging task.

Toward automated classification of consumers' cancer-related questions with a new taxonomy of expected answer types.

This article examines methods for automated question classification applied to cancer-related questions that people have asked on the web. This work is part of a broader effort to provide automated question answering for health education. We created a new corpus of consumer-health questions related to cancer and a new taxonomy for those questions. We then compared the effectiveness of different statistical methods for developing classifiers, including weighted classification and resampling. Basic methods for building classifiers were limited by the high variability in the natural distribution of questions and typical refinement approaches of feature selection and merging categories achieved only small improvements to classifier accuracy. Best performance was achieved using weighted classification and resampling methods, the latter yielding an accuracy of F1 = 0.963. Thus, it would appear that statistical classifiers can be trained on natural data, but only if natural distributions of classes are smoothed. Such classifiers would be useful for automated question answering, for enriching web-based content, or assisting clinical professionals to answer questions.

E-Patient Reputation in Health Forums.

Online health forums are increasingly used by patients to get information and help related to their health. However, information reliability in these forums is unfortunately not always guaranteed. Obviously, consequences of self-diagnosis may be severe on the patient's health if measures are taken without consulting a doctor. Many works on trust issues related to social media have been proposed, but most of them mainly focus only on the structure part of the social network (number of posts, number of likes, etc.). In the case of online health forums, a lot of trust and distrust is expressed inside the posted messages and cannot be inferred by only considering the structure. In this study, we rather suggest inferring the user's trustworthiness from the replies he receives in the forum. The proposed method is divided into three main steps: First, the recipient(s) of each post must be identified. Next, the trust or distrust expressed in these posts is evaluated. Finally, the user's reputation is computed by aggregating all the posts he received. Conducted experiments using a manually annotated corpus are encouraging.

Characterizing Patient-Generated Clinical Data and Associated Implications for Electronic Health Records.

Patient-facing technologies are increasingly utilized for direct patient data entry for potential incorporation into the electronic health record. We analyzed patient-entered data during implementation of a patient-facing data entry technology using an online patient portal and clinic-based tablet computers at a University-based tertiary medical center clinic, including entries for past medical history, past surgical history, and social history. Entries were assessed for granularity, clinical accuracy, and the addition of novel information into the record. We found that over half of patient-generated diagnoses were duplicates of lesser or equal granularity compared to previous provider-entered diagnoses. Approximately one fifth of patient-generated diagnoses were found to meet the criteria for new, meaningful additions to the medical record. Our findings demonstrate that while patient-generated data provides important additional information, it may also present challenges including generating inaccurate or less granular information.

Monitoring the Amount of Practical Use of eHealth on National Level by Use of Log Data: Lessons Learned.

This paper set out to define the lessons learned from the process of characterizing the amount of practical use of eHealth on national level by collecting and comparing log data harvested from national logs in the Nordic countries. The health systems of the Nordic countries are quite similar in structure and their eHealth strategies include similar elements, however when confronted with the specific context in the different systems it proved challenging to define a common set of indicators for monitoring the practical use of eHealth. A thorough analysis of context leading to the definitions of the indicators is the basis needed due to the complexity of the data in the national logs. A comprehensive knowledge of the structure that underlines these logs is of utmost importance when striving for collecting comparable data. Although challenging, the process of defining indicators for practical use of eHealth by data harvested trough national logs is not an impossible task, but a task that requires in depth discussions of definitions of indicators as well as a substantial insight into the architecture and content of the national databases. There is need for continuous work on these indicators to ensure their quality and thus make sure that the defined indicators can meaningfully inform eHealth policies.

A prerecognition model for hot topic discovery based on microblogging data.

The microblogging is prevailing since its easy and anonymous information sharing at Internet, which also brings the issue of dispersing negative topics, or even rumors. Many researchers have focused on how to find and trace emerging topics for analysis. When adopting topic detection and tracking techniques to find hot topics with streamed microblogging data, it will meet obstacles like streamed microblogging data clustering, topic hotness definition, and emerging hot topic discovery. This paper schemes a novel prerecognition model for hot topic discovery. In this model, the concepts of the topic life cycle, the hot velocity, and the hot acceleration are promoted to calculate the change of topic hotness, which aims to discover those emerging hot topics before they boost and break out. Our experiments show that this new model would help to discover potential hot topics efficiently and achieve considerable performance.

Secure Teleassistance towards endless medical litigations: identification of liabilities through a protocol using Joint Watermarking-Encryption Evidences.

Teleassistance is defined by the help provided through a telemedicine network by a medical practitioner to one other medical practitioner faced to a difficult case. One of the main limiting factors of its development is the fear of the practitioners to be involved in a litigation. In such a situation, the main issue is to determine as quick and as certain as possible if the damage is in relation with the tort of negligence and the liabilities of each involved physician. After a brief summary of the legal context, we present a protocol combining joint watermarking-encryption and a third party to enforce exchange traceability and therefore to bring valuable electronic evidence in case of teleassistance litigations.

Learning the preferences of physicians for the organization of result lists of medical evidence articles.

BACKGROUND: Online medical knowledge repositories such as MEDLINE and The Cochrane Library are increasingly used by physicians to retrieve articles to aid with clinical decision making. The prevailing approach for organizing retrieved articles is in the form of a rank-ordered list, with the assumption that the higher an article is presented on a list, the more relevant it is. OBJECTIVES: Despite this common list-based organization, it is seldom studied how physicians perceive the association between the relevance of articles and the order in which articles are presented. In this paper we describe a case study that captured physician preferences for 3-element lists of medical articles in order to learn how to organize medical knowledge for decision-making. METHODS: Comprehensive relevance evaluations were developed to represent 3-element lists of hypothetical articles that may be retrieved from an online medical knowledge source such as MEDLINE or The Cochrane Library. Comprehensive relevance evaluations asses not only an article's relevance for a query, but also whether it has been placed on the correct list position. In other words an article may be relevant and correctly placed on a result list (e.g. the most relevant article appears first in the result list), an article may be relevant for a query but placed on an incorrect list position (e.g. the most relevant article appears second in a result list), or an article may be irrelevant for a query yet still appear in the result list. The relevance evaluations were presented to six senior physicians who were asked to express their preferences for an article's relevance and its position on a list by pairwise comparisons representing different combinations of 3-element lists. The elicited preferences were assessed using a novel GRIP (Generalized Regression with Intensities of Preference) method and represented as an additive value function. Value functions were derived for individual physicians as well as the group of physicians. RESULTS: The results show that physicians assign significant value to the 1st position on a list and they expect that the most relevant article is presented first. Whilst physicians still prefer obtaining a correctly placed article on position 2, they are also quite satisfied with misplaced relevant article. Low consideration of the 3rd position was uniformly confirmed. CONCLUSIONS: Our findings confirm the importance of placing the most relevant article on the 1st position on a list and the importance paid to position on a list significantly diminishes after the 2nd position. The derived value functions may be used by developers of clinical decision support applications to decide how best to organize medical knowledge for decision making and to create personalized evaluation measures that can augment typical measures used to evaluate information retrieval systems.

Automatically classifying question types for consumer health questions.

We present a method for automatically classifying consumer health questions. Our thirteen question types are designed to aid in the automatic retrieval of medical answers from consumer health resources. To our knowledge, this is the first machine learning-based method specifically for classifying consumer health questions. We demonstrate how previous approaches to medical question classification are insufficient to achieve high accuracy on this task. Additionally, we describe, manually annotate, and automatically classify three important question elements that improve question classification over previous techniques. Our results and analysis illustrate the difficulty of the task and the future directions that are necessary to achieve high-performing consumer health question classification.

Categorizing biomedicine images using novel image features and sparse coding representation.

BACKGROUND: Images embedded in biomedical publications carry rich information that often concisely summarize key hypotheses adopted, methods employed, or results obtained in a published study. Therefore, they offer valuable clues for understanding main content in a biomedical publication. Prior studies have pointed out the potential of mining images embedded in biomedical publications for automatically understanding and retrieving such images' associated source documents. Within the broad area of biomedical image processing, categorizing biomedical images is a fundamental step for building many advanced image analysis, retrieval, and mining applications. Similar to any automatic categorization effort, discriminative image features can provide the most crucial aid in the process. METHOD: We observe that many images embedded in biomedical publications carry versatile annotation text. Based on the locations of and the spatial relationships between these text elements in an image, we thus propose some novel image features for image categorization purpose, which quantitatively characterize the spatial positions and distributions of text elements inside a biomedical image. We further adopt a sparse coding representation (SCR) based technique to categorize images embedded in biomedical publications by leveraging our newly proposed image features. RESULTS: we randomly selected 990 images of the JPG format for use in our experiments where 310 images were used as training samples and the rest were used as the testing cases. We first segmented 310 sample images following the our proposed procedure. This step produced a total of 1035 sub-images. We then manually labeled all these sub-images according to the two-level hierarchical image taxonomy proposed by 1. Among our annotation results, 316 are microscopy images, 126 are gel electrophoresis images, 135 are line charts, 156 are bar charts, 52 are spot charts, 25 are tables, 70 are flow charts, and the remaining 155 images are of the type "others". A serial of experimental results are obtained. Firstly, each image categorizing results is presented, and next image categorizing performance indexes such as precision, recall, F-score, are all listed. Different features which include conventional image features and our proposed novel features indicate different categorizing performance, and the results are demonstrated. Thirdly, we conduct an accuracy comparison between support vector machine classification method and our proposed sparse representation classification method. At last, our proposed approach is compared with three peer classification method and experimental results verify our impressively improved performance. CONCLUSIONS: Compared with conventional image features that do not exploit characteristics regarding text positions and distributions inside images embedded in biomedical publications, our proposed image features coupled with the SR based representation model exhibit superior performance for classifying biomedical images as demonstrated in our comparative benchmark study.

Bioinformatics clouds for big data manipulation.

UNLABELLED: As advances in life sciences and information technology bring profound influences on bioinformatics due to its interdisciplinary nature, bioinformatics is experiencing a new leap-forward from in-house computing infrastructure into utility-supplied cloud computing delivered over the Internet, in order to handle the vast quantities of biological data generated by high-throughput experimental technologies. Albeit relatively new, cloud computing promises to address big data storage and analysis issues in the bioinformatics field. Here we review extant cloud-based services in bioinformatics, classify them into Data as a Service (DaaS), Software as a Service (SaaS), Platform as a Service (PaaS), and Infrastructure as a Service (IaaS), and present our perspectives on the adoption of cloud computing in bioinformatics. REVIEWERS: This article was reviewed by Frank Eisenhaber, Igor Zhulin, and Sandor Pongor.

Should you search the Internet for information about your acute symptom?

OBJECTIVE: To determine if symptom-related Web sites give sufficient information for users to seek urgent care when warranted. MATERIALS AND METHODS: We reviewed 120 Web sites (15 sites for each of eight acute symptoms). Symptom-related sites were identified with Google, Yahoo!®, and Bing™ searches and focused on potentially hazardous symptoms such as chest pain, shortness of breath, abdominal pain, and syncope. We reviewed each symptom-related site for the presence of critical symptom indicators (key symptom characteristics and associated factors) that triage the user to urgent care. RESULTS: Of the 120 sites reviewed, 41 (33%) contained no critical symptom indicators. No site contained a complete set of critical symptom indicators. Overall, out of the 1,020 total critical symptoms searched for in the sites, we only found 329 (32%). When present, critical symptom indicators were found on the top half of the first page of the site in only 34%. Specific recommendations for further care were absent in 42% of the cases where critical symptom indicators were identified. CONCLUSIONS: Symptom-related sites ranked highly by major search engines lack much of the information needed to make a decision about whether a symptom needs urgent attention. When present, this information is usually not located where users can rapidly access it and often lacks prescriptive guidance for users to seek care. Until more sites contain at least minimal triage advice, relying on an Internet search to help determine the urgency of a symptom could be risky.

The Bellevue Classification System: nursing's voice upon the library shelves.

This article examines the inspiration, construction, and meaning of the Bellevue Classification System (BCS), created during the 1930s for use in the Bellevue School of Nursing Library. Nursing instructor Ann Doyle, with assistance from librarian Mary Casamajor, designed the BCS after consulting with library leaders and examining leading contemporary classification systems, including the Dewey Decimal Classification and Library of Congress, Ballard, and National Health Library classification systems. A close textual reading of the classes, subclasses, and subdivisions of these classification systems against those of the resulting BCS, reveals Doyle's belief that the BCS was created not only to organize the literature, but also to promote the burgeoning intellectualism and professionalism of early twentieth-century American nursing.

A new algorithm for reducing the workload of experts in performing systematic reviews.

OBJECTIVE: To determine whether a factorized version of the complement naïve Bayes (FCNB) classifier can reduce the time spent by experts reviewing journal articles for inclusion in systematic reviews of drug class efficacy for disease treatment. DESIGN: The proposed classifier was evaluated on a test collection built from 15 systematic drug class reviews used in previous work. The FCNB classifier was constructed to classify each article as containing high-quality, drug class-specific evidence or not. Weight engineering (WE) techniques were added to reduce underestimation for Medical Subject Headings (MeSH)-based and Publication Type (PubType)-based features. Cross-validation experiments were performed to evaluate the classifier's parameters and performance. MEASUREMENTS: Work saved over sampling (WSS) at no less than a 95% recall was used as the main measure of performance. RESULTS: The minimum workload reduction for a systematic review for one topic, achieved with a FCNB/WE classifier, was 8.5%; the maximum was 62.2% and the average over the 15 topics was 33.5%. This is 15.0% higher than the average workload reduction obtained using a voting perceptron-based automated citation classification system. CONCLUSION: The FCNB/WE classifier is simple, easy to implement, and produces significantly better results in reducing the workload than previously achieved. The results support it being a useful algorithm for machine-learning-based automation of systematic reviews of drug class efficacy for disease treatment.

Classifying health questions asked by the public using the ICPC-2 classification and a taxonomy of generic clinical questions: an empirical exploration of the feasibility.

In case of an overload of information, structure is needed to make the content of the information accessible and the information flow well-ordered. If we wish to gain insight into the health information needs of the public, a specific research tool is needed. The aim of this study was to investigate the feasibility of using two professional classification structures for medical information to classify health questions asked by the public: one classification for the subject of the question, the International Classification of Primary Care (ICPC-2), and one classification for the nature and type of the question, the Taxonomy for Generic Clinical Questions (TGCQ). Health questions asked during online consultations with health care providers were retrieved (452 subjects for coding) and were given two codes: one code according to the ICPC-2 and one according to the TGCQ. The problems encountered during coding were recorded and analyzed. Nine different clusters of problems arose during classification with the ICPC-2, including issues regarding specificity, lay versus professional terminology, a combination of diverse complaints not complying with a clinical syndrome, and preclinical issues. Nine types of problems were encountered during the classification with the TGCQ: questions about preclinical issues, preventive procedures, name finding, health promotion, where to go for a diagnostic test or therapy, justification of the choice of a test or treatment, and common knowledge. The results of this study are promising, and further investigation of the validity, reliability, and use of these two classification systems to classify health questions asked by the public is desirable. The problems that were encountered should be solved before these professional systems can be used to classify the health information needs of the general public.