Multiple valued logic approach for matching patient records in multiple databases

Many problems arise when linking medical records from multiple databases. Matching these data to other data is problematic since even small errors, such as data entry errors, different text format, and missing data, can prevent the exact-match algorithms. Evidence from previous studies suggested that approximate field matching represent a solution to resolve the problem by identifying equivalent string values in different representations. The purpose of this article is to explore the effectiveness of a medical record matching method using a fuzzy logic framework. This article considers quantitative measures of the typical elements in medical records, and fuzzy logic is applied to link to the linguistic concepts. Moreover, this article discusses the medical record matching from the developed framework, which is tested on a public data set. The results from the test on a public data set indicate that the medical record matching method using fuzzy logic framework provides an effective solution for dealing with linkage problems, and illustrate that the multiple valued logic method outlined can potentially be applied to address similar problems in other databases.     

Unambiguous identification of hospital patients: case study at the university departments of the General Hospital, Vienna

This article considers the problem of identifying patients in one or more heterogeneous personal databases. The unambiguous identification of patients is an essential prerequisite for an efficient patient care system. We discuss the problems involved in this task and suggest how they can be dealt with. The solution of automatic consolidation of patient records sequires programming, organisational and work psychology measures. Following a survey of conventional identification methods, the method developed at the Department of Medical Computer Sciences, which is based on the current clinical situation at the General Hospital in Vienna (AKH--Allgemeines KrankenHaus), is described in detail. The basic principle is to identify patients unambiguously by means of an ID (IZAHL) derived directly from the personal data. Thereby a deterministic technique without probability weighting is used-all compared information must correspond completely. The article closes with a critical survey of experience gathered to date.     

A peripheral reading deficit under conditions of diffuse visual attention

We report a single case study of a patient, FL, who has a peripheral dyslexia with symptoms resembling attentional dyslexia. FL demonstrated impaired identification of letters within strings, and better identification of words than their constituent letters. We found that FL was impaired at both letter counting and same-case letter matching with letter strings, and his matching and naming performance were strongly affected by letter spacing. The effects of these visual variables on performance suggest an early locus to FL's deficit. We propose that letter identification was disrupted by abnormal lateral masking and poor location coding within words. These peripheral processing deficits were reduced when the task required focused attention on the central letter location. Nevertheless, even with impaired letter coding, word representations could be accessed to some degree, via supra-letter units. We discuss the implications of the data for understanding normal reading.     

A peripheral reading deficit under conditions of diffuse visual attention

We report a single case study of a patient, FL, who has a peripheral dyslexia with symptoms resembling attentional dyslexia. FL demonstrated impaired identification of letters within strings, and better identification of words than their constituent letters. We found that FL was impaired at both letter counting and same-case letter matching with letter strings, and his matching and naming performance were strongly affected by letter spacing. The effects of these visual variables on performance suggest an early locus to FL's deficit. We propose that letter identification was disrupted by abnormal lateral masking and poor location coding within words. These peripheral processing deficits were reduced when the task required focused attention on the central letter location. Nevertheless, even with impaired letter coding, word representations could be accessed to some degree, via supra-letter units. We discuss the implications of the data for understanding normal reading.     

Temporal compartmental dosing effects for robotic prostate stereotactic body radiotherapy

The rate of dose accumulation within a given area of a target volume tends to vary significantly for non-isocentric delivery systems such as Cyberknife stereotactic body radiotherapy. In this study, we investigated whether intra-target temporal dose distributions produce significant variations in the biological equivalent dose. For the study, time courses of ten patients were reconstructed and calculation of a biologically equivalent uniform dose (EUD) was performed using a formula derived from the linear quadratic model (α/β = 3 for prostate cancer cells). The calculated EUD values obtained for the actual patient treatments were then compared with theoretical EUD values for delivering the same physical dose distribution except that the whole target being irradiated continuously (e.g. large-field 'dose-bathing' type of delivery). For all the case, the EUDs for the actual treatment delivery were found to correlate strongly with the EUDs for the large-field delivery: a linear correlation coefficient of R2 = 0.98 was obtained and the average EUD for the actual Cyberknife delivery was somewhat higher (5.0 ± 4.7%) than that for the large-field delivery. However, no statistical significance was detected between the two types of delivery (p = 0.21). We concluded that non-isocentric small-field Cyberknife delivery produced consistent biological dosing that tracked well with the constant-dose-rate, large-field-type delivery for prostate stereotactic body radiotherapy.     

Towards the development of a conceptual distance metric for the UMLS

The objective of this work is to investigate the feasibility of conceptual similarity metrics in the framework of the Unified Medical Language System (UMLS). We have investigated an approach based on the minimum number of parent links between concepts, and evaluated its performance relative to human expert estimates on three sets of concepts for three terminologies within the UMLS (i.e., MeSH, ICD9CM, and SNOMED). The resulting quantitative metric enables computer-based applications that use decision thresholds and approximate matching criteria. The proposed conceptual matching supports problem solving and inferencing (using high-level, generic concepts) based on readily available data (typically represented as low-level, specific concepts). Through the identification of semantically similar concepts, conceptual matching also enables reasoning in the absence of exact, or even approximate, lexical matching. Finally, conceptual matching is relevant for terminology development and maintenance, machine learning research, decision support system development, and data mining research in biomedical informatics and other fields.     

Methodological framework for evaluating clinical processes: A cognitive informatics perspective

We propose a methodological framework for evaluating clinical cognitive activities in complex real-world environments that provides a guiding framework for characterizing the patterns of activities. This approach, which we refer to as a process-based approach, is particularly relevant to cognitive informatics (CI) research—an interdisciplinary domain utilizing cognitive approaches in the study of computing systems and applications—as it provides new ways for understanding human information processing, interactions, and behaviors. Using this approach involves the identification of a process of interest (e.g., a clinical workflow), and the contributing sequences of activities in that process (e.g., medication ordering). A variety of analytical approaches can then be used to characterize the inherent dependencies and relations within the contributing activities within the considered process. Using examples drawn from our own research and the extant research literature, we describe the theoretical foundations of the process-based approach, relevant practical and pragmatic considerations for using such an approach, and a generic framework for applying this approach for evaluation studies in clinical settings. We also discuss the potential for this approach in future evaluations of interactive clinical systems, given the need for new approaches for evaluation, and significant opportunities for automated, unobtrusive data collection.     

Data mining and clinical data repositories: Insights from a 667,000 patient data set

Clinical repositories containing large amounts of biological, clinical, and administrative data are increasingly becoming available as health care systems integrate patient information for research and utilization objectives. To investigate the potential value of searching these databases for novel insights, we applied a new data mining approach, HealthMiner, to a large cohort of 667,000 inpatient and outpatient digital records from an academic medical system. HealthMiner approaches knowledge discovery using three unsupervised methods: CliniMiner, Predictive Analysis, and Pattern Discovery. The initial results from this study suggest that these approaches have the potential to expand research capabilities through identification of potentially novel clinical disease associations.     

Discrete-space versus continuous-space lesion boundary and area definitions

Measurement of the size of anatomic regions of interest in medical images is used to diagnose disease, track growth, and evaluate response to therapy. The discrete nature of medical images allows for both continuous and discrete definitions of region boundary. These definitions may, in turn, support several methods of area calculation that give substantially different quantitative values. This study investigated several boundary definitions (e.g., continuous polygon, internal discrete, and external discrete) and area calculation methods (pixel counting and Green's theorem). These methods were applied to three separate databases: A synthetic image database, the Lung Image Database Consortium database of lung nodules and a database of adrenal gland outlines. Average percent differences in area on the order of 20% were found among the different methods applied to the clinical databases. These results support the idea that inconsistent application of region boundary definition and area calculation may substantially impact measurement accuracy.     

Using a resequencing microarray as a multiple respiratory pathogen detection assay

Simultaneous testing for detection of infectious pathogens that cause similar symptoms (e.g., acute respiratory infections) is invaluable for patient treatment, outbreak prevention, and efficient use of antibiotic and antiviral agents. In addition, such testing may provide information regarding possible coinfections or induced secondary infections, such as virally induced bacterial infections. Furthermore, in many cases, detection of a pathogen requires more than genus/species-level resolution, since harmful agents (e.g., avian influenza virus) are grouped with other, relatively benign common agents, and for every pathogen, finer resolution is useful to allow tracking of the location and nature of mutations leading to strain variations. In this study, a previously developed resequencing microarray that has been demonstrated to have these capabilities was further developed to provide individual detection sensitivity ranging from 10(1) to 10(3) genomic copies for more than 26 respiratory pathogens while still retaining the ability to detect and differentiate between close genetic neighbors. In addition, the study demonstrated that this system allows unambiguous and reproducible sequence-based strain identification of the mixed pathogens. Successful proof-of-concept experiments using clinical specimens show that this approach is potentially very useful for both diagnostics and epidemic surveillance.     

Electronic signatures for long-lasting storage purposes in electronic archives

Communication and co-operation in healthcare and welfare require a certain set of trusted third party (TTP) services describing both status and relation of communicating principals as well as their corresponding keys and attributes. Additional TTP services are needed to provide trustworthy information about dynamic issues of communication and co-operation such as time and location of processes, workflow relations, and system behaviour. Legal and ethical requirements demand securely stored patient information and well-defined access rights. Among others, electronic signatures based on asymmetric cryptography are important means for securing the integrity of a message or file as well as for accountability purposes including non-repudiation of both origin and receipt. Electronic signatures along with certified time stamps or time signatures are especially important for electronic archives in general, electronic health records (EHR) in particular, and especially for typical purposes of long-lasting storage. Apart from technical storage problems (e.g. lifetime of the storage devices, interoperability of retrieval and presentation software), this paper identifies mechanisms of e.g. re-signing and re-stamping of data items, files, messages, sets of archived items or documents, archive structures, and even whole archives.     

Assertion modeling and its role in clinical phenotype identification

This paper describes an approach to assertion classification and an empirical study on the impact this task has on phenotype identification, a real world application in the clinical domain. The task of assertion classification is to assign to each medical concept mentioned in a clinical report (e.g., pneumonia, chest pain) a specific assertion category (e.g., present, absent, and possible). To improve the classification of medical assertions, we propose several new features that capture the semantic properties of special cue words highly indicative of a specific assertion category. The results obtained outperform the current state-of-the-art results for this task. Furthermore, we confirm the intuition that assertion classification contributes in significantly improving the results of phenotype identification from free-text clinical records.     

Citizen participation in the police domain: The role of citizens’ attitude and morality

Although there is a large potential of citizen capital in fighting crime and creating safer neighborhoods, in reality, only a small fraction of citizens is actively participating. This study examines the relationship between different types of actual participation behavior in the police domain from a citizen's stance and 3 different but interconnected psychological drivers: the attitude toward citizen participation, moral values, and moral emotions. A total of 217 Dutch citizens filled out an online questionnaire, assessing these drivers and the actual participatory actions they engaged in over the past year. The results show that 4 broad categories of participation behavior can be distinguished: social control (e.g., correcting others regarding their behavior); responsive participation (e.g., calling the police); collaborative participation (e.g., meeting with a police officer); and detection (e.g., joining a neighborhood watch). As expected, moral values had an indirect influence on participation via attitude and moral emotions. The attitude toward citizen participation was positively related to all four types of reported behavior, while the influence of moral emotions only related to social control and responsive behavior. These results can be used in the design and testing of interventions to stimulate citizen participation.     

De-identification of unstructured paper-based health records for privacy-preserving secondary use

Whenever personal data is processed, privacy is a serious issue. Especially in the document-centric e-health area, the patients’ privacy must be preserved in order to prevent any negative repercussions for the patient. Clinical research, for example, demands structured health records to carry out efficient clinical trials, whereas legislation (e.g. HIPAA) regulates that only de-identified health records may be used for research. However, unstructured and often paper-based data dominates information technology, especially in the healthcare sector. Existing approaches are geared towards data in English-language documents only and have not been designed to handle the recognition of erroneous personal data which is the result of the OCR-based digitization of paper-based health records.     

Electronic search strategies to identify reports of cluster randomized trials in MEDLINE: low precision will improve with adherence to reporting standards

Background  

Cluster randomized trials (CRTs) present unique methodological and ethical challenges. Researchers conducting systematic reviews of CRTs (e.g., addressing methodological or ethical issues) require efficient electronic search strategies (filters or hedges) to identify trials in electronic databases such as MEDLINE. According to the CONSORT statement extension to CRTs, the clustered design should be clearly identified in titles or abstracts; however, variability in terminology may make electronic identification challenging. Our objectives were to (a) evaluate sensitivity ("recall") and precision of a well-known electronic search strategy ("randomized controlled trial" as publication type) with respect to identifying CRTs, (b) evaluate the feasibility of new search strategies targeted specifically at CRTs, and (c) determine whether CRTs are appropriately identified in titles or abstracts of reports and whether there has been improvement over time.