Linking Patient Records at Scale with a Hybrid Approach Combining Contrastive Learning and Deterministic Rules

ABSTRACTIntegrating data from multiple sources expands research opportunities at low cost. However, due to different data collection processes and privacy constraints, unique identifiers are unavailable. Record linkage (RL) algorithms address this by probabilistically linking records based on partially identifying variables. Since these variables lack the strength to perfectly combine information, RL procedures yield an imperfect set of linked records. Therefore, assessing the false discovery proportion (FDP) in RL is crucial for ensuring the reliability of subsequent analyses. In this paper, we introduce a novel method for estimating the FDP in RL for two overlapping data sets. We synthesize data from their estimated empirical distribution and use it along with real data in the linkage process. Since synthetic records cannot form links with real entities, they provide a means to estimate the amount of falsely linked pairs. Notably, this method applies to all RL techniques and across diverse settings where links and non‐links have similar distributions–typical in complex tasks with poorly discriminative linking variables and multiple records sharing similar information while representing different entities. By identifying the FDP in RL and selecting suitable model parameters, our approach enables to assess and improve the reliability of linked data. We evaluate its performance using established RL algorithms and benchmark data applications before deploying it to link siblings from the Netherlands Perinatal Registry, where the reliability of previous RL applications has never been confirmed. Through this application, we highlight the importance of accounting for linkage errors when studying mother‐child dynamics in healthcare records.

In this paper, a probabilistic analytical approach for reliability evaluation of power systems with high penetration of wind and solar photovoltaic (PV) renewable power generation is presented. Due to intermittent nature of wind and solar power, the traditional deterministic method cannot properly address such uncertainties, probabilistic methods need to be utilized. In this paper, the loss of load method, one of the most effective probabilistic analytical methods, is adopted. The generation model is represented by the capacity outage probability table (COPT). The load model is presented by the load duration curve (LDC). Both generation and load models are used to obtain the system reliability considering wind and PV sources with certain forced outage rate (FOR). Simulation results are obtained using MATLAB. It is found that renewable energy sources can significantly improve the system reliability, but not as good as conventional power generators with the same rated capacity due to the actual reduced capacity value of renewable energy sources caused by intermittency. This analytical analysis is useful for the power system planner to quantify reliability improvements by installing grid-connected hybrid renewable power generation.

Integrated Data System Person Identification: Accuracy Requirements and Methods

Examine the accuracy of privacy preserving record linkage (PPRL) matches in real world data (RWD). We conducted a systematic literature review to identify articles evaluating PPRL methods from January 1, 2013 to June 15, 2023. Eligible studies included original research reporting quantitative metrics such as precision and recall in health-related data sources. Covidence software was used to manage the review process. Five studies met our inclusion criteria. Tokenization and hash functions were used to hash and encrypt personally identifiable information (PII) including first and last names, dates of birth (DOB), and Social Security Numbers (SSNs) in a variety of RWD. All identified studies utilized deterministic matching. Combinations of tokenized or hashed PII that included "quasi-identifiers" like names and DOBs had consistently high precision (>95%) but lower recall, likely due to misspelled or inconsistently spelled names and name changes. SSN-based combinations demonstrated high precision but variable recall due to incomplete SSN data in RWD. Studies that employed algorithms in which at least one match was identified from a specified set of PII combinations provided high precision and high recall. The systematic review indicates that PPRL methods generally provide highly accurate patient data linkage while maintaining privacy. Researchers should carefully consider the completeness and stability of each PII element selected for PPRL and may want to employ a strategy that allows for patient records to be matched if they meet at least one of several combinations of PII.

Slope stability analyses are usually carried out using deterministic methods, which aim the calculation of a single safety factor assuming the values of the shear strength parameters as representative and fixed. These methods fail to assess the uncertainties in soil properties and do not indicate the proportion of influence that each resistance parameter has on the final value of the safety factor. The probabilistic methods, based on probability, reliability and statistical theories, allow the estimation of the influence of these uncertainties on the deterministic calculations, with the possibility to broadly predict the risk of failure associated with a geotechnical slope stability project. This dissertation studies the application of three probabilistic methods (Monte Carlo, Latin Hypercube, and Alternative Point Estimates) in the evaluation of slope stability, with aid of limit equilibrium methods for the calculation of safety factors. In order to infer the impact of random variables on the estimates of probability and reliability, as well as the importance of an adequate quantification of the standard deviation values, comparisons are made among the results obtained with probabilistic and deterministic methods (limit equilibrium method, finite element method), discussing the main advantages, difficulties and limitations in their application to soil slope stability problems.

Probabilistic and deterministic methods for reserves and resources evaluation are commonly used in isolation and often considered mutually exclusive. Subsurface uncertainties are critical factors impacting projects and reserves/resources especially in projects/areas where large sums of capital investment are required. Probabilistic methods allow a rigorous use of information on the ranges of uncertainty on key reservoir parameters like porosity, water saturation, permeability aquifer size for reserves estimation. A key output of probabilistic methods is the confidence levels associated with the reserves. Deterministic methods cannot provide confidence levels associated with reserves and resources assessments reason why its successful application is often relaying on the expert knowledge of the evaluator and the strict use of reserves or resources definitions. Technological advances in computing in the last decades have played a key role in advancing computationally intensive probabilistic methodologies including artificial intelligence. These advances have allowed integrated teams to perform studies using sophisticated workflows in feasible timeframes. In this paper a detailed analysis of the different probabilistic methods is presented with a review of the level of regulatory compliance with the SEC1 and other industry guidelines like SPE PRMS2 that each of the different methods have. A group of workflows using a combination of these methods is also analysed.

With free movement of labour in Europe, European guidelines on cardiovascular care and the enlargement of the European Union to include countries with disparate health care systems, it is important to develop common quality standards for cardiovascular prevention and risk management across Europe. Panels from nine European countries (Austria, Belgium, Finland, France, Germany, Netherlands, Slovenia, United Kingdom and Switzerland) developed quality indicators for the prevention and management of cardiovascular disease in primary care. A two-stage modified Delphi process was used to identify indicators that were judged valid for necessary care. Forty-four out of 202 indicators (22%) were rated as valid. These focused predominantly on secondary prevention and management of established cardiovascular disease and diabetes. Less agreement on indicators of preventive care or on indicators for the management of hypertension and hypercholesterolaemia in patients without established disease was observed. Although 85% of the 202 potential indicators assessed were rated valid by at least one panel, lack of consensus among panels meant that the set that could be agreed upon among all panels was much smaller. Indicators for the management of established cardiovascular disease have been developed, which can be used to measure the quality of cardiovascular care across a wide range of countries. Less agreement on how the quality of preventive care should be assessed was observed, probably caused by differences in health systems, culture and attitudes to prevention.

Aim: To analyse trends in organ transplantation in South Africa, a developing country with limited government support, two disparate health care systems and issues related to organ trafficking and HIV. Methods: Transplantation statistics voluntarily submitted to the Organ Donor Foundation from 1991 to 2011 were analysed in a relational database. Results: A total of 7280 transplants were performed across 10 public and 13 private hospitals with the majority being kidney transplants (84%).Figure: No Caption available.Deceased donors decreased from 4.2 per million population (pmp) in 1991 to 1.7 pmp in 2011. An increase in living donors kept transplantation activity relatively constant. A marked shift towards private sector transplantation was noted (most marked in liver and cardiac transplantation).Table 1: Comparison of workload between public and private health care systems: 1991, 2001 and 2011Lung and multi-organ transplants were almost exclusively performed in the private sector (141 vs 5). Liver transplantation in the private sector began in 2004 but increased dramatically and has expanded into the public sector.Table 2: Liver transplantation: 2004 - 2011The majority of transplants were performed in the Gauteng (46%) and Western Cape (42%) provinces. Kwa-Zulu Natal performed 20% of South Africa’s transplants in 2002 prior to prosecutions in the private sector for organ trafficking. In 2010 <1% of South Africa’s transplants were performed in Kwa-Zulu Natal. Conclusions: Public and private health care can work together to offer transplantation in the resource constrained developing world. Government oversight and funding is essential to ensure adequate regulation and equitable access to transplantation.

An American Society of Mechanical Engineers (ASME) standard for design of facilities with potential for significant hazards to the health and safety of the public, the worker, and the environment is being developed. The standard includes requirements and guidance for design organizations to incorporate risk informed probabilistic design methodologies with traditional deterministic design methods using reliability and availability targets. Probabilistic design processes include selecting the systems and products to be evaluated, defining the failure modes, establishing system life cycle models, defining the acceptance criteria, selecting the appropriate probabilistic design method, processing the analysis results, and, where applicable, recommending design changes. The order of these activities as well as the inputs and outputs from each step are defined. Probabilistic design methods are addressed at both the system and product level and their interaction. A toolbox of probabilistic design methods is included as an appendix to the standard. A description of the current plans for each activity, the order of activities, and the input and outputs from each activity are described along with the interaction to the other activities in the plant system design process. The basic content and organization of the toolbox is described. This paper provides an overview of the probabilistic analysis section rather than a detailed description of the draft document or examples.

To enhance interoperability among disparate healthcare systems, standardized terminologies and transmission protocols are in the process of initiation across Korean medical institutions. A survey was conducted among 182 employees of healthcare facilities including doctors, nurses, administrative staff, and public health workers. This study investigates healthcare employee' perceptions and acceptance of implementing healthcare data standards, using the Unified Theory of Acceptance and Use of Technology (UTAUT) framework. In PLS-SEM analysis, several significant relationships between behavioral intention and other factors were identified.

Comparison of glioblastoma outcomes in two geographically and ethnically distinct patient populations in disparate healthcare systems

Blockchain technology, renowned for its decentralized and immutable nature, is poised to revolutionize the landscape of smart healthcare. In this transformative paradigm, patient data becomes securely stored and efficiently managed, fostering unprecedented levels of trust, transparency, and interoperability among healthcare stakeholders. By leveraging blockchain, institutions can ensure the integrity and confidentiality of medical records, mitigating the risk of data breaches. Smart contracts embedded within the blockchain facilitate automated execution of agreements, streamlining processes such as insurance claims, medication adherence monitoring, and patient consent management. Moreover, it enables the creation of comprehensive patient-centric networks, facilitating seamless data sharing across disparate healthcare systems while maintaining patient privacy. This technology holds immense promise for reforming smart healthcare, fostering a future where data is securely shared, patient outcomes are optimized, and the delivery of healthcare services is truly patient-centric.

Stability analyses of geotechnical structures in rock are traditionally performed using deterministic methods. In Europe, Eurocode7, introduced in the beginning of the 21st century, adopts limit state design and semi-probabilistic methods, using partial factors for the design of geotechnical structures. Meanwhile, reliability-based design, using probabilistic methods, is becoming more common in practical cases. The paper considers an intentionally simple case study—the analysis of a slope in a rock mass with one discontinuity, considered in a discrete way, forming a rock block to be stabilised by anchors—to compare the results obtained with the different methods. The objective is to calculate the force applied by the anchors so that the ultimate limit states of sliding of the rock block is verified. Deterministic-based design optimization considering both the traditional global safety factor approach and the partial factor approach following the Eurocode 7 are first applied. A reliability-based design optimization procedure—which takes geometrical and mechanical properties of the discontinuities as random variables—is then used, and the results are compared to the former ones. A discussion is presented concerning the consistency of the obtained results.

The rapid growth of the Internet has been accompanied by a proliferation of e-services targeting consumers. E-services are available for banking, shopping, learning, government online, and healthcare. However, each of these services requires a consumer’s personally identifiable information (PII) in one form or another. This leads to concerns over privacy. In order for e-services to be successful, privacy must be protected (Ackerman, Cranor, &amp; Reagle, 1999). An effective and flexible way of handling privacy is management via privacy policies. In this approach, a consumer of an e-service has a personal privacy policy that describes what private information the consumer is willing to give up to the e-service, with which parties the provider of the e-service may share the private information, and how long the private information may be kept by the provider. The provider likewise has a provider privacy policy describing similar privacy constraints as in the consumer’s policy, but from the viewpoint of the provider, (i.e., the nature of the private information and the disclosure/retention requirements that are needed by the e-service). Before the consumer engages the e-service, the provider’s privacy policy must match with the consumer’s privacy policy. In this way, the consumer’s privacy is protected, assuming that the provider complies with the consumer’s privacy policy. Note that policy compliance is outside the scope of this work but see Yee and Korba (July, 2004). Initial attempts at conserving consumer privacy for e-services over the last few years have focused on the use of Web site privacy policies that state the privacy rules or preferences of the Web site or service provider. Some of these policies are merely statements in plain English and it is up to the consumer to read it. This has the drawback that very few consumers take the trouble to read it. Even when they do take the time to look at it, online privacy policies have been far too complicated for consumers to understand and suffer from other deficiencies (Lichtenstein, Swatman, &amp; Babu, 2003; Jensen &amp; Potts, 2004). Still other privacy policies are specified using P3P (W3C) that allows a consumer’s browser to automatically check the privacy policy via a browser plug-in. This, of course, is better than plain English policies but a major drawback is that it is a “take-it-or-leave-it” approach. There is no recourse for the consumer who has a conflict with the Web site’s P3P policy, except to try another Web site. In this case, we have advocated a negotiations approach to resolve the conflict (Yee &amp; Korba, Jan., May, 2003). However, this requires a machine-processable personal privacy policy for the consumer. We assume that providers in general have sufficient resources to generate their privacy policies. Certainly, the literature is full of works relating to enterprise privacy policies and models (e.g., Barth &amp; Mitchell, 2005; Karjoth &amp; Schunter 2002). Consumers, on the other hand, need help in formulating machine-processable privacy policies. In addition, the creation of such policies needs to be as easy as possible or consumers would simply avoid using them. Existing privacy specification languages such as P3P, APPEL (W3C; W3C, 2002), and EPAL (IBM) are far too complicated for the average internet user to understand. Understanding or changing a privacy policy expressed in these languages effectively requires knowing how to program. Moreover, most of these languages suffer from inadequate expressiveness (Stufflebeam, Anton, He, &amp; Jain, 2004). What is needed is an easy, semi-automated way of seeding a personal privacy policy with a consumer’s privacy preferences. In this work, we present two semi-automated approaches for obtaining consumer personal privacy policies for e-services through seeding. This article is based on our work in Yee and Korba (2004). The section “Background” examines related work and the content of personal privacy policies. The section “Semi-Automated Seeding of Personal Privacy Policies” shows how personal privacy policies can be semi-automatically seeded or generated. The section “Future Trends” identifies some of the developments we see in this area over the next few years. We end with ”Conclusion”.

Quantitative diffusion tensor MRI fiber tractography of sensorimotor white matter development in premature infants