Evaluation of patient re-identification using laboratory test orders and mitigation via latent space variables.

Anonymized electronic health records (EHR) are often used for biomedical research. One persistent concern with this type of research is the risk for re-identification of patients from their purportedly anonymized data. Here, we use the EHR of 731,850 de-identified patients to demonstrate that the average patient is unique from all others 98.4% of the time simply by examining what laboratory tests have been ordered for them. By the time a patient has visited the hospital on two separate days, they are unique in 72.3% of cases. We further present a computational study to identify how accurately the records from a single day of care can be used to re-identify patients from a set of 99 other patients. We show that, given a single visit's laboratory orders (even without result values) for a patient, we can re-identify the patient at least 25% of the time. Furthermore, we can place this patient among the top 10 most similar patients 47% of the time. Finally, we present a proof-of-concept technique using a variational autoencoder to encode laboratory results into a lower-dimensional latent space. We demonstrate that releasing latentspace encoded laboratory orders significantly improves privacy compared to releasing raw laboratory orders (<5% re-identification), while preserving information contained within the laboratory orders (AUC of >0.9 for recreating encoded values). Our findings have potential consequences for the public release of anonymized laboratory tests to the biomedical research community. We note that our findings do not imply that laboratory tests alone are personally identifiable. In the attack scenario presented here, reidentification would require a threat actor to possess an external source of laboratory values which are linked to personal identifiers at the start.

No exchange, same pain, no gain: Risk-reward of wearable healthcare disclosure of health personally identifiable information for enhanced pain treatment.

Wearable technologies have created fascinating opportunities for patients to treat chronic pain in a discreet, mobile fashion. However, many of these health wearables require patients to disclose sensitive information, including health information (e.g., heart rate, glucose levels) and personal information (location, email, name, etc.). Individuals using wearables for treatment of chronic pain may sacrifice social health elements, including their privacy, in exchange for better physical and mental health. Utilizing communication privacy management, a popular disclosure theory, this article explores the policy and ethical ramifications of patients disclosing sensitive health information in exchange for better health treatment and relief of chronic pain. The article identifies scenarios where a user must disclose information, and what factors motivate or dissuade disclosure, and ultimately the use of a health wearable. Practical implications of this conceptual article include an improved understanding of how and why consumers may disclose personal data to health wearables, and potential impacts for public policy and ethics regarding how wearables and their manufacturers entice disclosure of private health information.

Estudio comparativo de la diversidad de las características clínicas dentales en poblaciones militares de Portugal y España / Diversity in dental clinical characteristics in Portuguese and Spanish military populations

Introducción: Se ha realizado una comparación de las características dentales de 2poblaciones militares de Portugal y España. El objetivo principal de esta investigación fue identificar aquellas características dentales que podrían ser de utilidad para diferenciar estas poblaciones en un análisis forense. Material y métodos: El estudio se realizó en una muestra compuesta por 5.136 militares profesionales de las fuerzas armadas, el 31,9% eran militares portugueses y el 68,1% del total de la muestra pertenecían a las fuerzas armadas españolas. Los datos dentales se registraron empleando los símbolos dentales descritos en Forensic Dental Symbols(R), gestionados con la base de datos Dental Encoder(R). Resultados: La población de estudio estaba constituida por un 86,6% de hombres (88,1% en la muestra española y 83,4% en la muestra portuguesa) y un 13,4% de mujeres (11,9% en la muestra española y 16,6% en la muestra portuguesa). La frecuencia de dientes no restaurados fue menor para los primeros molares en todos los cuadrantes, mientras que la mayor frecuencia de esta característica (>90%) se observó en los dientes anteriores, superiores e inferiores, y en los primeros premolares inferiores. Las frecuencias más altas de tratamientos restauradores fueron encontradas para los primeros y segundos molares en todos los cuadrantes, y las mayores frecuencias de ausencias dentarias se observaron en los terceros molares (superior al 28% en todos los cuadrantes). El análisis de concordancia mostró que las correlaciones entre los dientes contralaterales fueron significativamente mayores que entre los dientes antagonistas, para ambas muestras poblacionales de estudio. Conclusiones: Nuestros resultados proporcionan información potencialmente útil sobre la importancia de las bases de datos de registros dentales y el análisis de las características dentales con fines de identificación

Introduction: Dental characteristics were compared in population samples of Spanish and Portuguese military personnel. The main aim of this study was to identify those dental characteristics that could potentially serve to differentiate between these populations in a forensic analysis. Material and methods: A sample of 5136 individuals belonging to the professional military staff of the Portuguese and Spanish armed forces was studied. Dental data were recorded with the Forensic Dental Symbols(R) for the Dental Encoder(R) database. The population sample analysed in this study consisted of 68.1% Spanish and 31.9% Portuguese individuals. Results: The population was mostly male, with 86.6% men (88.1% in the Spanish sample versus 83.4% in the Portuguese sample), and 13.4% women (11.9% Spanish and 16.6% Portuguese). The frequency of unrestored teeth was lowest for first molars in all quadrants, and the highest frequency of unrestored teeth (>90%) was for the upper and lower anterior teeth and lower first premolars. The highest frequencies of restorative treatment were found for the first and second molars in all quadrants, and the highest frequencies of missing teeth were found for the third molars (always >28%). Concordance analysis showed that correlations between contralateral teeth were significantly higher than between antagonist teeth in both samples. Conclusions: Our findings provide potentially useful information on the importance of dental record databases and their value for identification purposes

Efficient Record Linkage Algorithms Using Complete Linkage Clustering.

Data from different agencies share data of the same individuals. Linking these datasets to identify all the records belonging to the same individuals is a crucial and challenging problem, especially given the large volumes of data. A large number of available algorithms for record linkage are prone to either time inefficiency or low-accuracy in finding matches and non-matches among the records. In this paper we propose efficient as well as reliable sequential and parallel algorithms for the record linkage problem employing hierarchical clustering methods. We employ complete linkage hierarchical clustering algorithms to address this problem. In addition to hierarchical clustering, we also use two other techniques: elimination of duplicate records and blocking. Our algorithms use sorting as a sub-routine to identify identical copies of records. We have tested our algorithms on datasets with millions of synthetic records. Experimental results show that our algorithms achieve nearly 100% accuracy. Parallel implementations achieve almost linear speedups. Time complexities of these algorithms do not exceed those of previous best-known algorithms. Our proposed algorithms outperform previous best-known algorithms in terms of accuracy consuming reasonable run times.

Anonymisation of geographical distance matrices via Lipschitz embedding.

BACKGROUND: Anonymisation of spatially referenced data has received increasing attention in recent years. Whereas the research focus has been on the anonymisation of point locations, the disclosure risk arising from the publishing of inter-point distances and corresponding anonymisation methods have not been studied systematically. METHODS: We propose a new anonymisation method for the release of geographical distances between records of a microdata file--for example patients in a medical database. We discuss a data release scheme in which microdata without coordinates and an additional distance matrix between the corresponding rows of the microdata set are released. In contrast to most other approaches this method preserves small distances better than larger distances. The distances are modified by a variant of Lipschitz embedding. RESULTS: The effects of the embedding parameters on the risk of data disclosure are evaluated by linkage experiments using simulated data. The results indicate small disclosure risks for appropriate embedding parameters. CONCLUSION: The proposed method is useful if published distance information might be misused for the re-identification of records. The method can be used for publishing scientific-use-files and as an additional tool for record-linkage studies.