A New Era in Pharmacovigilance: Toward Real-World Data and Digital Monitoring

A Comparative Analysis of ADRs under Obeticholic Acid and Ursodeoxycholic Acid in Cholestatic Liver Diseases Using the FAERS Database

BACKGROUND

The objective of this study was to compare the safety profiles of OCA and UDCA for the treatment of PBC using the FDA Adverse Event Reporting System database.

METHODS

We extracted reports for OCA from 2016 to 2023 and UDCA from 2004 to 2023. Demographic details, adverse events (AEs), and concomitant medications were analyzed using descriptive statistics and signal detection methods.

RESULTS

The most common for OCA were pruritus (1345 cases, ROR 20.96) and fatigue (528 cases, ROR 3.46). UDCA was more frequently associated with hepatocellular carcinoma (22 cases, ROR 16.37) and type I hypersensitivity reactions (11 cases, ROR 12.77). OCA was also linked to a higher frequency of constipation (161 cases, ROR 3.92) and increased blood alkaline phosphatase levels (145 cases, ROR 44.27).

CONCLUSION

This study reveals distinct safety profiles for OCA and UDCA in the treatment of PBC. OCA is associated with a higher frequency of pruritus, fatigue, constipation, and increased blood alkaline phosphatase levels, while UDCA is linked to hepatocellular carcinoma and type I hypersensitivity reactions. These findings support personalized treatment approaches based on individual patient characteristics.

Barriers in reporting adverse effects of medical devices: a literature review

Medical devices play an essential role in the delivery of healthcare but its use is not entirely risk free. There are several instances where it causes mortality or morbidity among users. It is important to evaluate the risks involved at every stage of its application to bring improvement in the standard of healthcare. For the purpose Materiovigilance Program of India was launched on July 6, 2015. Despite these efforts, available data suggests that reporting of adverse events is very low. The present study aims to identify barriers that influence the reporting of adverse events of medical devices and outline a strategy to overcome these barriers. Systemic review method has been adopted to achieve these ends. Thirty-one papers have been selected based on the inclusion criteria related to objective of the study. Lack of awareness, attitude, and resources are found to be major barriers at the individual level for not reporting adverse effects of medical devices. The organizational factors such as hierarchical set up, lack of time and incentives, and furthermore lack of industry responsiveness have been identified as prominent barriers to the reporting of adverse events. In order to improve the reporting level, it is important to make access and contact easier with the reporting system. Engaging healthcare professionals at various levels by acknowledging and appreciating their contribution. The adverse events of medical devices should not be restricted to physicians; only rather other health care professional such as nurses, pharmacists, and technicians should also be encouraged to report any adverse event of medical devices.

A discovery and verification approach to pharmacovigilance using electronic healthcare data

Introduction

Pharmacovigilance is vital for drug safety. The process typically involves two key steps: initial signal generation from spontaneous reporting systems (SRSs) and subsequent expert review to assess the signals' (potential) causality and decide on the appropriate action.

Methods

We propose a novel discovery and verification approach to pharmacovigilance based on electronic healthcare data. We enhance the signal detection phase by introducing an ensemble of methods which generated signals are combined using Borda count ranking; a method designed to emphasize consensus. Ensemble methods tend to perform better when data is noisy and leverage the strengths of individual classifiers, while trying to mitigate some of their limitations. Additionally, we offer the committee of medical experts with the option to perform an in-depth investigation of selected signals through tailored pharmacoepidemiological studies to evaluate their plausibility or spuriousness. To illustrate our approach, we utilize data from the German Pharmacoepidemiological Research Database, focusing on drug reactions to the direct oral anticoagulant rivaroxaban.

Results

In this example, the ensemble method is built upon the Bayesian confidence propagation neural network, longitudinal Gamma Poisson shrinker, penalized regression and random forests. We also conduct a pharmacoepidemiological verification study in the form of a nested active comparator case-control study, involving patients diagnosed with atrial fibrillation who initiated anticoagulant treatment between 2011 and 2017.

Discussion

The case study reveals our ability to detect known adverse drug reactions and discover new signals. Importantly, the ensemble method is computationally efficient. Hasty false conclusions can be avoided by a verification study, which is, however, time-consuming to carry out. We provide an online tool for easy application: https://borda.bips.eu.

A Real‑World Pharmacovigilance Study of FDA Adverse Event Reporting System (FAERS) for Mavacamten

BACKGROUND

Mavacamten is a first-in-class cardiac myosin inhibitor approved by the US Food and Drug Administration (FDA) for symptomatic obstructive hypertrophic cardiomyopathy (HCM). This pharmacovigilance study aimed to assess mavacamten-related adverse drug reactions (ADRs) in the real world as reported in the FDA Adverse Event Reporting System (FAERS).

METHODS

We conducted disproportionality analyses with four signal detection algorithms-reporting odds ratio (ROR), proportional reporting ratio (PRR), Bayesian confidence propagation neural network, and the multi-item gamma Poisson shrinker to identify mavacamten-related ADRs.

RESULTS

Out of 4,500,131 reports from the FAERS database, 1004 mavacamten-related ADRs were identified from 1 January 2022 to 30 September 2023. A total of 26 significant disproportionality preferred terms (PTs) conforming to the four signal detection algorithms were noted. Some of the statistically significant cardiac ADRs at PT level include decreased ejection fraction (EF) [ROR 33.60 (95% confidence interval, CI 21.79-51.82), PRR 32.86 (χ2 615.96), information component (IC) 5.03, IC025 4.61, empirical Bayesian geometric mean (EBGM) 32.77, EBGM05 21.25], cardiac failure [ROR 9.39 (95% CI 6.49-13.60), PRR 9.13 (χ2 202.42), IC 3.19, IC025 2.83, EBGM 9.12, EBGM05 6.30], and atrial fibrillation [ROR 16.63 (95% CI 12.72-21.75), PRR 15.66 (χ2 769.93), IC 3.97, IC025 3.71, EBGM 15.64, EBGM05 11.96].

CONCLUSIONS

The results of our study were consistent with the safety data of clinical trials, including reduced ejection fraction, atrial fibrillation, dyspnea, and syncope. We also found potential new and unexpected ADR signals, such as urinary tract infection, gout, and peripheral edema.

Serious gaming as potential training tool for recognition of adverse drug reactions: side-effect exposure-medical education (SeeMe)

PURPOSE

The recognition of adverse drug reactions (ADRs) is an important part of daily clinical work. However, medical education in this field is mostly drug-based and does not address adequately the complexity of this field regarding individual risk factors and polypharmacy. This study investigates the potential of the web-based serious game SeeMe (side-effect exposure-medical education) in pharmacological education of medical students to improve the recognition of relevant ADRs.

METHODS

One hundred fifty-seven medical students were recruited to evaluate the serious game SeeMe. SeeMe was developed to improve knowledge and recognition of ADRs in clinical practice. Players take on the role of a physician trying to understand fictional patients with ADRs. Before and after an 8-week playing period, an evaluation was carried out through a pre- and post-questionnaire and a pre- and post- knowledge test.

RESULTS

The students achieved significantly better results in the knowledge test, as almost twice as many exam-relevant questions were answered correctly (p < 0.001). The serious game had a positive effect on the students' perception of the importance of ADRs.

CONCLUSION

This study demonstrates the potential of web- and case-based fictional serious games in medical education. The improved recognition of side effects represents a crucial step for education and training in clinical pharmacology. Future versions of the serious game may take this further and focus on training in the treatment of ADRs and their relevance in various healthcare professions.

Leveraging Real-World Data in Safety Signal Assessment

PURPOSE

TransCelerate BioPharma surveyed its member biopharmaceutical companies to understand current practices and identify opportunities to complement safety signal assessment with rapid real-world data (RWD) analysis.

METHODS

A voluntary 30-question questionnaire regarding the use of RWD in safety signal assessment was disseminated to subject matter experts at all TransCelerate member companies in July 2022. Responses were blinded, aggregated, summarized, and presented.

RESULTS

Eighteen of 20 member companies provided responses to the questionnaire. Sixteen (89%) companies reported actively leveraging RWD in their signal assessment processes. Of 18 respondent companies, 8 (44%) routinely use rapid approaches to RWD analysis, 7 (39%) utilize rapid RWD analysis non-routinely or in a pilot setting, 2 (11%) are considering using rapid RWD analysis, and 1 (6%) has no plans to use rapid RWD analysis for their signal assessment. Most companies reported that RWD adds context to and improves quality of signal assessments. To conduct RWD analysis for signal assessment, 16 of 17 (94%) respondent companies utilize or plan to utilize internally available data, 8 (47%) utilize both internal and external data, and 3 (18%) utilize data networks. Respondents identified key challenges to rapidly performing RWD analyses, including data access/availability, time for analysis execution, and uncertainties regarding acceptance of minimal or non-protocolized approaches by health authorities.

CONCLUSION

Biopharmaceutical companies reported that they see value in the use of rapid RWD analyses for complementing signal assessments. Future work is recommended to offer a framework and process for use of rapid use of RWD analyses in signal assessment.

In-silico approaches to assessing multiple high-level drug-drug and drug-disease adverse drug effects

INTRODUCTION

Pharmacovigilance plays a pivotal role in monitoring adverse events (AEs) related to chemical substances in human/animal populations. With increasing spontaneous-reporting systems, researchers turned to in-silico approaches to efficiently analyze drug safety profiles. Here, we review in-silico methods employed for assessing multiple drug-drug/drug-disease AEs covered by comparative analyses and visualization strategies.

AREAS COVERED

Disproportionality, involving multi-stage statistical methodologies and data processing, identifies safety signals among drug-AE pairs. By stratifying data based on disease indications/demographics, researchers address confounders and assess drug safety. Comparative analyses, including clustering techniques and visualization techniques, assess drug similarities, patterns, and trends, calculate correlations, and identify distinct toxicities. Furthermore, we conducted a thorough Scopus search on 'pharmacovigilance,' yielding 5,836 publications spanning 2003 to 2023.

EXPERT OPINION

Pharmacovigilance relies on diverse data sources, presenting challenges in the integration of in-silico approaches and requiring compliance with regulations and AI adoption. Systematic use of statistical analyses enables identifications of potential risks with drugs. Frequentist and Bayesian methods are used in disproportionalities, each with its strengths and weaknesses. Integration of pharmacogenomics with pharmacovigilance enables personalized medicine, with AI further enhancing patient engagement. This multidisciplinary approach holds promise, improving drug efficacy and safety, and should be a core mission of One-Health studies.

Hospital registration of adverse drug reactions in electronic health records: importance and contribution to pharmacovigilance

BACKGROUND

Information on registered adverse drug reactions (ADRs) in hospitals may provide a large real-world data source that can be used to ensure patients' safety. This study aimed to assess the potential contribution of hospital registration of ADRs in electronic health records (EHR) to pharmacovigilance.

RESEARCH DESIGN AND METHODS

An observational retrospective descriptive study using data from the Jeroen Bosch Hospital in the Netherlands in 2019. 'Serious and/or severe' and 'previously unknown' ADRs registered systematically in the corresponding field of EHRs were assessed.

RESULTS

ADR data concerning 1010 patients were included. In total, 1630 ADRs were registered in EHRs. Fifty-eight serious and/or severe ADRs (5.2%) were registered. Tubulointerstitial nephritis was the most frequently registered severe ADR and was mainly associated with antibacterials for systemic use. A total of 82 previously unknown ADRs (5%) were registered. 'Migraine' and 'chest pain' were the most frequently registered unknown ADRs. Additionally, 25 ADRs (1.5%) were registered that may be attributable to 10 drugs 'under additional monitoring.'

CONCLUSIONS

Hospital registrations of ADRs in EHRs provide information on ADRs, which are challenging to assess during clinical trials. However, improvements are required to optimize this registration before it can serve as a valuable data source for pharmacovigilance purposes.

Machine Learning Techniques for Predicting Drug-Related Side Effects: A Scoping Review

BACKGROUND

Drug safety relies on advanced methods for timely and accurate prediction of side effects. To tackle this requirement, this scoping review examines machine-learning approaches for predicting drug-related side effects with a particular focus on chemical, biological, and phenotypical features.

METHODS

This was a scoping review in which a comprehensive search was conducted in various databases from 1 January 2013 to 31 December 2023.

RESULTS

The results showed the widespread use of Random Forest, k-nearest neighbor, and support vector machine algorithms. Ensemble methods, particularly random forest, emphasized the significance of integrating chemical and biological features in predicting drug-related side effects.

CONCLUSIONS

This review article emphasized the significance of considering a variety of features, datasets, and machine learning algorithms for predicting drug-related side effects. Ensemble methods and Random Forest showed the best performance and combining chemical and biological features improved prediction. The results suggested that machine learning techniques have some potential to improve drug development and trials. Future work should focus on specific feature types, selection techniques, and graph-based methods for even better prediction.

Leveraging patient experience data to guide medicines development, regulation, access decisions and clinical care in the EU

Patient experience data (PED), provided by patients/their carers without interpretation by clinicians, directly capture what matters more to patients on their medical condition, treatment and impact of healthcare. PED can be collected through different methodologies and these need to be robust and validated for its intended use. Medicine regulators are increasingly encouraging stakeholders to generate, collect and submit PED to support both scientific advice in development programs and regulatory decisions on the approval and use of these medicines. This article reviews the existing definitions and types of PED and demonstrate the potential for use in different settings of medicines' life cycle, focusing on Patient-Reported Outcomes (PRO) and Patient Preferences (PP). Furthermore, it addresses some challenges and opportunities, alluding to important regulatory guidance that has been published, methodological aspects and digitalization, highlighting the lack of guidance as a key hurdle to achieve more systematic inclusion of PED in regulatory submissions. In addition, the article discusses opportunities at European and global level that could be implemented to leverage PED use. New digital tools that allow patients to collect PED in real time could also contribute to these advances, but it is equally important not to overlook the challenges they entail. The numerous and relevant initiatives being developed by various stakeholders in this field, including regulators, show their confidence in PED's value and create an ideal moment to address challenges and consolidate PED use across medicines' life cycle.

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.

Innovations in pharmacovigilance studies of medicines in older people

Pharmacovigilance is defined by the World Health Organization as "the science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other medicine/vaccine related problem". Pharmacovigilance studies are critical for detecting and assessing adverse events of medicines that may not have been observed in clinical trials. This activity is especially important in older people who are often excluded from clinical trials as they have multiple chronic conditions and use multiple medicines for longer durations than the clinical trials. In this narrative review we describe innovative methods in pharmacovigilance studies of medicines in older people that leverage the increasing availability of digital health technologies, electronic health records and real-world health data to identify and quantify medication related harms in older people.

Overcoming Barriers to Discovery and Implementation of Equitable Pharmacogenomic Testing in Oncology

Pharmacogenomics (PGx), the study of inherited genomic variation and drug response or safety, is a vital tool in precision medicine. In oncology, testing to identify PGx variants offers patients the opportunity for customized treatments that can minimize adverse effects and maximize the therapeutic benefits of drugs used for cancer treatment and supportive care. Because individuals of shared ancestry share specific genetic variants, PGx factors may contribute to outcome disparities across racial and ethnic categories when genetic ancestry is not taken into account or mischaracterized in PGx research, discovery, and application. Here, we examine how the current scientific understanding of the role of PGx in differential oncology safety and outcomes may be biased toward a greater understanding and more complete clinical implementation of PGx for individuals of European descent compared with other genetic ancestry groups. We discuss the implications of this bias for PGx discovery, access to care, drug labeling, and patient and provider understanding and use of PGx approaches. Testing for somatic genetic variants is now the standard of care in treatment of many solid tumors, but the integration of PGx into oncology care is still lacking despite demonstrated actionable findings from PGx testing, reduction in avoidable toxicity and death, and return on investment from testing. As the field of oncology is poised to expand and integrate germline genetic variant testing, it is vital that PGx discovery and application are equitable for all populations. Recommendations are introduced to address barriers to facilitate effective and equitable PGx application in cancer care.

Development and application of pharmacological statin-associated muscle symptoms phenotyping algorithms using structured and unstructured electronic health records data

Importance

Statins are widely prescribed cholesterol-lowering medications in the United States, but their clinical benefits can be diminished by statin-associated muscle symptoms (SAMS), leading to discontinuation.

Objectives

In this study, we aimed to develop and validate a pharmacological SAMS clinical phenotyping algorithm using electronic health records (EHRs) data from Minnesota Fairview.

Materials and Methods

We retrieved structured and unstructured EHR data of statin users and manually ascertained a gold standard set of SAMS cases and controls using the published SAMS-Clinical Index tool from clinical notes in 200 patients. We developed machine learning algorithms and rule-based algorithms that incorporated various criteria, including ICD codes, statin allergy, creatine kinase elevation, and keyword mentions in clinical notes. We applied the best-performing algorithm to the statin cohort to identify SAMS.

Results

We identified 16 889 patients who started statins in the Fairview EHR system from 2010 to 2020. The combined rule-based (CRB) algorithm, which utilized both clinical notes and structured data criteria, achieved similar performance compared to machine learning algorithms with a precision of 0.85, recall of 0.71, and F1 score of 0.77 against the gold standard set. Applying the CRB algorithm to the statin cohort, we identified the pharmacological SAMS prevalence to be 1.9% and selective risk factors which included female gender, coronary artery disease, hypothyroidism, and use of immunosuppressants or fibrates.

Discussion and Conclusion

Our study developed and validated a simple pharmacological SAMS phenotyping algorithm that can be used to create SAMS case/control cohort to enable further analysis which can lead to the development of a SAMS risk prediction model.

From randomised controlled trials to real-world data: Clinical evidence to guide management of diabetic macular oedema

Randomised clinical trials (RCTs) are generally considered the gold-standard for providing scientific evidence for treatments' effectiveness and safety but their findings may not always be generalisable to the broader population treated in routine clinical practice. RCTs include highly selected patient populations that fit specific inclusion and exclusion criteria. Although they may have a lower level of certainty than RCTs on the evidence hierarchy, real-world data (RWD), such as observational studies, registries and databases, provide real-world evidence (RWE) that can complement RCTs. For example, RWE may help satisfy requirements for a new indication of an already approved drug and help us better understand long-term treatment effectiveness, safety and patterns of use in clinical practice. Many countries have set up registries, observational studies and databases containing information on patients with retinal diseases, such as diabetic macular oedema (DMO). These DMO RWD have produced significant clinical evidence in the past decade that has changed the management of DMO. RWD and medico-administrative databases are a useful resource to identify low frequency safety signals. They often have long-term follow-up with a large number of patients and minimal exclusion criteria. We will discuss improvements in healthcare information exchange technologies, such as blockchain technology and FHIR (Fast Healthcare Interoperability Resources), which will connect and extend databases already available. These registries can be linked with existing or emerging retinal imaging modalities using artificial intelligence to aid diagnosis, treatment decisions and provide prognostic information. The results of RCTs and RWE are combined to provide evidence-based guidelines.

An Exploratory Study of the Impact of COVID-19 Vaccine Spontaneous Reporting on Masking Signal Detection in EudraVigilance

INTRODUCTION

During the signal detection process, statistical methods are used to identify drug-event combinations (DECs) which are disproportionately reported when compared with other drugs and events in the entire database. We hypothesise that the high volume of COVID-19 vaccine adverse drug reaction (ADR) reports transmitted to EudraVigilance may have affected the performance of disproportionality statistics used in routine signal detection, potentially resulting in signals either being masked, or false associations being flagged as potential signals.

OBJECTIVE

Our aim was to study the impact of COVID-19 vaccine spontaneous reporting on statistical signal detection in EudraVigilance.

METHODS

We recalculated the reporting odds ratio (ROR) for signals that were previously discussed at the level of the Pharmacovigilance Risk Assessment Committee, or signals that were retrieved from EudraVigilance, by omitting COVID-19 vaccine reports from the standard ROR calculation and then comparing the lower confidence interval (LCI) of the recalculated ROR to the LCI of the actual ROR in EudraVigilance.

RESULTS

In total, 52 signals for 38 active substances were reviewed. For 35 signals, the LCI of the recalculated ROR value was lower than the LCI of the actual ROR (suggesting that COVID-19 vaccine ADR reporting had a positive effect on the strength of the signal) while for 15 signals the LCI of the recalculated ROR value was higher than the LCI of the actual ROR (suggesting that COVID-19 vaccine ADR reporting had an attenuating effect on the strength of the signal). For two signals, no change in the ROR was observed. In our analysis, six significant results were found. Five DECs were found to be masked: bleomycin and immune thrombocytopenia (actual ROR LCI = 0.94, recalculated ROR LCI = 1.02), vortioxetine and heavy menstrual bleeding (actual ROR LCI = 0.3, recalculated ROR LCI = 1.06), caplacizumab and heavy menstrual bleeding (actual ROR LCI = 0.98, recalculated ROR LCI = 3.47), ziprasidone and amenorrhoea (actual ROR LCI = 0.84, recalculated ROR LCI = 1.67), and azacitidine and pericarditis (actual ROR LCI = 0.81, recalculated ROR LCI = 2.01). For the DEC of adalimumab and immune reconstitution inflammatory syndrome, the LCI of the actual ROR value was 1.14 and removing COVID-19 vaccine reporting resulted in an LCI of the recalculated ROR value of 0.94 (below threshold).

CONCLUSIONS

We demonstrated five cases of masking and one case of false-positive association due to the influence of COVID-19 vaccine spontaneous reporting on the ROR. This suggests that the high number of adverse drug reaction reports for COVID-19 vaccines in EudraVigilance has the potential to affect routine statistical signal detection activities. The impact of COVID-19 vaccine ADR reports on current signal detection practices requires further evaluation and solutions to tackle masking issues in EudraVigilance may need to be developed.

Bidirectional Encoder Representations from Transformers-like large language models in patient safety and pharmacovigilance: A comprehensive assessment of causal inference implications

Causality assessment is vital in patient safety and pharmacovigilance (PSPV) for safety signal detection, adverse reaction management, and regulatory submission. Large language models (LLMs), especially those designed with transformer architecture, are revolutionizing various fields, including PSPV. While attempts to utilize Bidirectional Encoder Representations from Transformers (BERT)-like LLMs for causal inference in PSPV are underway, a detailed evaluation of "fit-for-purpose" BERT-like model selection to enhance causal inference performance within PSPV applications remains absent. This study conducts an in-depth exploration of BERT-like LLMs, including generic pre-trained BERT LLMs, domain-specific pre-trained LLMs, and domain-specific pre-trained LLMs with safety knowledge-specific fine-tuning, for causal inference in PSPV. Our investigation centers around (1) the influence of data complexity and model architecture, (2) the correlation between the BERT size and its impact, and (3) the role of domain-specific training and fine-tuning on three publicly accessible PSPV data sets. The findings suggest that (1) BERT-like LLMs deliver consistent predictive power across varied data complexity levels, (2) the predictive performance and causal inference results do not directly correspond to the BERT-like model size, and (3) domain-specific pre-trained LLMs, with or without safety knowledge-specific fine-tuning, surpass generic pre-trained BERT models in causal inference. The findings are valuable to guide the future application of LLMs in a broad range of application.

A causal roadmap for generating high-quality real-world evidence

Increasing emphasis on the use of real-world evidence (RWE) to support clinical policy and regulatory decision-making has led to a proliferation of guidance, advice, and frameworks from regulatory agencies, academia, professional societies, and industry. A broad spectrum of studies use real-world data (RWD) to produce RWE, ranging from randomized trials with outcomes assessed using RWD to fully observational studies. Yet, many proposals for generating RWE lack sufficient detail, and many analyses of RWD suffer from implausible assumptions, other methodological flaws, or inappropriate interpretations. The Causal Roadmap is an explicit, itemized, iterative process that guides investigators to prespecify study design and analysis plans; it addresses a wide range of guidance within a single framework. By supporting the transparent evaluation of causal assumptions and facilitating objective comparisons of design and analysis choices based on prespecified criteria, the Roadmap can help investigators to evaluate the quality of evidence that a given study is likely to produce, specify a study to generate high-quality RWE, and communicate effectively with regulatory agencies and other stakeholders. This paper aims to disseminate and extend the Causal Roadmap framework for use by clinical and translational researchers; three companion papers demonstrate applications of the Causal Roadmap for specific use cases.

A nationwide pharmacovigilance investigation on trends and seriousness of adverse events induced by anti-obesity medication

Introduction

Despite rising concerns regarding the safety of anti-obesity medications, there is a lack of comprehensive pharmacovigilance investigations utilising real-world data. We aimed to characterise the prevalence and seriousness of adverse drug events (ADEs) related to anti-obesity medications and to identify predictors associated with increased risk of serious adverse events (SAE), thereby conveying evidence on drug safety.

Methods

We conducted a cross-sectional analysis on ADE cases spontaneously reported to the Korea Adverse Event Reporting System Database (KIDS-KD). ADE reports pertaining to anti-obesity medications prescribed for overweight, obesity (International Classification of Disease, 10th revision (ICD-10) code E66) and abnormal weight gain (ICD-10 code E63.5) were included in the analysis. We performed a disproportionality to detect the association of the system organ class-based ADEs with their seriousness an individual's sex by estimating reporting odds ratios (RORs) and their 95% confidence intervals (CIs). We performed logistic regression to investigate factors that are substantially associated with increased SAE risks by estimating odds ratio (OR) and their 95% CIs.

Results

The most common causative anti-obesity medication was phentermine, followed by liraglutide. ADEs associated with psychiatric disorders (ROR = 1.734; 95% CI = 1.111-2.707), liver and biliary system disorders (ROR = 22.948; 95% CI = 6.613-70.635), cardiovascular disorders (ROR = 5.707; 95% CI = 1.965-16.574), and respiratory disorders (ROR = 4.567; 95% CI = 1.774-11.762) were more likely to be serious events. Additionally, men are more likely to experience ADEs related gastrointestinal disorders (ROR = 1.411) and less likely to have heart and rhythm disorders (ROR = 0.507). The risk of SAE incidences was positively correlated with being male (OR = 2.196; 95% CI = 1.296-3.721), dual or triple combination of anti-obesity medications (OR = 3.258; 95% CI = 1.633-6.501 and OR = 8.226; 95% CI = 3.046-22.218, respectively), and concomitant administration of fluoxetine (OR = 5.236; 95% CI = 2.218-12.365).

Conclusions

Seriousness of anti-obesity medication-related ADEs differs among system-organ class, while sex-related differences in ADE profiles are also present. The predictors substantially increasing risk of SAE incidences include being male, having a higher number of concomitant medications (including multiple combination of anti-obesity medications), and concurrent use of fluoxetine. Nonetheless, further pharmacovigilance investigation and monitoring are needed to enhance awareness on ADEs induced by anti-obesity medications.

Artificial Intelligence and Machine Learning in Pharmacological Research: Bridging the Gap Between Data and Drug Discovery

Artificial intelligence (AI) has transformed pharmacological research through machine learning, deep learning, and natural language processing. These advancements have greatly influenced drug discovery, development, and precision medicine. AI algorithms analyze vast biomedical data identifying potential drug targets, predicting efficacy, and optimizing lead compounds. AI has diverse applications in pharmacological research, including target identification, drug repurposing, virtual screening, de novo drug design, toxicity prediction, and personalized medicine. AI improves patient selection, trial design, and real-time data analysis in clinical trials, leading to enhanced safety and efficacy outcomes. Post-marketing surveillance utilizes AI-based systems to monitor adverse events, detect drug interactions, and support pharmacovigilance efforts. Machine learning models extract patterns from complex datasets, enabling accurate predictions and informed decision-making, thus accelerating drug discovery. Deep learning, specifically convolutional neural networks (CNN), excels in image analysis, aiding biomarker identification and optimizing drug formulation. Natural language processing facilitates the mining and analysis of scientific literature, unlocking valuable insights and information. However, the adoption of AI in pharmacological research raises ethical considerations. Ensuring data privacy and security, addressing algorithm bias and transparency, obtaining informed consent, and maintaining human oversight in decision-making are crucial ethical concerns. The responsible deployment of AI necessitates robust frameworks and regulations. The future of AI in pharmacological research is promising, with integration with emerging technologies like genomics, proteomics, and metabolomics offering the potential for personalized medicine and targeted therapies. Collaboration among academia, industry, and regulatory bodies is essential for the ethical implementation of AI in drug discovery and development. Continuous research and development in AI techniques and comprehensive training programs will empower scientists and healthcare professionals to fully exploit AI's potential, leading to improved patient outcomes and innovative pharmacological interventions.

Reasons for Discontinuation or Change of Selective Serotonin Reuptake Inhibitors in Online Drug Reviews

Importance

Selective serotonin reuptake inhibitors (SSRIs) are a commonly prescribed medication class to treat a variety of mental disorders. However, adherence to SSRIs is low, and uncovering the reasons for discontinuation among SSRI users is an important first step to improving medication persistence.

Objective

To identify the reasons SSRIs are discontinued or changed, as reported by patients and caregivers in online drug reviews.

Design, Setting, and Participants

This qualitative study used natural language processing and machine learning to extract mentions of changes in SSRI intake from 667 drug reviews posted on the online health forum WebMD from September 1, 2007, to August 31, 2021. The type of medication change, including discontinuation, switch to another medication, or dose change and the reason for the change were manually annotated. In each instance in which an adverse event was reported, the event was categorized using Medical Dictionary for Regulatory Activities primary system organ class (SOC) codes, and its relative frequency was compared with that in spontaneous reporting systems maintained by the US Food and Drug Administration and the UK Medicines and Healthcare Products Regulatory Agency.

Main Outcomes and Measures

Reasons for SSRI medication change as assessed using SOC codes.

Results

In total, 667 reviews posted by 659 patients or caregivers (516 [78%] of patients were female; 410 [62%] 25-54 years of age) were identified that indicated a medication change: 335 posts indicated SSRI discontinuation, 188 posts indicated dose change, and 179 posts indicated switched medications. Most authors 625 (95%) were patients. The most common reason for medication discontinuation or switching was adverse events experienced, and the most common reason for dose change was titration. Both uptitration and downtitration were initiated by either a health care professional or patient. The most common adverse events were classified by SOC codes as psychiatric disorders, including insomnia, loss of libido, and anxiety. Compared with those in regulatory data, psychiatric adverse events, adverse events recorded by investigations (mostly weight gain) and adverse events associated with the reproductive system (mostly erectile dysfunction) were reported disproportionately more often.

Conclusions and Relevance

This qualitative study of online drug reviews found that useful information was provided directly by patients or their caregivers regarding their medication behavior, specifically, information regarding SSRI treatment changes that may inform interventions to improve adherence. These findings suggest that these reported adverse events may be associated with SSRI persistence and that people may feel more inclined to report such events on social media than to clinicians or regulatory agencies.

Recent advances in dosage form design for the elderly: a review

INTRODUCTION

With the increase in the elderly population and the prevalence of multiple medical conditions, medication adherence, and efficacy have become crucial for the effective management of their health. The aging population faces unique challenges that need to be addressed through advancements in drug delivery systems and formulation technologies.

AREAS COVERED

The current review highlights the recent advances in dosage form design for older individuals, with consideration of their specific physiological and cognitive changes. Various dosage forms, such as modified-release tablets/capsules, chewable tablets, and transdermal patches, can be tailored to meet the specific needs of elderly patients. Advancements in drug delivery systems, such as nanotherapeutics, additive manufacturing (three-dimensional printing), and drug-food combinations, improve drug delivery and efficacy and overcome challenges, such as dysphagia and medication adherence.

EXPERT OPINION

Regulatory guidelines and considerations are crucial in ensuring the safe utilization of medications among older adults. Important factors to consider include geriatric-specific guidelines, safety considerations, labeling requirements, clinical trial considerations, and adherence and accessibility considerations.

Severe cutaneous adverse reactions to drugs: A real-world pharmacovigilance study using the FDA Adverse Event Reporting System database

Background: Sound drug safety information is important to optimize patient management, but the widely recognized comprehensive landscape of culprit-drugs that cause severe cutaneous adverse reactions (SCARs) is currently lacking.

Objective: The main aim of the study is to provide a comprehensive landscape of culprit-drugs for SCARs to guide clinical practice.

Methods: We analyzed reports associated with SCARs in the FDA Adverse Event Reporting System database between 1 January 2004 and 31 December 2021 and compiled a list of drugs with potentially serious skin toxicity. According to this list, we summarized the reporting proportions of different drugs and drug classes and conducted disproportionality analysis for all the drugs. In addition, the risk characteristic of SCARs due to different drugs and drug classes was summarized by the positive–negative distribution based on the results of the disproportionality analysis.

Results: A total of 77,789 reports in the FDA Adverse Event Reporting System database were considered SCAR-related, of which lamotrigine (6.2%) was the most reported single drug followed by acetaminophen (5.8%) and allopurinol (5.8%) and antibacterials (20.6%) was the most reported drug class followed by antiepileptics (16.7%) and antineoplastics (11.3%). A total of 1,219 drugs were reported as culprit-drugs causing SCARs in those reports, and the largest number of drugs belonged to antineoplastics. In disproportionality analysis, 776 drugs showed at least one positive pharmacovigilance signal. Drugs with the most positive signals were lamotrigine, acetaminophen, furosemide, and sulfamethoxazole/trimethoprim.

Conclusion: Our study provided a real-world overview of SCARs to drugs, and the investigation of SCAR positive–negative distribution across different drugs revealed its risk characteristics, which may help optimize patient management.

Exploratory pharmacovigilance with machine learning in big patient data: A focused scoping review

BACKGROUND

Machine learning can operationalize the rich and complex data in electronic patient records for exploratory pharmacovigilance endeavours.

OBJECTIVE

The objective of this review is to identify applications of machine learning and big patient data in exploratory pharmacovigilance.

METHODS

We searched PubMed and Embase and included original articles with an exploratory pharmacovigilance purpose, focusing on medicinal interventions and reporting the use of machine learning in electronic patient records with ≥1000 patients collected after market entry.

FINDINGS

Of 2557 studies screened, seven were included. Those covered six countries and were published between 2015 and 2021. The most prominent machine learning methods were random forests, logistic regressions, and support vector machines. Two studies used artificial neural networks or naive Bayes classifiers. One study used formal concept analysis for association mining, and another used temporal difference learning. Five studies compared several methods against each other. The numbers of patients in most data sets were in the order of thousands; two studies used what can more reasonably be considered big data with >1 000 000 patients records.

CONCLUSION

Despite years of great aspirations for combining machine learning and clinical data for exploratory pharmacovigilance, only few studies still seem to deliver somewhat on these expectations.

The evolution of pharmacovigilance ecosystems: Does Moore's law invite the use of Occam's razor?

AIMS

Moore's law predicts the doubling of complexity of integrated circuits every 2 years; Kryder's corollary assumes a doubling of data storage every 13 months. With the increasing volume of legislation, pharmacovigilance systems today are inherently complex, and the emphasis has shifted from reactive (responding to emerging risks) to planned, active, risk-proportionate approaches operating throughout the life cycle of medicines.

METHODS

Exploration of the drivers for increasing complexity of pharmacovigilance systems, focusing on regulatory environment, data management and evaluation.

RESULTS

Evaluation of postmarketing data plays an increasingly important role in pharmacovigilance. There is great interest on the part of all stakeholders in optimizing the use of these data. Innovative approaches, including pharmacogenetics and passive measures (sensors), will lead to increased complexity and volumes of data and inevitably to an increase in the volume of case reports. There is a multiplicity of regulations and guidelines on how to manage these data, with an inherent lack of harmonization.

CONCLUSION

We summarize the current characterization of safety data types, sources and the classification of these data. Using this benchmark, we discuss the future requirements of an effective pharmacovigilance ecosystem, keeping the principle of parsimony in mind. In this complex, continuously and rapidly changing environment, there is a need for a return to simplicity and pragmatism. The application of Occam's razor could help to support the rapid provision of new, affordable medicines with a positive benefit to risk profile.

Evaluation of Risk of Cardiac or Cerebrovascular Events in Romosozumab Users Focusing on Comorbidities: Analysis of the Japanese Adverse Drug Event Report Database

Background: Romosozumab is associated with an increased risk of cardiac or cerebrovascular events. Identifying the risk factors for these events could contribute to the safe use of romosozumab. Objective: This study aimed to investigate risk factors for cardiac or cerebrovascular events in romosozumab users. Methods: First, disproportionality analysis was performed to compare the frequency of cardiac or cerebrovascular events, using data from the Japanese Adverse Drug Event Report database. Next, multivariate logistic analysis was performed to investigate risk factors for cardiac or cerebrovascular events in romosozumab users. Results: In total, 859 romosozumab users were identified. A disproportionality of both cardiac and cerebrovascular events was observed in only romosozumab users. Multivariate logistic analysis revealed that the risk of cardiac events in romosozumab users was significantly increased in patients with cardiac disease (odds ratio [OR]: 5.9, 95% confidence interval [CI] 3.5-9.9; P < 0.01) and hypertension (OR: 1.6, 95% CI 1.0-2.7; P = 0.047). In addition, the risk of cerebrovascular events in romosozumab users was significantly increased in the presence of cerebrovascular disease (OR: 2.7, 95% CI 1.2-6.2; P = 0.02) and hypertension (OR: 2.6, 95% CI 1.7-3.9; P < 0.01). Conclusion: Our findings suggest that hypertension may increase the risk of cardiac or cerebrovascular events in romosozumab users. Although additional studies are needed to assess other associated factors, these findings may contribute to the appropriate use of romosozumab and limit adverse events.

Pharmacovigilance in China: Evolution and future challenges

Drug-related adverse reactions are among the main reasons for harm to patients under care worldwide and even their deaths. The pharmacovigilance system has been proven to be an effective method of avoiding or alleviating such adverse events. In 2019, after two decades of implementation of the drug-related adverse reaction reporting system, China formally implemented a pharmacovigilance system with the Pharmacovigilance Quality Management Standards and a series of supporting technical documents created to improve the safety of medication given to patients. China's pharmacovigilance system has faced many problems and challenges during its implementation. This spontaneous reporting system is the main source of data for China's medication vigilance activities, but it has not provided sufficiently powerful evidence for regulatory decision-making. In conformity with the health-centred drug regulatory concept, the Chinese government has accelerated the speed of examination and approval of urgently needed clinical drugs and orphan drugs along with the requirement to improve the safety supervision of these drugs after their listing. China's marketing authorization holders (MAHs) must strengthen their pharmacovigilance capabilities as the primary responsible departments for drug safety. Chinese medical schools generally lack professional courses on pharmacovigilance. The regulatory authorities have recognized such problems and have made efforts to improve the professional capacity of pharmacovigilance personnel and to strengthen cooperation with stakeholders through the implementation of an action plan of medication surveillance and the establishment of a patient-based adverse events reporting system and active surveillance systems, which will help China bridge the gap to bring its pharmacovigilance practice up to standards.

A bibliometric analysis of research progress on pharmacovigilance and cancer from 2002 to 2021

The complexity of cancer itself and treatment makes pharmacovigilance critical in oncology. Despite rapid progress on pharmacovigilance and cancer research in the past two decades, there has been no bibliometric analysis in this field. Therefore, based on the Web of Science database, we used CiteSpace, VOS-viewer and R-bibliometrix to analyze and visualize publications, and described the development trend and research hot spots in this field. 502 publications were included. The development of pharmacovigilance and cancer research has continued to grow. The USA has the largest number of publications and citations, followed by France and UK. Vanderbilt University and Sorbonne University are the institutions that contribute the most papers, and 5 of the top 10 high-yield institutions are from France. Salem JE and Lebrun-Vignes B of Sorbonne University have published the most papers, and they have a strong cooperative relationship. Salem JE has the highest H index. Drug Safety has the largest number of publications in the field of pharmacovigilance and cancer, with a high impact factor (IF). In recent years, immune checkpoint inhibitors (ICIs) have been identified as a hot topic and will continue to be maintained. This paper can help researchers get familiar with the current situation and trend of pharmacovigilance and cancer research, and provide valuable reference for the selection of future research directions.

An overview of methodological flaws of real-world studies investigating drug safety in the post-marketing setting

INTRODUCTION

The evaluation of the post-marketing safety profile of drugs is a continuous monitoring process for approved and marketed medicines and it is crucial for detecting new adverse drug reactions. As such, real-world studies are essential to complement pre-marketing evidence with information concerning drug risk-benefit profile and use in wider patient populations and they have a great potential to support post-marketing drug safety evaluations.

AREAS COVERED

A detailed description of the main limitations of real-world data sources (i.e. claims databases, electronic healthcare records, drug/disease registers and spontaneous reporting system databases) and of the main methodological challenges of real-world studies in generating real-world evidence is provided.

EXPERT OPINION

Real-world evidence biases can be ascribed to both the methodological approach and the specific limitations of the different real-world data sources used to carry out the study. As such, it is crucial to characterize the quality of real-world data, by establishing guidelines and best practices for the assessment of data fitness for purpose. On the other hand, it is important that real-world studies are conducted using a rigorous methodology, aimed at minimizing the risk of bias.

Developing a deep learning natural language processing algorithm for automated reporting of adverse drug reactions

The detection of adverse drug reactions (ADRs) is critical to our understanding of the safety and risk-benefit profile of medications. With an incidence that has not changed over the last 30 years, ADRs are a significant source of patient morbidity, responsible for 5%-10% of acute care hospital admissions worldwide. Spontaneous reporting of ADRs has long been the standard method of reporting, however this approach is known to have high rates of under-reporting, a problem that limits pharmacovigilance efforts. Automated ADR reporting presents an alternative pathway to increase reporting rates, although this may be limited by over-reporting of other drug-related adverse events. We developed a deep learning natural language processing algorithm to identify ADRs in discharge summaries at a single academic hospital centre. Our model was developed in two stages: first, a pre-trained model (DeBERTa) was further pre-trained on 1.1 million unlabelled clinical documents; secondly, this model was fine-tuned to detect ADR mentions in a corpus of 861 annotated discharge summaries. This model was compared to a version without the pre-training step, and a previously published RoBERTa model pretrained on MIMIC III, which has demonstrated strong performance on other pharmacovigilance tasks. To ensure that our algorithm could differentiate ADRs from other drug-related adverse events, the annotated corpus was enriched for both validated ADR reports and confounding drug-related adverse events using. The final model demonstrated good performance with a ROC-AUC of 0.955 (95% CI 0.933 - 0.978) for the task of identifying discharge summaries containing ADR mentions, significantly outperforming the two comparator models.

A Rule-Based Inference Framework to Explore and Explain the Biological Related Mechanisms of Potential Drug-Drug Interactions

As more drugs are developed and the incidence of polypharmacy increases, it is becoming critically important to anticipate potential DDIs before they occur in the clinic, along with those for which effects might go unobserved. However, traditional methods for DDI identification are unable to coalesce interaction mechanisms out of vast lists of potential or known DDIs, much less study them accurately. Computational methods have great promise but have realized only limited clinical utility. This work develops a rule-based inference framework to predict DDI mechanisms and support determination of their clinical relevance. Given a drug pair, our framework interrogates and describes DDI mechanisms based on a knowledge graph that integrates extensive available biomedical resources through semantic web technologies and backward chaining inference, effectively identifying facts within the graph that prove and explain the mechanisms of the drugs' interaction. The framework was evaluated through a case study combining a chemotherapy agent, irinotecan, and a widely used antibiotic, levofloxacin. The mutual interactions identified indicate that our framework can effectively explore and explain the mechanisms of potential DDIs. This approach has the potential to improve drug discovery and design and to support rapid and cost-effective identification of DDIs along with their putative mechanisms, a key step in determining clinical relevance and supporting clinical decision-making.

A smart hospital-driven approach to precision pharmacovigilance

Researchers, regulatory agencies, and the pharmaceutical industry are moving towards precision pharmacovigilance as a comprehensive framework for drug safety assessment, at the service of the individual patient, by clustering specific risk groups in different databases. This article explores its implementation by focusing on: (i) designing a new data collection infrastructure, (ii) exploring new computational methods suitable for drug safety data, and (iii) providing a computer-aided framework for distributed clinical decisions with the aim of compiling a personalized information leaflet with specific reference to a drug's risks and adverse drug reactions. These goals can be achieved by using 'smart hospitals' as the principal data sources and by employing methods of precision medicine and medical statistics to supplement current public health decisions.

Drug-Drug Interaction of the Sodium Glucose Co-Transporter 2 Inhibitors with Statins and Myopathy: A Disproportionality Analysis Using Adverse Events Reporting Data

INTRODUCTION

An increased risk of myopathy due to a potential interaction between sodium glucose co-transporter-2 inhibitors (SGLT-2i) and HMG-CoA reductase inhibitors (statins) has been suggested by case reports.

OBJECTIVE

We aimed to assess if the reporting of myopathy is disproportionally higher among people using both SGLT-2i and statins compared to using either SGLT-2i or statins alone.

METHODS

We conducted a disproportionality analysis using data from the US Food and Drug Administration Adverse Event Reporting System (FAERS). We included reports with at least one antihyperglycemic agent. We compared the proportion of myopathy cases to non-cases between those not using SGLT-2i or statins, using SGLT-2i only, statins only, or both. We calculated the reporting odds ratio and 95% confidence interval. We further stratified by individual SGLT-2i and selected statins (rosuvastatin or atorvastatin).

RESULTS

We included 688,388 reports with at least one antihyperglycemic agent recorded, of which 9.80% had at least one SGLT-2i agent. Among all included reports, there were a total of 2202 myopathy cases with the majority, 61.26%, occurring among those using statins alone and only 2.72% of myopathy cases were among those using both SGLT-2i and statins together. Reporting of myopathy was not disproportionally higher among those reporting the use of SGLT-2i with statins (reporting odds ratio 2.95, 95% confidence interval 2.27-3.85) compared to statins alone (reporting odds ratio 6.41, 95% confidence interval 5.86-7.02).

CONCLUSIONS

Reports of myopathy were not disproportionally higher among those using SGLT-2i with statins compared to SGLT-2i or statins alone at the class level. Further observational studies may be needed to better assess this interaction at the agent level.

The value of real world evidence: The case of medical cannabis

Randomised controlled trials (RCTs) have long been considered the gold standard of medical evidence. In relation to cannabis based medicinal products (CBMPs), this focus on RCTs has led to very restrictive guidelines in the UK, which are limiting patient access. There is general agreement that RCT evidence in relation to CBPMs is insufficient at present. As well as commercial reasons, a major problem is that RCTs do not lend themselves well to the study of whole plant medicines. One solution to this challenge is the use of real world evidence (RWE) with patient reported outcomes (PROs) to widen the evidence base. Such data increasingly highlights the positive impact medical cannabis can have on patients' lives. This paper outlines the value of this approach which involves the study of interventions and patients longitudinally under medical care. In relation to CBMPs, RWE has a broad range of advantages. These include the study of larger groups of patients, the use of a broader range and ratio of components of CBMPs, and the inclusion of more and rarer medical conditions. Importantly, and in contrast to RCTs, patients with significant comorbidities-and from a wider demographic profile-can also be studied, so providing higher ecological validity and increasing patient numbers, whilst offering significant cost savings. We conclude by outlining 12 key recommendations of the value of RWE in relation to medical cannabis. We hope that this paper will help policymakers and prescribers understand the importance of RWE in relation to medical cannabis and help them develop approaches to overcome the current situation which is detrimental to patients.

A contextual multi-task neural approach to medication and adverse events identification from clinical text

Effective wide-scale pharmacovigilance calls for accurate named entity recognition (NER) of medication entities such as drugs, dosages, reasons, and adverse drug events (ADE) from clinical text. The scarcity of adverse event annotations and underlying semantic ambiguities make accurate scope identification challenging. The current research explores integrating contextualized language models and multi-task learning from diverse clinical NER datasets to mitigate this challenge. We propose a novel multi-task adaptation method to refine the embeddings generated by the Bidirectional Encoder Representations from Transformers (BERT) language model to improve inter-task knowledge sharing. We integrated the adapted BERT model into a unique hierarchical multi-task neural network comprised of the medication and auxiliary clinical NER tasks. We validated the model using two different versions of BERT on diverse well-studied clinical tasks: Medication and ADE (n2c2 2018/n2c2 2009), Clinical Concepts (n2c2 2010/n2c2 2012), Disorders (ShAReCLEF 2013). Overall medication extraction performance enhanced by up to +1.19 F1 (n2c2 2018) while generalization enhanced by +5.38 F1 (n2c2 2009) as compared to standalone BERT baselines. ADE recognition enhanced significantly (McNemar's test), out-performing prior baselines. Similar benefits were observed on the auxiliary clinical and disorder tasks. We demonstrate that combining multi-dataset BERT adaptation and multi-task learning out-performs prior medication extraction methods without requiring additional features, newer training data, or ensembling. Taken together, the study contributes an initial case study towards integrating diverse clinical datasets in an end-to-end NER model for clinical decision support.

Quantifying the Severity of Adverse Drug Reactions Using Social Media: Network Analysis

BACKGROUND

Adverse drug reactions (ADRs) affect the health of hundreds of thousands of individuals annually in the United States, with associated costs of hundreds of billions of dollars. The monitoring and analysis of the severity of ADRs is limited by the current qualitative and categorical systems of severity classification. Previous efforts have generated quantitative estimates for a subset of ADRs but were limited in scope because of the time and costs associated with the efforts.

OBJECTIVE

The aim of this study is to increase the number of ADRs for which there are quantitative severity estimates while improving the quality of these severity estimates.

METHODS

We present a semisupervised approach that estimates ADR severity by using social media word embeddings to construct a lexical network of ADRs and perform label propagation. We used this method to estimate the severity of 28,113 ADRs, representing 12,198 unique ADR concepts from the Medical Dictionary for Regulatory Activities.

RESULTS

Our Severity of Adverse Events Derived from Reddit (SAEDR) scores have good correlations with real-world outcomes. The SAEDR scores had Spearman correlations of 0.595, 0.633, and -0.748 for death, serious outcome, and no outcome, respectively, with ADR case outcomes in the Food and Drug Administration Adverse Event Reporting System. We investigated different methods for defining initial seed term sets and evaluated their impact on the severity estimates. We analyzed severity distributions for ADRs based on their appearance in boxed warning drug label sections, as well as for ADRs with sex-specific associations. We found that ADRs discovered in the postmarketing period had significantly greater severity than those discovered during the clinical trial (P<.001). We created quantitative drug-risk profile (DRIP) scores for 968 drugs that had a Spearman correlation of 0.377 with drugs ranked by the Food and Drug Administration Adverse Event Reporting System cases resulting in death, where the given drug was the primary suspect.

CONCLUSIONS

Our SAEDR and DRIP scores are well correlated with the real-world outcomes of the entities they represent and have demonstrated utility in pharmacovigilance research. We make the SAEDR scores for 12,198 ADRs and the DRIP scores for 968 drugs publicly available to enable more quantitative analysis of pharmacovigilance data.