The EU-ADR Web Platform: delivering advanced pharmacovigilance tools

Navigating duplication in pharmacovigilance databases: a scoping review

OBJECTIVES

Pharmacovigilance databases play a critical role in monitoring drug safety. The duplication of reports in pharmacovigilance databases, however, undermines their data integrity. This scoping review sought to provide a comprehensive understanding of duplication in pharmacovigilance databases worldwide.

DESIGN

A scoping review.

DATA SOURCES

Reviewers comprehensively searched the literature in PubMed, Web of Science, Wiley Online Library, EBSCOhost, Google Scholar and other relevant websites.

ELIGIBILITY CRITERIA

Peer-reviewed publications and grey literature, without language restriction, describing duplication and/or methods relevant to duplication in pharmacovigilance databases from inception to 1 September 2023.

DATA EXTRACTION AND SYNTHESIS

We used the Joanna Briggs Institute guidelines for scoping reviews and conformed with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews. Two reviewers independently screened titles, abstracts and full texts. One reviewer extracted the data and performed descriptive analysis, which the second reviewer assessed. Disagreements were resolved by discussion and consensus or in consultation with a third reviewer.

RESULTS

We screened 22 745 unique titles and 156 were eligible for full-text review. Of the 156 titles, 58 (47 peer-reviewed; 11 grey literature) fulfilled the inclusion criteria for the scoping review. Included titles addressed the extent (5 papers), prevention strategies (15 papers), causes (32 papers), detection methods (25 papers), management strategies (24 papers) and implications (14 papers) of duplication in pharmacovigilance databases. The papers overlapped, discussing more than one field. Advances in artificial intelligence, particularly natural language processing, hold promise in enhancing the efficiency and precision of deduplication of large and complex pharmacovigilance databases.

CONCLUSION

Duplication in pharmacovigilance databases compromises risk assessment and decision-making, potentially threatening patient safety. Therefore, efficient duplicate prevention, detection and management are essential for more reliable pharmacovigilance data. To minimise duplication, consistent use of worldwide unique identifiers as the key case identifiers is recommended alongside recent advances in artificial intelligence.

Do peripheral neuropathies differ among immune checkpoint inhibitors? Reports from the European post-marketing surveillance database in the past 10 years

Although the immunotherapy advent has revolutionized cancer treatment, it, unfortunately, does not spare cancer patients from possible immune-related adverse events (irAEs), which can also involve the peripheral nervous system. Immune checkpoint inhibitors (ICIs), blocking cytotoxic T-lymphocyteassociated protein 4 (CTLA-4), programmed cell death protein 1 (PD-1), or programmed cell death ligand 1 (PD-L1), can induce an immune imbalance and cause different peripheral neuropathies (PNs). Considering the wide range of PNs and their high impact on the safety and quality of life for cancer patients and the availability of large post-marketing surveillance databases, we chose to analyze the characteristics of ICI-related PNs reported as suspected drug reactions from 2010 to 2020 in the European real-world context. We analyzed data collected in the European pharmacovigilance database, Eudravigilance, and conducted a systematic and disproportionality analysis. In our study, we found 735 reports describing 766 PNs occurred in patients treated with ICIs. These PNs included Guillain-Barré syndrome, Miller-Fisher syndrome, neuritis, and chronic inflammatory demyelinating polyradiculoneuropathy. These ADRs were often serious, resulting in patient disability or hospitalization. Moreover, our disproportionality analysis revealed an increased reporting frequency of PNs with tezolizumab compared to other ICIs. Guillain-Barré syndrome is a notable potential PN related to ICIs, as it is associated with a significant impact on patient safety and has had unfavorable outcomes, including a fatal one. Continued monitoring of the safety profile of ICIs in real-life settings is necessary, especially considering the increased frequency of PNs associated with atezolizumab compared with other ICIs.

Reporting of clinical trial safety results in ClinicalTrials.gov for FDA-approved drugs: A cross-sectional analysis

BACKGROUND

Adverse events identified during clinical trials can be important early indicators of drug safety, but complete and timely data on safety results have historically been difficult to access. The aim was to compare the availability, completeness, and concordance of safety results reported in ClinicalTrials.gov and peer-reviewed publications.

METHODS

We analyzed clinical trials used in the Food and Drug Administration safety assessment of new drugs approved between 1 July 2018 and 30 June 2019. The key safety outcomes examined were all-cause mortality, serious adverse events, adverse events, and withdrawals due to adverse events. Availability of safety results was measured by the presence and timing of a record of trial-level results in ClinicalTrials.gov and a corresponding peer-reviewed publication. For the subset of trials with available results, completeness was defined as the reporting of safety results for all participants and compared between ClinicalTrials.gov and publications. To assess concordance, we compared the numeric results for safety outcomes reported in ClinicalTrials.gov and publications to results in Food and Drug Administration trial reports.

RESULTS

Among 156 trials studying 52 drugs, 91 (58.3%) trials reported safety results in ClinicalTrials.gov and 106 (67.9%) in peer-reviewed publications (risk difference = -9.6%, 95% confidence interval = -20.3 to 1.0). All-cause mortality was reported sooner in published articles compared with ClinicalTrials.gov (log-rank test, p = 0.01). There was no difference in time to reporting for serious adverse events (p = 0.05), adverse events (p = 0.09), or withdrawals due to adverse events (p = 0.20). Complete reporting of all-cause mortality was similar in ClinicalTrials.gov and publications (74.7% vs 78.3%, respectively; risk difference = -3.6%, 95% confidence interval = -15.5 to 8.3) and higher in ClinicalTrials.gov for serious adverse events (100% vs 79.2%; risk difference = 20.8%, 95% confidence interval = 13.0 to 28.5) and adverse events (100% vs 86.8%; risk difference = 13.2%, 95% confidence interval = 6.8 to 19.7). Withdrawals due to adverse events were less often completely reported in ClinicalTrials.gov (62.6% vs 92.5%; risk difference = -29.8%, 95% confidence interval = -40.1 to -18.7). No difference was found in concordance of results between ClinicalTrials.gov and publications for all-cause mortality, serious adverse events, or withdrawals due to adverse events.

CONCLUSION

Safety results were available in ClinicalTrials.gov at a similar rate as in peer-reviewed publications, with more complete reporting of certain safety outcomes in ClinicalTrials.gov. Future efforts should consider adverse event reporting in ClinicalTrials.gov as an accessible data source for post-marketing surveillance and other evidence synthesis tasks.

Standardizing registry data to the OMOP Common Data Model: experience from three pulmonary hypertension databases

Background

The Observational Medical Outcomes Partnership (OMOP) Common Data Model (CDM) can be used to transform observational health data to a common format. CDM transformation allows for analysis across disparate databases for the generation of new, real-word evidence, which is especially important in rare disease where data are limited. Pulmonary hypertension (PH) is a progressive, life-threatening disease, with rare subgroups such as pulmonary arterial hypertension (PAH), for which generating real-world evidence is challenging. Our objective is to document the process and outcomes of transforming registry data in PH to the OMOP CDM, and highlight challenges and our potential solutions.

Methods

Three observational studies were transformed from the Clinical Data Interchange Standards Consortium study data tabulation model (SDTM) to OMOP CDM format. OPUS was a prospective, multi-centre registry (2014–2020) and OrPHeUS was a retrospective, multi-centre chart review (2013–2017); both enrolled patients newly treated with macitentan in the US. EXPOSURE is a prospective, multi-centre cohort study (2017–ongoing) of patients newly treated with selexipag or any PAH-specific therapy in Europe and Canada. OMOP CDM version 5.3.1 with recent OMOP CDM vocabulary was used. Imputation rules were defined and applied for missing dates to avoid exclusion of data. Custom target concepts were introduced when existing concepts did not provide sufficient granularity.

Results

Of the 6622 patients in the three registry studies, records were mapped for 6457. Custom target concepts were introduced for PAH subgroups (by combining SNOMED concepts or creating custom concepts) and World Health Organization functional class. Per the OMOP CDM convention, records about the absence of an event, or the lack of information, were not mapped. Excluding these non-event records, 4% (OPUS), 2% (OrPHeUS) and 1% (EXPOSURE) of records were not mapped.

Conclusions

SDTM data from three registries were transformed to the OMOP CDM with limited exclusion of data and deviation from the SDTM database content. Future researchers can apply our strategy and methods in different disease areas, with tailoring as necessary. Mapping registry data to the OMOP CDM facilitates more efficient collaborations between researchers and establishment of federated data networks, which is an unmet need in rare diseases.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01434-3.

Machine learning guided association of adverse drug reactions with in vitro target-based pharmacology

BACKGROUND

Adverse drug reactions (ADRs) are one of the leading causes of morbidity and mortality in health care. Understanding which drug targets are linked to ADRs can lead to the development of safer medicines.

METHODS

Here, we analyse in vitro secondary pharmacology of common (off) targets for 2134 marketed drugs. To associate these drugs with human ADRs, we utilized FDA Adverse Event Reports and developed random forest models that predict ADR occurrences from in vitro pharmacological profiles.

FINDINGS

By evaluating Gini importance scores of model features, we identify 221 target-ADR associations, which co-occur in PubMed abstracts to a greater extent than expected by chance. Amongst these are established relations, such as the association of in vitro hERG binding with cardiac arrhythmias, which further validate our machine learning approach. Evidence on bile acid metabolism supports our identification of associations between the Bile Salt Export Pump and renal, thyroid, lipid metabolism, respiratory tract and central nervous system disorders. Unexpectedly, our model suggests PDE3 is associated with 40 ADRs.

INTERPRETATION

These associations provide a comprehensive resource to support drug development and human biology studies.

FUNDING

This study was not supported by any formal funding bodies.

Streamlined mail-based methods for large randomised trials: lessons learnt from the ASCEND study

Reliable assessment of the effects of an intervention usually requires large randomised trials but such studies are becoming increasingly complex and costly to run. 'Streamlined' trials are needed in which every aspect of the trial design and conduct is simplified, retaining only those elements needed to answer the research question and ensure the safety of the individual participants. In this review we discuss how the trial 'A Study of Cardiovascular Events iN Diabetes' (ASCEND) was streamlined. The study included a two-by-two factorial design: it assessed the effects of low-dose aspirin and, separately, supplementation with n-3 fatty acids on serious vascular events in 15,480 people with diabetes but no overt cardiovascular disease. Other key streamlined design features, such as mail-based recruitment and follow-up, mainly by post, with no in-person visits and use of a run-in period, are also described. We go on to discuss the success of the study and other studies that have employed a similar mail-based approach, and the type of clinical trials that are suitable for mail-based design. Finally, we consider the limitations of the study, and how these could be circumvented in future studies. ASCEND randomised large numbers of eligible participants, achieved good adherence rates and almost complete follow-up at a fraction of the cost of traditional clinic-based trials. Such studies are necessary if researchers are to address the important clinical questions most relevant to improving health.

Analysis of Adverse Drug Reactions Identified in Nursing Notes Using Reinforcement Learning

Objectives

Electronic Health Records (EHRs)-based surveillance systems are being actively developed for detecting adverse drug reactions (ADRs), but this is being hindered by the difficulty of extracting data from unstructured records. This study performed the analysis of ADRs from nursing notes for drug safety surveillance using the temporal difference method in reinforcement learning (TD learning).

Methods

Nursing notes of 8,316 patients (4,158 ADR and 4,158 non-ADR cases) admitted to Ajou University Hospital were used for the ADR classification task. A TD(λ) model was used to estimate state values for indicating the ADR risk. For the TD learning, each nursing phrase was encoded into one of seven states, and the state values estimated during training were employed for the subsequent testing phase. We applied logistic regression to the state values from the TD(λ) model for the classification task.

Results

The overall accuracy of TD-based logistic regression of 0.63 was comparable to that of two machine-learning methods (0.64 for a naïve Bayes classifier and 0.63 for a support vector machine), while it outperformed two deep learning-based methods (0.58 for a text convolutional neural network and 0.61 for a long short-term memory neural network). Most importantly, it was found that the TD-based method can estimate state values according to the context of nursing phrases.

Conclusions

TD learning is a promising approach because it can exploit contextual, time-dependent aspects of the available data and provide an analysis of the severity of ADRs in a fully incremental manner.

Deconstructing the Pharmacovigilance Hype Cycle

Data science is making increasing contributions to pharmacovigilance. Although the technical innovation of these works are indisputable, efficient progress in real-world pharmacovigilance signal detection may be hampered by corresponding technology life cycle effects, with a resulting tendency to conclude that, with large enough datasets and intricate algorithms, "the numbers speak for themselves," discounting the importance of clinical and scientific judgment. A practical consequence is overzealous declarations regarding the safety or lack of safety of drugs. We describe these concerns through a critical discussion of key results and conclusions from case studies selected to illustrate these points.

Pharmacovigilance in oncology

Background Side effects of cancer therapy are one of the most important issues faced by cancer patients during their illness. Pharmacovigilance, namely the science and activities aimed at monitoring the safety of drugs, is particularly important in oncology, due to the intrinsic biologic toxicity of antineoplastic agents, their narrow therapeutic windows, and the high doses and rigid timing of treatment regimens. Aim of the review To identify the main issues in carrying out an effective pharmacovigilance activity in oncology. Method We searched PubMed for articles about pharmacovigilance in relation to chemotherapy, radiotherapy and targeted therapy for cancer, using MeSH terms and text words. We also searched Embase, CINAHL, Scopus, Micromedex, the Cochrane Library, two pharmacovigilance databases and the gray literature for articles published in 2012-2018. Overall, 137 articles were considered potentially relevant and were critically appraised independently by two authors, leading to the inclusion of 44 relevant studies, guidelines and reviews. Another 10 important research reports were included in the review. Results Eight critical issues of pharmacovigilance in oncology were identified. These issues pertain to: terminology; range of side effects; targeted therapy and immunotherapy; chemoradiotherapy; generic drugs and biosimilars; drug interactions, pharmacogenetics and polypharmacy; special patient categories; and under-reporting of ADRs. Conclusion The importance of pharmacovigilance in oncology must be highlighted with every effort, to improve safety and offer cancer patients every possible help to improve their quality of life during such a critical period of their lives.

Reporting to Improve Reproducibility and Facilitate Validity Assessment for Healthcare Database Studies V1.0

Purpose

Defining a study population and creating an analytic dataset from longitudinal healthcare databases involves many decisions. Our objective was to catalogue scientific decisions underpinning study execution that should be reported to facilitate replication and enable assessment of validity of studies conducted in large healthcare databases.

Methods

We reviewed key investigator decisions required to operate a sample of macros and software tools designed to create and analyze analytic cohorts from longitudinal streams of healthcare data. A panel of academic, regulatory, and industry experts in healthcare database analytics discussed and added to this list.

Conclusion

Evidence generated from large healthcare encounter and reimbursement databases is increasingly being sought by decision‐makers. Varied terminology is used around the world for the same concepts. Agreeing on terminology and which parameters from a large catalogue are the most essential to report for replicable research would improve transparency and facilitate assessment of validity. At a minimum, reporting for a database study should provide clarity regarding operational definitions for key temporal anchors and their relation to each other when creating the analytic dataset, accompanied by an attrition table and a design diagram.

A substantial improvement in reproducibility, rigor and confidence in real world evidence generated from healthcare databases could be achieved with greater transparency about operational study parameters used to create analytic datasets from longitudinal healthcare databases.

Standard-based comprehensive detection of adverse drug reaction signals from nursing statements and laboratory results in electronic health records

Objective. We propose 2 Medical Dictionary for Regulatory Activities–enabled pharmacovigilance algorithms, MetaLAB and MetaNurse, powered by a per-year meta-analysis technique and improved subject sampling strategy.

Matrials and methods. This study developed 2 novel algorithms, MetaLAB for laboratory abnormalities and MetaNurse for standard nursing statements, as significantly improved versions of our previous electronic health record (EHR)–based pharmacovigilance method, called CLEAR. Adverse drug reaction (ADR) signals from 117 laboratory abnormalities and 1357 standard nursing statements for all precautionary drugs (n  = 101) were comprehensively detected and validated against SIDER (Side Effect Resource) by MetaLAB and MetaNurse against 11 817 and 76 457 drug-ADR pairs, respectively.

Results. We demonstrate that MetaLAB (area under the curve, AUC = 0.61 ± 0.18) outperformed CLEAR (AUC = 0.55 ± 0.06) when we applied the same 470 drug-event pairs as the gold standard, as in our previous research. Receiver operating characteristic curves for 101 precautionary terms in the Medical Dictionary for Regulatory Activities Preferred Terms were obtained for MetaLAB and MetaNurse (0.69 ± 0.11; 0.62 ± 0.07), which complemented each other in terms of ADR signal coverage. Novel ADR signals discovered by MetaLAB and MetaNurse were successfully validated against spontaneous reports in the US Food and Drug Administration Adverse Event Reporting System database.

Discussion. The present study demonstrates the symbiosis of laboratory test results and nursing statements for ADR signal detection in terms of their system organ class coverage and performance profiles.

Conclusion. Systematic discovery and evaluation of the wide spectrum of ADR signals using standard-based observational electronic health record data across many institutions will affect drug development and use, as well as postmarketing surveillance and regulation.

The role of national registries in improving patient safety for hip and knee replacements

Background

The serious adverse events associated with metal on metal hip replacements have highlighted the importance of improving methods for monitoring surgical implants.

The new European Union (EU) device regulation will enforce post-marketing surveillance based on registries among other surveillance tools. Europe has a common regulatory environment, a common market for medical devices, and extensive experience with joint replacement registries. In this context, we elaborate how joint replacement registries, while building on existing structure and data, can better ensure safety and balance risks and benefits.

Main text

Actions to improve registry-based implant surveillance include: enriching baseline and diversifying outcomes data collection; improving methodology to limit bias; speeding-up failure detection by active real-time monitoring; implementing risk-benefit analysis; coordinating collaboration between registries; and translating knowledge gained from the data into clinical decision-making and public health policy.

Conclusions

The changes proposed here will improve patient safety, enforce the application of the new legal EU requirements, augment evidence, improve clinical decision-making, facilitate value-based health-care delivery, and provide up-to-date guidance for public health.

Reporting to Improve Reproducibility and Facilitate Validity Assessment for Healthcare Database Studies V1.0

PURPOSE

Defining a study population and creating an analytic dataset from longitudinal healthcare databases involves many decisions. Our objective was to catalogue scientific decisions underpinning study execution that should be reported to facilitate replication and enable assessment of validity of studies conducted in large healthcare databases.

METHODS

We reviewed key investigator decisions required to operate a sample of macros and software tools designed to create and analyze analytic cohorts from longitudinal streams of healthcare data. A panel of academic, regulatory, and industry experts in healthcare database analytics discussed and added to this list.

CONCLUSION

Evidence generated from large healthcare encounter and reimbursement databases is increasingly being sought by decision-makers. Varied terminology is used around the world for the same concepts. Agreeing on terminology and which parameters from a large catalogue are the most essential to report for replicable research would improve transparency and facilitate assessment of validity. At a minimum, reporting for a database study should provide clarity regarding operational definitions for key temporal anchors and their relation to each other when creating the analytic dataset, accompanied by an attrition table and a design diagram. A substantial improvement in reproducibility, rigor and confidence in real world evidence generated from healthcare databases could be achieved with greater transparency about operational study parameters used to create analytic datasets from longitudinal healthcare databases.

Markov Logic Networks for Adverse Drug Event Extraction from Text

Adverse drug events (ADEs) are a major concern and point of emphasis for the medical profession, government, and society. A diverse set of techniques from epidemiology, statistics, and computer science are being proposed and studied for ADE discovery from observational health data (e.g., EHR and claims data), social network data (e.g., Google and Twitter posts), and other information sources. Methodologies are needed for evaluating, quantitatively measuring, and comparing the ability of these various approaches to accurately discover ADEs. This work is motivated by the observation that text sources such as the Medline/Medinfo library provide a wealth of information on human health. Unfortunately, ADEs often result from unexpected interactions, and the connection between conditions and drugs is not explicit in these sources. Thus, in this work we address the question of whether we can quantitatively estimate relationships between drugs and conditions from the medical literature. This paper proposes and studies a state-of-the-art NLP-based extraction of ADEs from text.

Conducting Privacy-Preserving Multivariable Propensity Score Analysis When Patient Covariate Information Is Stored in Separate Locations

Distributed networks of health-care data sources are increasingly being utilized to conduct pharmacoepidemiologic database studies. Such networks may contain data that are not physically pooled but instead are distributed horizontally (separate patients within each data source) or vertically (separate measures within each data source) in order to preserve patient privacy. While multivariable methods for the analysis of horizontally distributed data are frequently employed, few practical approaches have been put forth to deal with vertically distributed health-care databases. In this paper, we propose 2 propensity score-based approaches to vertically distributed data analysis and test their performance using 5 example studies. We found that these approaches produced point estimates close to what could be achieved without partitioning. We further found a performance benefit (i.e., lower mean squared error) for sequentially passing a propensity score through each data domain (called the "sequential approach") as compared with fitting separate domain-specific propensity scores (called the "parallel approach"). These results were validated in a small simulation study. This proof-of-concept study suggests a new multivariable analysis approach to vertically distributed health-care databases that is practical, preserves patient privacy, and warrants further investigation for use in clinical research applications that rely on health-care databases.

Social Media Listening for Routine Post-Marketing Safety Surveillance

INTRODUCTION

Post-marketing safety surveillance primarily relies on data from spontaneous adverse event reports, medical literature, and observational databases. Limitations of these data sources include potential under-reporting, lack of geographic diversity, and time lag between event occurrence and discovery. There is growing interest in exploring the use of social media ('social listening') to supplement established approaches for pharmacovigilance. Although social listening is commonly used for commercial purposes, there are only anecdotal reports of its use in pharmacovigilance. Health information posted online by patients is often publicly available, representing an untapped source of post-marketing safety data that could supplement data from existing sources.

OBJECTIVES

The objective of this paper is to describe one methodology that could help unlock the potential of social media for safety surveillance.

METHODS

A third-party vendor acquired 24 months of publicly available Facebook and Twitter data, then processed the data by standardizing drug names and vernacular symptoms, removing duplicates and noise, masking personally identifiable information, and adding supplemental data to facilitate the review process. The resulting dataset was analyzed for safety and benefit information.

RESULTS

In Twitter, a total of 6,441,679 Medical Dictionary for Regulatory Activities (MedDRA(®)) Preferred Terms (PTs) representing 702 individual PTs were discussed in the same post as a drug compared with 15,650,108 total PTs representing 946 individual PTs in Facebook. Further analysis revealed that 26 % of posts also contained benefit information.

CONCLUSION

Social media listening is an important tool to augment post-marketing safety surveillance. Much work remains to determine best practices for using this rapidly evolving data source.

Scaling up health knowledge at European level requires sharing integrated data: an approach for collection of database specification

Computerized health care databases have been widely described as an excellent opportunity for research. The availability of "big data" has brought about a wave of innovation in projects when conducting health services research. Most of the available secondary data sources are restricted to the geographical scope of a given country and present heterogeneous structure and content. Under the umbrella of the European Innovation Partnership on Active and Healthy Ageing, collaborative work conducted by the partners of the group on "adherence to prescription and medical plans" identified the use of observational and large-population databases to monitor medication-taking behavior in the elderly. This article describes the methodology used to gather the information from available databases among the Adherence Action Group partners with the aim of improving data sharing on a European level. A total of six databases belonging to three different European countries (Spain, Republic of Ireland, and Italy) were included in the analysis. Preliminary results suggest that there are some similarities. However, these results should be applied in different contexts and European countries, supporting the idea that large European studies should be designed in order to get the most of already available databases.

Challenges and Opportunities for Exploring Patient-Level Data

The proper exploration of patient-level data will pave the way towards personalised medicine. To better assess the state of the art in this field we identify the challenges and uncover the opportunities for the exploration of patient-level data through the review of well-known initiatives and projects focusing on the exploration of patient-level data. These cover a broad array of topics, from genomics to patient registries up to rare diseases research, among others. For each, we identified basic goals, involved partners, defined strategies and key technological and scientific outcomes, establishing the foundation for our analysis framework with four pillars: control, sustainability, technology, and science. Substantial research outcomes have been produced towards the exploration of patient-level data. The potential behind these data will be essential to realise the personalised medicine premise in upcoming years. Hence, relevant stakeholders continually push forward new developments in this domain, bringing novel opportunities that are ripe for exploration. Despite last decade's translational research advances, personalised medicine is still far from being a reality. Patients' data underlying potential goes beyond daily clinical practice. There are miscellaneous challenges and opportunities open for the exploration of these data by academia and business stakeholders.

Large-Scale Predictive Drug Safety: From Structural Alerts to Biological Mechanisms

The recent explosion of data linking drugs, proteins, and pathways with safety events has promoted the development of integrative systems approaches to large-scale predictive drug safety. The added value of such approaches is that, beyond the traditional identification of potentially labile chemical fragments for selected toxicity end points, they have the potential to provide mechanistic insights for a much larger and diverse set of safety events in a statistically sound nonsupervised manner, based on the similarity to drug classes, the interaction with secondary targets, and the interference with biological pathways. The combined identification of chemical and biological hazards enhances our ability to assess the safety risk of bioactive small molecules with higher confidence than that using structural alerts only. We are still a very long way from reliably predicting drug safety, but advances toward gaining a better understanding of the mechanisms leading to adverse outcomes represent a step forward in this direction.

Targets of drugs are generally, and targets of drugs having side effects are specifically good spreaders of human interactome perturbations

Network-based methods are playing an increasingly important role in drug design. Our main question in this paper was whether the efficiency of drug target proteins to spread perturbations in the human interactome is larger if the binding drugs have side effects, as compared to those which have no reported side effects. Our results showed that in general, drug targets were better spreaders of perturbations than non-target proteins, and in particular, targets of drugs with side effects were also better spreaders of perturbations than targets of drugs having no reported side effects in human protein-protein interaction networks. Colorectal cancer-related proteins were good spreaders and had a high centrality, while type 2 diabetes-related proteins showed an average spreading efficiency and had an average centrality in the human interactome. Moreover, the interactome-distance between drug targets and disease-related proteins was higher in diabetes than in colorectal cancer. Our results may help a better understanding of the network position and dynamics of drug targets and disease-related proteins, and may contribute to develop additional, network-based tests to increase the potential safety of drug candidates.

Secure Secondary Use of Clinical Data with Cloud-based NLP Services. Towards a Highly Scalable Research Infrastructure

OBJECTIVES

The secondary use of clinical data provides large opportunities for clinical and translational research as well as quality assurance projects. For such purposes, it is necessary to provide a flexible and scalable infrastructure that is compliant with privacy requirements. The major goals of the cloud4health project are to define such an architecture, to implement a technical prototype that fulfills these requirements and to evaluate it with three use cases.

METHODS

The architecture provides components for multiple data provider sites such as hospitals to extract free text as well as structured data from local sources and de-identify such data for further anonymous or pseudonymous processing. Free text documentation is analyzed and transformed into structured information by text-mining services, which are provided within a cloud-computing environment. Thus, newly gained annotations can be integrated along with the already available structured data items and the resulting data sets can be uploaded to a central study portal for further analysis.

RESULTS

Based on the architecture design, a prototype has been implemented and is under evaluation in three clinical use cases. Data from several hundred patients provided by a University Hospital and a private hospital chain have already been processed.

CONCLUSIONS

Cloud4health has shown how existing components for secondary use of structured data can be complemented with text-mining in a privacy compliant manner. The cloud-computing paradigm allows a flexible and dynamically adaptable service provision that facilitates the adoption of services by data providers without own investments in respective hardware resources and software tools.

An effective method of large scale ontology matching

Background

We are currently facing a proliferation of heterogeneous biomedical data sources accessible through various knowledge-based applications. These data are annotated by increasingly extensive and widely disseminated knowledge organisation systems ranging from simple terminologies and structured vocabularies to formal ontologies. In order to solve the interoperability issue, which arises due to the heterogeneity of these ontologies, an alignment task is usually performed. However, while significant effort has been made to provide tools that automatically align small ontologies containing hundreds or thousands of entities, little attention has been paid to the matching of large sized ontologies in the life sciences domain.

Results

We have designed and implemented ServOMap, an effective method for large scale ontology matching. It is a fast and efficient high precision system able to perform matching of input ontologies containing hundreds of thousands of entities. The system, which was included in the 2012 and 2013 editions of the Ontology Alignment Evaluation Initiative campaign, performed very well. It was ranked among the top systems for the large ontologies matching.

Conclusions

We proposed an approach for large scale ontology matching relying on Information Retrieval (IR) techniques and the combination of lexical and machine learning contextual similarity computing for the generation of candidate mappings. It is particularly adapted to the life sciences domain as many of the ontologies in this domain benefit from synonym terms taken from the Unified Medical Language System and that can be used by our IR strategy. The ServOMap system we implemented is able to deal with hundreds of thousands entities with an efficient computation time.

Pharmacovigilance in oncology: evaluation of current practice and future perspectives

RATIONALE, AIMS AND OBJECTIVES

Pharmacovigilance (PV), or drug safety monitoring, aims to improve patient safety through the detection and management of drug-related adverse reactions. It is implemented both by spontaneous reporting of adverse drug reactions (ADRs) and by careful detection of signals suggestive of drug toxicity. PV is an important clinical topic in clinical practice and pharmacotherapy, assuring the maintenance of a safe risk/benefit ratio throughout the commercial life cycle of a drug.

METHODS

We conducted a structured literature search on PubMed, Scopus, Cinahl and the Cochrane Library. We also performed manual searches in international databases of ADR individual reports to outline a structured profile on the topic. Our goal was to review key elements that affect safety monitoring of cancer drugs and their appropriate use, highlighting the strengths and weaknesses of PV in oncology.

RESULTS

This paper provides an understanding of the methodologies used by PV in current clinical practice and particularly in cancer drug therapy; a focus upon reporting of ADRs by health professionals and patients; and a focus upon methods used by PV to detect new signals of risk/harm related to medicines utilization.

CONCLUSION

To our knowledge, few articles focus upon the importance of PV and post-marketing surveillance of cancer drug therapies. Structured management of spontaneous reports of ADRs and data collection is essential to monitoring the safe use of drugs in this field in which pharmacotherapy is affected by high incidence of drug-related complications and by a narrow benefit/risk ratio.

"Big data" and the electronic health record

OBJECTIVES

Implementation of Electronic Health Record (EHR) systems continues to expand. The massive number of patient encounters results in high amounts of stored data. Transforming clinical data into knowledge to improve patient care has been the goal of biomedical informatics professionals for many decades, and this work is now increasingly recognized outside our field. In reviewing the literature for the past three years, we focus on "big data" in the context of EHR systems and we report on some examples of how secondary use of data has been put into practice.

METHODS

We searched PubMed database for articles from January 1, 2011 to November 1, 2013. We initiated the search with keywords related to "big data" and EHR. We identified relevant articles and additional keywords from the retrieved articles were added. Based on the new keywords, more articles were retrieved and we manually narrowed down the set utilizing predefined inclusion and exclusion criteria.

RESULTS

Our final review includes articles categorized into the themes of data mining (pharmacovigilance, phenotyping, natural language processing), data application and integration (clinical decision support, personal monitoring, social media), and privacy and security.

CONCLUSION

The increasing adoption of EHR systems worldwide makes it possible to capture large amounts of clinical data. There is an increasing number of articles addressing the theme of "big data", and the concepts associated with these articles vary. The next step is to transform healthcare big data into actionable knowledge.

Identifying plausible adverse drug reactions using knowledge extracted from the literature

Pharmacovigilance involves continually monitoring drug safety after drugs are put to market. To aid this process; algorithms for the identification of strongly correlated drug/adverse drug reaction (ADR) pairs from data sources such as adverse event reporting systems or Electronic Health Records have been developed. These methods are generally statistical in nature, and do not draw upon the large volumes of knowledge embedded in the biomedical literature. In this paper, we investigate the ability of scalable Literature Based Discovery (LBD) methods to identify side effects of pharmaceutical agents. The advantage of LBD methods is that they can provide evidence from the literature to support the plausibility of a drug/ADR association, thereby assisting human review to validate the signal, which is an essential component of pharmacovigilance. To do so, we draw upon vast repositories of knowledge that has been extracted from the biomedical literature by two Natural Language Processing tools, MetaMap and SemRep. We evaluate two LBD methods that scale comfortably to the volume of knowledge available in these repositories. Specifically, we evaluate Reflective Random Indexing (RRI), a model based on concept-level co-occurrence, and Predication-based Semantic Indexing (PSI), a model that encodes the nature of the relationship between concepts to support reasoning analogically about drug-effect relationships. An evaluation set was constructed from the Side Effect Resource 2 (SIDER2), which contains known drug/ADR relations, and models were evaluated for their ability to "rediscover" these relations. In this paper, we demonstrate that both RRI and PSI can recover known drug-adverse event associations. However, PSI performed better overall, and has the additional advantage of being able to recover the literature underlying the reasoning pathways it used to make its predictions.

Gathering and exploring scientific knowledge in pharmacovigilance

Pharmacovigilance plays a key role in the healthcare domain through the assessment, monitoring and discovery of interactions amongst drugs and their effects in the human organism. However, technological advances in this field have been slowing down over the last decade due to miscellaneous legal, ethical and methodological constraints. Pharmaceutical companies started to realize that collaborative and integrative approaches boost current drug research and development processes. Hence, new strategies are required to connect researchers, datasets, biomedical knowledge and analysis algorithms, allowing them to fully exploit the true value behind state-of-the-art pharmacovigilance efforts. This manuscript introduces a new platform directed towards pharmacovigilance knowledge providers. This system, based on a service-oriented architecture, adopts a plugin-based approach to solve fundamental pharmacovigilance software challenges. With the wealth of collected clinical and pharmaceutical data, it is now possible to connect knowledge providers' analysis and exploration algorithms with real data. As a result, new strategies allow a faster identification of high-risk interactions between marketed drugs and adverse events, and enable the automated uncovering of scientific evidence behind them. With this architecture, the pharmacovigilance field has a new platform to coordinate large-scale drug evaluation efforts in a unique ecosystem, publicly available at http://bioinformatics.ua.pt/euadr/.

Improving case-based reasoning systems by combining k-nearest neighbour algorithm with logistic regression in the prediction of patients' registration on the renal transplant waiting list

Introduction

Case-based reasoning (CBR) is an emerging decision making paradigm in medical research where new cases are solved relying on previously solved similar cases. Usually, a database of solved cases is provided, and every case is described through a set of attributes (inputs) and a label (output). Extracting useful information from this database can help the CBR system providing more reliable results on the yet to be solved cases.

Objective

We suggest a general framework where a CBR system, viz. K-Nearest Neighbour (K-NN) algorithm, is combined with various information obtained from a Logistic Regression (LR) model, in order to improve prediction of access to the transplant waiting list.

Methods

LR is applied, on the case database, to assign weights to the attributes as well as the solved cases. Thus, five possible decision making systems based on K-NN and/or LR were identified: a standalone K-NN, a standalone LR and three soft K-NN algorithms that rely on the weights based on the results of the LR. The evaluation was performed under two conditions, either using predictive factors known to be related to registration, or using a combination of factors related and not related to registration.

Results and Conclusion

The results show that our suggested approach, where the K-NN algorithm relies on both weighted attributes and cases, can efficiently deal with non relevant attributes, whereas the four other approaches suffer from this kind of noisy setups. The robustness of this approach suggests interesting perspectives for medical problem solving tools using CBR methodology.

Drug-induced acute myocardial infarction: identifying 'prime suspects' from electronic healthcare records-based surveillance system

BACKGROUND

Drug-related adverse events remain an important cause of morbidity and mortality and impose huge burden on healthcare costs. Routinely collected electronic healthcare data give a good snapshot of how drugs are being used in 'real-world' settings.

OBJECTIVE

To describe a strategy that identifies potentially drug-induced acute myocardial infarction (AMI) from a large international healthcare data network.

METHODS

Post-marketing safety surveillance was conducted in seven population-based healthcare databases in three countries (Denmark, Italy, and the Netherlands) using anonymised demographic, clinical, and prescription/dispensing data representing 21,171,291 individuals with 154,474,063 person-years of follow-up in the period 1996-2010. Primary care physicians' medical records and administrative claims containing reimbursements for filled prescriptions, laboratory tests, and hospitalisations were evaluated using a three-tier triage system of detection, filtering, and substantiation that generated a list of drugs potentially associated with AMI. Outcome of interest was statistically significant increased risk of AMI during drug exposure that has not been previously described in current literature and is biologically plausible.

RESULTS

Overall, 163 drugs were identified to be associated with increased risk of AMI during preliminary screening. Of these, 124 drugs were eliminated after adjustment for possible bias and confounding. With subsequent application of criteria for novelty and biological plausibility, association with AMI remained for nine drugs ('prime suspects'): azithromycin; erythromycin; roxithromycin; metoclopramide; cisapride; domperidone; betamethasone; fluconazole; and megestrol acetate.

LIMITATIONS

Although global health status, co-morbidities, and time-invariant factors were adjusted for, residual confounding cannot be ruled out.

CONCLUSION

A strategy to identify potentially drug-induced AMI from electronic healthcare data has been proposed that takes into account not only statistical association, but also public health relevance, novelty, and biological plausibility. Although this strategy needs to be further evaluated using other healthcare data sources, the list of 'prime suspects' makes a good starting point for further clinical, laboratory, and epidemiologic investigation.