Perspectives on the use of data mining in pharmaco-vigilance

Antineoplastic agents associated with neutropenic enterocolitis in patients with malignancy: A quantitative safety signal analysis

BACKGROUND

The use of certain chemotherapy agents is associated with the development of a condition called "chemotherapy-associated neutropenic enterocolitis" (CANE).

OBJECTIVE

To determine the risk of CANE associated with the use of each antineoplastic agent.

METHODS

The FDA FAERS database of spontaneous adverse reactions was searched for the occurrence of the MedDRA preferred term "neutropenic colitis."

RESULTS

The search resulted in 1134 records of patients (535 [47.3%] females, 479 [42.2%] males, sex not specified in 120 [10.6%]) with neutropenic colitis receiving immunosuppressive chemotherapy. The mean age of patients was 47 (SD 22). 22 antineoplastic agents were found to have a strong association (reported odds ratio [ROR] > 100) with the occurrence of CANE; 9 had ROR < 2.

CONCLUSION

Drug databases have several limitations in providing updated information about newly approved pharmaceutical adverse events. Signal detection is a diagnostic method recognized as practical in pharmacovigilance. It may be utilized in the FDA's adverse event reporting database and has demonstrated a reasonable predictive performance in signaling adverse events. Our study emphasized the substantial knowledge gap between what we know about the potential risk of CANE caused by antineoplastic agents and the reports of the FDA on their new approved products.

New exploration of signal detection of Regional Risks from the perspective of data mining: a pharmacovigilance analysis based on spontaneous reporting data in Zhenjiang, China

BACKGROUND

This study aimed to adopt the conventional signal detection methods to explore a new way of risk identification and to mine important drug risks from the perspective of big data based on Zhenjiang Adverse Event Reporting System (ZAERS).

RESEARCH DESIGN AND METHODS

Data were extracted from ZAERS database between 2012 and 2022. The risks of all the reported drug event combinations were identified at the preferred term level and the standardized MedDRA query level using disproportionality analysis. Then, we conducted signal assessment according to the descriptions of drug labels.

RESULTS

In total 41,473 ADE were reported and there were 12 risky signals. Signal assessment indicates the suspected causal associations in clindamycin-taste and smell disorders, valsartan-hepatic enzyme increased and valsartan-edema peripheral; the specific manifestations of allergic reactions triggered by clindamycin, cefotaxime, cefazodime, ShexiangZhuanggu plaster, ShexiangZhuifeng plaster, and Yanhuning need to be refined in drug labels. In addition, the drug labels of NiuHuangShangQing tablet/capsule, Fuyanxiao capsule, and BiYanLing tablet should be improved.

CONCLUSIONS

In this study, we attempted a new way to find potential drug risks using small spontaneous reporting data. Our findings also suggested the need for more precise identification of allergic risks and the improvement of traditional Chinese medicine labels.

Urinary tract infections and genital mycotic infections associated with SGLT‑2 inhibitors: an analysis of the FDA Adverse Event Reporting System

BACKGROUND

Sodium-glucose cotransporter-2 (SGLT-2) inhibitors are a class of antihyperglycemic agents, including canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin. Risk of urinary tract infections (UTIs) and genital mycotic infections (GMIs) associated with SGLT‑2 inhibitors is of great clinical significance. The study aimed to assess the association between SGLT-2 inhibitors and occurrences of UTIs and GMIs using the FDA Adverse Event Reporting System (FAERS) database.

METHODS

We used OpenVigil 2.1-MedDRA-v24 to query the FAERS database. Disproportionality analysis was performed to detect adverse event signals. Reporting Odds Ratio (ROR) and Proportional Reporting Ratio (PRR) were calculated to measure the disproportionality.

RESULTS

A total of 45,256 reports related to the use of SGLT-2 inhibitors, including 1,714 UTI cases and 438 GMI cases, were retrieved. Potential positive signals for UTIs and GMIs were identified for canagliflozin, dapagliflozin, empagliflozin, and ertugliflozin in adult patients of all ages and both sexes.

CONCLUSIONS

Data mining in the FAERS database suggests strong association between SGLT-2 inhibitors and UTIs/GMIs. These findings provide real-world evidence on the potential risk of UTIs/GMIs related to SGLT-2 inhibitors.

Myocarditis and Pericarditis Post-mRNA COVID-19 Vaccination: Insights from a Pharmacovigilance Perspective

Concerns remain regarding the rare cardiovascular adverse events, myocarditis and pericarditis (myo/pericarditis), particularly in younger individuals following mRNA COVID-19 vaccination. Our study aimed to comprehensively assess potential safety signals related to these cardiac events following the primary and booster doses, with a specific focus on younger populations, including children as young as 6 months of age. Using the Vaccine Adverse Events Reporting System (VAERS), the United States national passive surveillance system, we conducted a retrospective pharmacovigilance study analyzing spontaneous reports of myo/pericarditis. We employed both frequentist and Bayesian methods and conducted subgroup analyses by age, sex, and vaccine dose. We observed a higher reporting rate of myo/pericarditis following the primary vaccine series, particularly in males and mainly after the second dose. However, booster doses demonstrated a lower number of reported cases, with no significant signals detected after the fourth or fifth doses. In children and young adults, we observed notable age and sex differences in the reporting of myo/pericarditis cases. Males in the 12-17 and 18-24-year-old age groups had the highest number of cases, with significant signals for both males and females after the second dose. We also identified an increased reporting for a spectrum of cardiovascular symptoms such as chest pain and dyspnea, which increased with age, and were reported more frequently than myo/pericarditis. The present study identified signals of myo/pericarditis and related cardiovascular symptoms after mRNA COVID-19 vaccination, especially among children and adolescents. These findings underline the importance for continued vaccine surveillance and the need for further studies to confirm these results and to determine their clinical implications in public health decision-making, especially for younger populations.

Thromboembolism profiles associated with cyclin-dependent kinase 4/6 inhibitors: a real-world pharmacovigilance study and a systematic review

BACKGROUND

Thrombosis is the second leading cause of mortality in cancer patients. This study aimed to investigate the association between cyclin-dependent kinase 4 and 6 inhibitors (CDK4/6i) and thrombosis.

RESEARCH DESIGN AND METHODS

A retrospective pharmacovigilance analysis based on real-world data combined with a systematic review was used to explore the thrombotic risk profiles of CDK4/6i. The study has been registered with Prospero (CRD42021284218).

RESULT

In the pharmacovigilance analysis, CDK4/6i showed a higher rate of reported venous thromboembolism (VTE) (ROR = 2.78, 95% CI = 2.64-2.92), with the highest signal for trilaciclib (ROR = 27.55, 95% CI = 13.43-56.52) but only 9 cases, followed by abemaciclib (ROR = 3.73, 95% CI = 3.19-4.37). For arterial thromboembolism (ATE), only ribociclib increased the reporting rate (ROR = 2.14, 95% CI = 1.91-2.41). In the meta-analysis, palbociclib, abemaciclib, and trilaciclib all increased the risk of VTE (OR = 2.23, 3.17, and 3.90). In the subgroup analysis, only abemaciclib increased the risk of ATE (OR = 2.11, 95% CI = 1.12-3.99)    .

CONCLUSIONS

CDK4/6i had different profiles of thromboembolism. Palbociclib, abemaciclib, or trilaciclib increased the risk of VTE. Ribociclib and abemaciclib showed a weak association with the risk of ATE.

Prescribed Drugs and Self-Directed Violence: A Descriptive Study in the Spanish Pharmacovigilance Database

Self-inflicted violence is a major and growing public health problem and its prediction and prevention is challenging for healthcare systems worldwide. Our aim was to identify prescribed drugs associated with self-directed violent behaviors in Spain. A descriptive, longitudinal and retrospective study of spontaneous reports of adverse drug reactions corresponding to self-directed violence was recorded in the Spanish Pharmacovigilance Database (FEDRA^®) from 1984 to 31 March 2021. A total of 710 cases were reported in the study period. The mean age was 45.52 years (range 1-94). There were no gender differences except in children, where most reports were of male children. The main therapeutic groups that were involved included drugs for the nervous system (64.5%) and anti-infectives for systemic use (13.2%). The most commonly reported drugs were varenicline, fluoxetine, lorazepam, escitalopram, venlafaxine, veralipride, pregabalin, roflumilast and bupropion. There were reports of montelukast, hydroxychloroquine, isotretinoin, methylphenidate, infliximab, natalizumab, ribavirin and efavirenz, which were less known to be involved in self-directed violence. This study shows that self-directed violence is a rare adverse drug reaction, and can be related to the use of some medicines. It is important for healthcare professionals to consider this risk in their clinical praxis, implementing person-centred approaches. Further studies are needed, considering comorbidities and potential interactions.

Serious adverse reaction associated with the COVID-19 vaccines of BNT162b2, Ad26.COV2.S, and mRNA-1273: Gaining insight through the VAERS

Background and purpose: Serious adverse events following immunization (AEFI) associated with the COVID-19 vaccines, including BNT162b2 (Pfizer-BioNTech), Ad26.COV2.S (Janssen), and mRNA-1273 (Moderna), have not yet been fully investigated. This study was designed to evaluate the serious AEFI associated with these three vaccines. Methods: A disproportionality study was performed to analyze data acquired from the Vaccine Adverse Event-Reporting System (VAERS) between 1 January 2010 and 30 April 2021. The reporting odds ratio (ROR) method was used to identify the association between the COVID-19 vaccines BNT162b2, Ad26.COV2.S, and mRNA-1273 and each adverse event reported. Moreover, the ratio of the ROR value to the 95% CI span was applied to improve the credibility of the ROR. The median values of time from vaccination to onset (TTO) for the three vaccines were analyzed. Results: Compared with BNT162b2 and mRNA-1273, Ad26.COV2.S vaccination was associated with a lower death frequency (p < 0.05). Ad26.COV2.S vaccination was associated with a lower birth defect and emergency room visit frequency than BNT162b2 (p < 0.05). There were 6,605, 830, and 2,292 vaccine recipients who suffered from COVID-19-related symptoms after vaccination with BNT162b2, Ad26.COV2.S, and mRNA-1273, respectively, including people who were infected by COVID-19, demonstrated a positive SARS-CoV-2 test, and were asymptomatic. Serious AEFI, including thromboembolism, hemorrhage, thrombocytopenia, cardiac arrhythmia, hypertension, and hepatotoxicity, were associated with all three vaccines. Cardiac failure and acute renal impairment events were associated with BNT162b2 and mRNA-1273, while seizure events were associated with BNT162b2 and Ad26.COV2.S. The median values of TTO associated with the three vaccinations were similar. Conclusion: These findings may be useful for health workers and the general public prior to inoculation, especially for patients with underlying diseases; however, the risk/benefit profile of these vaccines remains unchanged. The exact mechanism of SARS-CoV-2 vaccine-induced AEFI remains unknown, and further studies are required to explore these phenomena.

Signaling COVID-19 Vaccine Adverse Events

Introduction

Statistical signal detection is a crucial tool for rapidly identifying potential risks associated with pharmaceutical products. The unprecedented environment created by the coronavirus disease 2019 (COVID-19) pandemic for vaccine surveillance predisposes commonly applied signal detection methodologies to a statistical issue called the masking effect, in which signals for a vaccine of interest are hidden by the presence of other reported vaccines. This masking effect may in turn limit or delay our understanding of the risks associated with new and established vaccines.

Objective

The aim is to investigate the problem of masking in the context of COVID-19 vaccine signal detection, assessing its impact, extent, and root causes.

Methods

Based on data underlying the Vaccine Adverse Event Reporting System, three commonly applied statistical signal detection methodologies, and a more advanced regression-based methodology, we investigate the temporal evolution of signals corresponding to five largely recognized adverse events and two potentially new adverse events.

Results

The results demonstrate that signals of adverse events related to COVID-19 vaccines may be undetected or delayed due to masking when generated by methodologies currently utilized by pharmacovigilance organizations, and that a class of advanced methodologies can partially alleviate the problem. The results indicate that while masking is rare relative to all possible statistical associations, it is much more likely to occur in COVID-19 vaccine signaling, and that its extent, direction, impact, and roots are not static, but rather changing in accordance with the changing nature of data.

Conclusions

Masking is an addressable problem that merits careful consideration, especially in situations such as COVID-19 vaccine safety surveillance and other emergency use authorization products.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40264-022-01186-z.

Social Media Analytics for Pharmacovigilance of Antiepileptic Drugs

Epilepsy is a common neurological disorder worldwide and antiepileptic drug (AED) therapy is the cornerstone of its treatment. It has a laudable aim of achieving seizure freedom with minimal, if any, adverse drug reactions (ADRs). Too often, AED treatment is a long-lasting journey, in which ADRs have a crucial role in its administration. Therefore, from a pharmacovigilance perspective, detecting the ADRs of AEDs is a task of utmost importance. Typically, this task is accomplished by analyzing relevant data from spontaneous reporting systems. Despite their wide adoption for pharmacovigilance activities, the passiveness and high underreporting ratio associated with spontaneous reporting systems have encouraged the consideration of other data sources such as electronic health databases and pharmaceutical databases. Social media is the most recent alternative data source with many promising potentials to overcome the shortcomings of traditional data sources. Although in the literature some attempts have investigated the validity and utility of social media for ADR detection of different groups of drugs, none of them was dedicated to the ADRs of AEDs. Hence, this paper presents a novel investigation of the validity and utility of social media as an alternative data source for the detection of AED ADRs. To this end, a dataset of consumer reviews from two online health communities has been collected. The dataset is preprocessed; the unigram, bigram, and trigram are generated; and the ADRs of each AED are extracted with the aid of consumer health vocabulary and ADR lexicon. Three widely used measures, namely, proportional reporting ratio, reporting odds ratio, and information component, are used to measure the association between each ADR and AED. The resulting list of signaled ADRs for each AED is validated against a widely used ADR database, called Side Effect Resource, in terms of the precision of ADR detection. The validation results indicate the validity of online health community data for the detection of AED ADRs. Furthermore, the lists of signaled AED ADRs are analyzed to answer questions related to the common ADRs of AEDs and the similarities between AEDs in terms of their signaled ADRs. The consistency of the drawn answers with the existing pharmaceutical knowledge suggests the utility of the data from online health communities for AED-related knowledge discovery tasks.

Hepatotoxicity-Related Adverse Effects of Proton Pump Inhibitors: A Cross-Sectional Study of Signal Mining and Analysis of the FDA Adverse Event Report System Database

Background and Objectives: Mounting evidence demonstrates that proton pump inhibitors (PPIs) are associated with a number of adverse effects. However, the literatures about hepatotoxicity-related adverse effects (HRAEs) of PPIs are mostly case reports and a few clinical studies.

Methods: We evaluated the association between PPIs and HAREs using the reporting odd ratio (ROR) for mining the adverse event report signals in the FDA Adverse Event Reporting System (FAERS) database.

Results: There were 23,825 reports of PPIs as primary suspect drug or second suspect drug, of which 3,253 reports were HRAEs. The top five HRAE signals caused by PPIs were hepatitis cholestatic, cholestasis, fulminant hepatitis, subacute hepatic failure, and acute hepatitis. We also summarized the signals of the HRAEs caused by each PPI. The simultaneous signals were cholestasis and hepatitis cholestatic. For the cholestasis signal, esomeprazole showed an ROR of 21.556 (95% CI 17.592–26.413); pantoprazole showed the highest ROR of 22.611 (95% CI 17.794–28.733) in the hepatic cholestatic signal; lansoprazole was the only PPI with expression in the coma hepatic signal, with an ROR of 10.424 (95% CI 3.340–32.532). By analyzing the reports of pantoprazole-induced hepatic encephalopathy, we found that patients aged over 65 years and males reported the highest rate. And from the combination of drugs and indications of drugs, no significant results were obtained.

Conclusions: The RORs of signals of “cholestasis” were generally higher than those of “hepatocellular injury.” And the signals about “cholestasis” in HRAE caused by PPIs are more reported.

New adaptive lasso approaches for variable selection in automated pharmacovigilance signal detection

Background

Adverse effects of drugs are often identified after market introduction. Post-marketing pharmacovigilance aims to detect them as early as possible and relies on spontaneous reporting systems collecting suspicious cases. Signal detection tools have been developed to mine these large databases and counts of reports are analysed with disproportionality methods. To address disproportionality method biases, recent methods apply to individual observations taking into account all exposures for the same patient. In particular, the logistic lasso provides an efficient variable selection framework, yet the choice of the regularization parameter is a challenging issue and the lasso variable selection may give inconsistent results.

Methods

We propose a new signal detection methodology based on the adaptive lasso. We derived two new adaptive weights from (i) a lasso regression using the Bayesian Information Criterion (BIC), and (ii) the class-imbalanced subsampling lasso (CISL), an extension of stability selection. The BIC is used in the adaptive lasso stage for variable selection. We performed an extensive simulation study and an application to real data, where we compared our methods to the existing adaptive lasso, and recent detection approaches based on lasso regression or propensity scores in high dimension. For both studies, we evaluate the methods in terms of false discoveries and sensitivity.

Results

In the simulations and the application, both proposed adaptive weights show equivalent or better performances than the other competitors, with an advantage for the CISL-based adaptive weights. CISL and lasso regression using BIC are solid alternatives.

Conclusion

Our proposed adaptive lasso is an appealing methodology for signal detection in pharmacovigilance. Although we cannot rely on test theory, our approaches show a low and stable False Discovery Rate in all simulation settings. All methods evaluated in this work are implemented in the adapt4pv R package.

Supplementary Information

The online version contains supplementary material available at (10.1186/s12874-021-01450-3).

Post-Marketing Safety Concerns with Esketamine: A Disproportionality Analysis of Spontaneous Reports Submitted to the FDA Adverse Event Reporting System

INTRODUCTION

Esketamine nasal spray received approval for treatment-resistant depression in March 2019.

OBJECTIVE

Using the FDA Adverse Event Reporting System (FAERS) database (March 2019-March 2020), we analysed esketamine-related adverse events (AEs) to detect and characterize relevant safety signals.

METHODS

We used the consolidated case/non-case approach to estimate the reporting odds ratio (ROR) and information component (IC) with relevant confidence intervals (95% CI) for esketamine-related AEs with ≥4 counts. Comparisons between serious and non-serious AEs were performed using non-parametric tests.

RESULTS

The FAERS database contained 962 cases of esketamine-related AEs, with signals detected for several AEs, such as dissociation (ROR = 1,612.64, 95% CI = 1,354.63, 1,919.79; IC = 8.19, 95% CI = 7.96, 8.35), sedation (ROR = 238.46, 95% CI = 202.98, 280.15; IC = 7, 95% CI = 6.75, 7.18), feeling drunk (ROR = 96.17, 95% CI = 61.42, 150.57; IC = 4.84, 95% CI = 4.09, 5.36), suicidal ideation (ROR = 24.03, 95% CI = 18.72, 30.84; IC = 4.31, 95% CI = 3.9, 4.61), and completed suicide (ROR = 5.75, 95% CI = 3.18, 10.41; IC = 2.25, 95% CI = 1.23, 2.94). Signals for suicidal and self-injurious ideation, but not suicide attempt and completed suicide, remained when comparing esketamine to venlafaxine. Females and patients receiving antidepressant polypharmacy, co-medication with mood stabilizers, antipsychotics, benzodiazepines, or somatic medications were more likely to suffer from serious versus non-serious AEs (χ2 = 125.29, p < 0.001, χ2 = 9.08, p = 0.003, χ2 = 8.14, p = 0.004, χ2 = 19.48, p < 0.001, χ2 = 25.62, p < 0.001, and χ2 = 16.79, p < 0.001, respectively).

CONCLUSIONS

Esketamine may carry a clear potential for serious AEs, which deserves urgent clarification by means of further prospective studies.

Pharmacovigilance Bibliometrics: Visualizing Thematic Development in the Category of Pharmacology and Pharmacy in Web of Science

Introduction: Pharmacovigilance studies include monitoring and preventing the occurrence of new, rare, or serious adverse drug reactions, making it possible to discover new safety issues without delay. Bibliometrics could assist scholars to analyze the development of pharmacovigilance.

Methods: The MeSH terms of both pharmacovigilance and “adverse drug reaction reporting system” were retrieved in the Science Citation Index Expanded. The articles from 1974 to July 2021 in the pharmacology and pharmacy category were recruited. The citation reports including the publication numbers, h-index, and sum and average cited times in terms of annuals, countries, organizations, authors and journals were tabulated. The coauthorship relations in the analysis units of countries, organizations, and authors; the top 10 burst references; the document citation network; and the author’s keywords co-occurrence overlay map were visualized by bibliometric software including the website (https://bibliometric.com/), VOSviewer, CiteSpace, and CitNetExplorer.

Results: From 1974 to the present, the most high-yield publication year, country, institute, author, and journal were 2020 (n = 222), France (n = 522), Netherlands Pharmacovigilance Centre Lareb (n = 82), Jean–Louis Montastruc (n = 125), Drug Safety (n = 384), respectively, in all 2,128 articles. Similarly, the United States, Institut National de la Sante et de la Recherche Medicale, and Jean–Louis Montastruc had the most coauthorship strength at the macrolevel (global), mesolevel (local), and microlevel (individual). The topics of burst references covered are the development of methodology, issues of patients reporting and under-reporting, evaluation of methods and databases, assessment of causality, and perspectives in pharmacovigilance. Eight clusters were grouped in the document citation network. “Pharmacovigilance,” “adverse drug reactions,” “pharmacoepidemiology,” “drug safety,” and “signal detection” were the research priorities, while “drug-related side effects and adverse reactions,” “VigiBase,” “disproportionality analysis,” “social media,” “FAERS,” “chemotherapy,” “patient safety,” “reporting odds ratio,” and “preventability” might be the future research hotspots.

Conclusion: Positive synergies can be observed in this study by employing the multiple software tools which established the relationship between the units of analysis. The bibliometric analysis can organize the thematic development and guide the hotspots of pharmacovigilance in pharmacology and pharmacy.

Vaccine adverse event mentions in social media: Mining the language of Twitter conversations (Preprint)

Background

Traditional monitoring for adverse events following immunization (AEFI) relies on various established reporting systems, where there is inevitable lag between an AEFI occurring and its potential reporting and subsequent processing of reports. AEFI safety signal detection strives to detect AEFI as early as possible, ideally close to real time. Monitoring social media data holds promise as a resource for this.

Objective

The primary aim of this study is to investigate the utility of monitoring social media for gaining early insights into vaccine safety issues, by extracting vaccine adverse event mentions (VAEMs) from Twitter, using natural language processing techniques. The secondary aims are to document the natural language processing techniques used and identify the most effective of them for identifying tweets that contain VAEM, with a view to define an approach that might be applicable to other similar social media surveillance tasks.

Methods

A VAEM-Mine method was developed that combines topic modeling with classification techniques to extract maximal VAEM posts from a vaccine-related Twitter stream, with high degree of confidence. The approach does not require a targeted search for specific vaccine reaction–indicative words, but instead, identifies VAEM posts according to their language structure.

Results

The VAEM-Mine method isolated 8992 VAEMs from 811,010 vaccine-related Twitter posts and achieved an F₁ score of 0.91 in the classification phase.

Conclusions

Social media can assist with the detection of vaccine safety signals as a valuable complementary source for monitoring mentions of vaccine adverse events. A social media–based VAEM data stream can be assessed for changes to detect possible emerging vaccine safety signals, helping to address the well-recognized limitations of passive reporting systems, including lack of timeliness and underreporting.

Association between cyclin-dependent kinase 4/6 inhibitors and venous thromboembolism: analysis of F.A.E.R.S. data

Objectives:Real world studies have started to emerged on occurrence of venous thromboembolism (VTE) with cyclin-dependent kinase 4 and 6 (CDK4/6) inhibitors, but still deserve constant surveillance and evaluation. This study was to analyze this association. Methods: Adverse event cases were acquired from the US Food and Drug Administration Adverse Event Reporting System (FAERS) database betweenJanuary 1st 2015 and December 31st 2020.Signals indicating association between CDK 4/6 inhibitors and VTE were identified by reporting odds ratio (ROR). Results: CDK 4/6 inhibitors had a total of 631 reports of VTE (ROR 1.44, 95% CI 1.33-1.55) compared with non-CDK 4/6 inhibitors. Palbociclib (ROR 1.42, 95% CI 1.09-1.88) demonstratedthe highest number of VTE reports, followed by ribociclib (ROR 1.41, 95% CI 1.29-1.54) and abemaciclib (ROR 0.92, 95% CI 0.72-1.17). Conclusions:Although it is not able to confirm the casual relationship between VTE and CDK4/6 inhibitors, this study suggested signal of VTE reporting in patients receiving CDK4/6 inhibitors, which is likely to reflect a potential association. The results may enhance physicians' awareness of the potential side effect of VTE associated with CDK 4/6 inhibitors. An early recognition of VTE signs/symptoms could decrease the morbidity and severity of such adverse events.

Anaplastic lymphoma kinase tyrosine kinase inhibitor-induced hepatic failure in lung cancer patients: A study of signal mining and analysis of the FDA adverse event reporting system database

WHAT IS KNOWN AND OBJECTIVE

Anaplastic Lymphoma Kinase Tyrosine Kinase Inhibitors (ALK TKIs) are standard first-line therapy for non-small cell lung cancer patients with ALK rearrangement. Although some cases of hepatotoxicity related to these drugs have been reported, there is still a lack of investigation on severe hepatotoxicity, such as hepatic failure, with ALK TKIs.

METHODS

We evaluated ALK TKI (crizotinib, alectinib, brigatinib, ceritinib and lorlatinib)-induced hepatic failure events (AIHFEs), by using the Reporting Odds Ratio (ROR) and Bayesian Confidence Propagation Neural Network method for mining the adverse event report signals in the FDA Adverse Event Reporting System (FAERS) database from Jan 2013 to Dec 2019.

RESULTS AND DISCUSSION

The AIHFEs of "Hepatic failure," "hepatitis fulminant" and "hepatic necrosis" were defined as exposure event signals caused by ALK TKIs. The RORs of "Hepatic failure" were 4.95 (2.36-10.42) in alectinib, 3.77 (1.69-8.40) in ceritinib and 2.45 (1.60-3.76) in crizotinib, respectively. The ROR of "hepatitis fulminant" was 7.86 (3.52-17.54) in crizotinib. The Information Component value of "hepatic necrosis" was 1.97 (0.15) in alectinib. In reports of exposure-event signals, the clinical outcome of eventual death was common and could occur within 3 months. In the reports of "hepatic failure," there was no significant difference in the number of reports between men and women [OR=1.86 (0.94-3.67), p = 0.09].

WHAT IS NEW AND CONCLUSIONS

By mining the adverse event report signals in the FAERS database, we found the exposure event signals of AIHFEs in ALK TKIs were "hepatic failure," "hepatitis fulminant" and "hepatic necrosis". AIHFEs were more likely to appear in the reports of ceritinib, crizotinib and alectinib.

Artificial Intelligence, Real-World Automation and the Safety of Medicines

Despite huge technological advances in the capabilities to capture, store, link and analyse data electronically, there has been some but limited impact on routine pharmacovigilance. We discuss emerging research in the use of artificial intelligence, machine learning and automation across the pharmacovigilance lifecycle including pre-licensure. Reasons are provided on why adoption is challenging and we also provide a perspective on changes needed to accelerate adoption, and thereby improve patient safety. Last, we make clear that while technologies could be superimposed on existing pharmacovigilance processes for incremental improvements, these great societal advances in data and technology also provide us with a timely opportunity to reconsider everything we do in pharmacovigilance operations to maximise the benefit of these advances.

Evaluation of signal detection algorithms within the Elanco Animal Health Pharmacovigilance database

Statistical algorithms for detecting safety signals are beginning to be applied to Animal Health Pharmacovigilance (PV) databases. How these signal detection algorithms (SDAs) perform in an animal health PV database is the subject of this report. Statistical methods and SDAs were assessed against a set of known signals in order to identify which SDAs were most appropriate for signal detection using the Elanco Animal Health PV database. A reference set of adverse events that should signal was created for 31 products across four species. Nine SDAs based on five disproportionality statistical methods were evaluated against the reference set. The performance metrics were sensitivity, precision, specificity, accuracy, and F score. For bovine and porcine products, the Observed-to-Expected (O/E) SDA was the closest in terms of geometric distance to 100% sensitivity and 100% precision. For canine and feline products, the Information Component (IC) SDA was geometrically closest to 100% sensitivity and 100% precision. Principal Component Analysis confirmed that the O/E and IC SDAs were unique performers with respect to one another and other SDAs. The performance of the SDAs was dependent on the choice of the statistical method with differences seen between animal species.

Linezolid and the risk of lactic acidosis: Data mining and analysis of the FDA Adverse Event Reporting System

WHAT IS KNOWN AND OBJECTIVE

Lactic acidosis (LA) is a rare but potentially lethal side effect of linezolid (LZD). However, limited by the study population, the number of patients with LA is insufficient to summarize all the clinical characteristics and risk factors.

METHODS

We evaluated the association between LZD and LA using the reporting odd ratio (ROR) for mining the adverse event report signals in the FDA Adverse Event Reporting System database from January 2013 to December 2019.

RESULTS AND DISCUSSION

There were 6218 reports of LZD as the primary suspected drug or secondary suspected drug, of which 275 (4.42%) reports were of LA. The ROR of LA with the use of LZD was 39.976 (95% CI 35.365-45.189). In the age composition of patients, elderly individuals (aged ≥60 years) accounted for the higher proportion, 42.54% (n = 117). LA usually occurred two weeks after LZD administration (n = 33). LZD was the unique suspected drug, accounting for 37.45% (n = 103) of all reports of LA. The drug with the most frequent occurrence of combination with LZD was 'meropenem' and 'warfarin'.

WHAT IS NEW AND CONCLUSIONS

The ROR of LA caused by LZD was very high, and the number of reports about LA caused by other antibiotics was significantly different from that of LA caused by LZD. The drug combined with LZD did not seem to affect the occurrence of LA, and the high occurrence of warfarin in the reports deserves the attention of doctors.

Real-world use of nonvitamin K antagonist oral anticoagulant in atrial fibrillation patients with liver disease: A meta-analysis

Several studies have investigated the effectiveness and safety of nonvitamin K antagonist oral anticoagulants (NOACs) vs vitamin K antagonists (VKAs) in patients with atrial fibrillation (AF) and liver disease. Herein, we conducted a meta‐analysis to compare the effect of NOACs with VKAs in patients with AF and liver disease. We also conducted a subsidiary analysis to compare the risk of liver injury between NOACs and VKA in AF patients. We systematically searched the PubMed and Embase databases from January 2009 to May 2020 for the relevant studies. Hazard ratios (HRs) with 95% confidence intervals (CIs) were selected and pooled using a random‐effects model. A total of six cohorts were included. Compared with VKA use, the use of NOACs was associated with reduced risks of stroke or systemic embolism (HR 0.68, 95% CI 0.49‐0.93), all‐cause death (HR 0.69, 95% CI 0.63‐0.75), and intracranial bleeding (HR 0.49, 95% CI 0.40‐0.59), whereas the outcomes of major bleeding (HR 0.72, 95% CI 0.51‐1.01) and gastrointestinal bleeding (HR 0.84, 95% CI 0.51‐1.36) were not significantly different between groups in AF patients with liver disease. Moreover, compared with VKA use, the use of NOACs was associated with a reduced risk of liver injury (HR 0.72, 95% CI 0.61‐0.84) in AF patients. Compared with VKAs, the use of NOACs was associated with reduced risks of stroke or systemic embolism, all‐cause death, and intracranial bleeding in AF patients with liver disease, and associated with a reduced risk of liver injury in AF patients.

Aromatase inhibitors associated osteonecrosis of jaw: signal refining to identify pseudo safety signals

Background Signal generation through data mining algorithms is an innovative and emerging field in pharmacovigilance. Early detection of safety signals is important for public health safety. However, the possibility of generating pseudo signals should not be overlooked. Objective Our study aimed to identify potential signals of aromatase inhibitors associated Osteonecrosis of Jaw and assess the possibilities of the safety signal to be a pseudo signal/false positive in FDA Adverse Event Reporting System (FAERS). Setting Spontaneously reported data in FAERS database. Methods Data for this study were obtained from the public release of data in FAERS. OpenVigil, a pharmacovigilance analytical tool was used to access FAERS data. Reporting Odds Ratio (ROR) was used to assess the relation between the drug and adverse event. A value of ROR-1.96SE  > 1, (SE-standard error) was considered positive. Main outcome measure Signal strength. Results FAERS database had a total of 15,178 reports for Osteonecrosis of Jaw. Amongst which 617 reports were associated with aromatase inhibitors. Signal strength ROR (lower bound of the 95% CI) for letrozole, anastrozole and exemestane associated Osteonecrosis of Jaw without any background correction was 8.34, 6.64 and 15.14 respectively. Upon removing the reports of concomitantly administered drugs (bisphosphonates and denosumab), signal strength drastically decreased to 0.03, 0.36 and 0.47 for letrozole, anastrozole and exemestane respectively. The signal strength of bisphosphonates and denosumab associated Osteonecrosis of Jaw was not changed significantly upon removal of aromatase inhibitors. Conclusion Our study concluded that the signal generated for aromatase inhibitors associated Osteonecrosis of Jaw in FAERS database can be false positive. Careful background corrections with identification of those risk factors are imperative to exclude false positive results.

Bayesian hierarchical methods in the detection of potentially teratogenic first-trimester medications

PURPOSE

Bayesian hierarchical models (BHMs) have been used to identify adverse drug reactions, allowing information sharing amongst adverse reactions and drugs expected to have similar properties. This study evaluated the use of BHMs in the routine signal detection analyses of potential first-trimester teratogens, where these models have not previously been applied.

METHODS

Data on 15 058 malformed foetuses exposed to first trimester medications (1995-2011) from 13 European congenital anomaly (CA) registries were analysed. The proportion of each CA in women taking a specific medication was compared with the proportion of that CA in all other women in the dataset (55 CAs × 523 medications). BHMs were grouped by either medications or CAs or by both simultaneously, and the results compared with analysing each medication-CA combination separately and adjusting for multiplicity using a double false discovery rate (FDR) procedure. The proportions of "high-risk" medications (medications which have been shown to carry a moderate to high risk of foetal malformations) identified as potential signals were compared, as well as the total number of potential signals requiring follow up (the effective workload).

RESULTS

BHMs identified more high-risk medications than the double FDR method, but the effective workload was larger. A BHM grouping both medications and CAs, for example, identified 23% of high-risk medications compared with 14% by the double FDR; however, there was an increase from 16 to 71 potential signals requiring follow up.

CONCLUSION

For comparable effective workloads, BHMs did not outperform the double FDR, which is comparatively straightforward to implement and is therefore recommended for continued use in teratogenic signal detection analyses.

Early signal detection of adverse events following influenza vaccination using proportional reporting ratio, Victoria, Australia

INTRODUCTION

Timely adverse event following immunisation (AEFI) signal event detection is essential to minimise further vaccinees receiving unsafe vaccines. We explored the proportional reporting ratio (PRR) ability to detect two known signal events with influenza vaccines with the aim of providing a model for prospective routine signal detection and improving vaccine safety surveillance in Australia.

METHODS

Passive AEFI surveillance reports from 2008-2017 relating to influenza vaccines were accessed from the Australian SAEFVIC (Victoria) database. Proportional reporting ratios were calculated for two vaccine-event categories; fever and allergic AEFI. Signal detection sensitivity for two known signal events were determined using weekly data; cumulative data by individual year and; cumulative for all previous years. Signal event thresholds of PRR ≥2 and Chi-square ≥4 were applied.

RESULTS

PRR provided sensitive signal detection when calculated cumulatively by individual year or by all previous years. Known signal events were detected 15 and 11 days earlier than traditional methods used at the time of the actual events.

CONCLUSION

Utilising a single jurisdiction's data, PRR improved vaccine pharmacovigilance and showed the potential to detect important safety signals much earlier than previously. It has potential to maximise immunisation safety in Australia. This study progresses the necessary work to establish national cohesion for passive surveillance signal detection and strengthen routine Australian vaccine pharmacovigilance.

A Review of the Food and Drug Administration Adverse Event Reporting System for Tramadol-Related Hypoglycemia

Background: Hypoglycemia is a rare adverse effect of tramadol that is described in the medical literature and package insert. Objective: The purpose of this study was to review reports of tramadol and hypoglycemia in the Food and Drug Administration Adverse Event Reporting System (FAERS) database to determine a potential association. Methods: Disproportionality analysis with Bayesian correction was used to compare tramadol and hypoglycemia with other medications in FAERS. The results were considered significant if the fifth percentile of the Empirical Bayesian geometric mean distribution (EB05) >2. Logistic regression odds ratios was used to determine if age, diabetes medications, and renal insufficiency masked the disproportionality of hypoglycemia, with the fifth percentile of the logistic regression odds ratio (LR05) >2 indicating a potential signal. The Interaction Signal Score (INTSS) was computed to determine the influence of predisposing risk factors on the signal. Results: A total of 605 cases of tramadol-associated hypoglycemia were reported, but our results were not significant (EB05: 1.590). Tramadol-associated hypoglycemia was significant in patients who did not take diabetes medications (EB05: 2.256; LR05: 2.2104). Renal insufficiency was not found to increase the risk of tramadol-associated hypoglycemia (INTSS: 0.865). There was a significant signal for tramadol-associated hypoglycemia in patients aged 0 to 1 year (LR05: 3.0240) and 2 to 4 years (LR05: 2.6853). Conclusion and Relevance: Results of our analysis suggest a potential signal between hypoglycemia and tramadol use in patients not taking diabetes medications. Our results do not support a predisposition for tramadol-associated hypoglycemia in patients with renal insufficiency, increasing age, and/or diabetes as noted in the tramadol package insert.

Evaluation of quantitative signal detection in EudraVigilance for orphan drugs: possible risk of false negatives

Different strategies have been studied to allow a better characterization of the safety profile of orphan drugs soon after their approval. At the end of the development phases only few data are available because of the small number of subjects exposed to an orphan medicine for the treatment of rare or ultra-rare conditions. As a consequence, the evaluation of the safety profile is limited at the time of the first approval. In the post-marketing period, all available sources should be combined for a better understanding of the safety of orphan drugs. These sources, include outputs from large databases such as the European Medicines Agency’s EudraVigilance database. Analyses of data from this source are required to be performed by marketing authorization holders (MAHs) as part of their signal management activities. In 2018, the Pharmacovigilance Risk Assessment Committee (PRAC) assessed 114 confirmed signals, 79% of which included data from EudraVigilance. MAHs have access to statistical calculations for drug–event combinations (DECs) from EudraVigilance, provided in the form of measures of disproportionality of ratios of the observed proportion of spontaneous cases for a DEC in relation to the proportion of cases that would be expected if no association existed between the drug and the event. However, such statistical summaries for orphan drugs could be misleading because of the very limited safety data available for orphan drugs (under-reporting together with low numbers of exposed patients). In addition, the applied statistical methodology in most instances is constrained by different confounding factors such as indications of specific medicines and the wide spectrum of medical conditions/diseases of patients from whom reporting of disproportionality ratios are derived (i.e. proportions of DECs for orphan drugs (ODECs) from a small patient population suffering the rare disease and the proportion of DECs in the rest of the population represented in the whole database who have been treated with other medicines for a wide range of indications, and prescribed to treat completely different medical conditions). As expected, these statistical calculations produced not only signals of disproportionate reporting (SDRs) that are false positives, but also not sensitive enough to detect certain SDRs, thus resulting in false negatives. In the context of rare/ultra-rare life-threatening diseases where new molecules have been made available on the market on the basis of their proven efficacy, but with only limited safety data at the time of approval, false negatives could be a special concern since unlikely converted in positives or becoming positives with notable delay. Subgroup analyses (using a limited dataset comprising ADRs within specific individual case safety reports (ICSRs), sorted by indication/disease relevant to the drug of interest could, at least in part, possibly reduce some of the weaknesses resulting from the abovementioned confounding factors. On the other hand it could also cause the loss of some identification of SDRs that would be captured if no database restrictions had been undertaken. Therefore, data subgroup analysis should not be selected as a preferred approach to quantitative signal detection for orphan drugs but rather evaluated as complementary possibly to confirm negatives or to further characterize detected SDRs. Some examples of false negatives originating from quantitative signal detection in EudraVigilance applied to orphan drugs are discussed in this article.

Risk Evaluation for Acute Kidney Injury Induced by the Concomitant Use of Valacyclovir, Analgesics, and Renin-Angiotensin System Inhibitors: The Detection of Signals of Drug-Drug Interactions

Background: Drug-related acute kidney disease is a common side effect of valacyclovir (VACV) treatment. Although analgesics are frequently administered concomitantly with VACV to treat the pain of herpes zoster, the differences between nonsteroidal anti-inflammatory drugs (NSAIDs) and acetaminophen in relation to VACV-related acute kidney injury (AKI) are unclear. The risk for AKI with concomitant use of VACV and renin–angiotensin system (RAS) inhibitors that can cause AKI via a similar mechanism to NSAIDs is also unknown. We therefore evaluated the association between concomitant use of these drugs and VACV-related AKI, which was characterized according to the Japanese Adverse Drug Event Report (JADER) database.

Methods: We analyzed data from the JADER database, which is a spontaneous reporting system. The reporting odds ratio was used to evaluate the signals of AKI.

Results: A high proportion of VACV-related AKI cases occurred in summer. There was an increase in AKI signal in cases with concomitant use of VACV and NSAIDs, while no increase was detected in cases with concomitant use of VACV and acetaminophen. AKI events in cases with concomitant use of VACV and NSAIDs were more frequent in older and female patients and those with hypertension. Additionally, a signal increase for VACV-related AKI was observed with concomitant use of RAS inhibitors, with or without NSAIDs.

Conclusions: We identified a seasonal variation in VACV-related AKI. Additionally, our findings indicate that acetaminophen might represent a safer analgesic than NSAIDs with respect to VACV-related AKI. We also identified candidate risk factors for AKI with concomitant use of NSAIDs, such as older age, female sex, and hypertension. Although further studies are warranted, our findings highlight the need to consider concomitant drug use and seasonal factors that lead to urinary output loss so that VACV-related AKI can be avoided.

Non-Vitamin K Antagonist Oral Anticoagulants and Risk of Serious Liver Injury

BACKGROUND

Non-vitamin K antagonist oral anticoagulants (NOACs) are relatively new drugs used for stroke prevention in nonvalvular atrial fibrillation (NVAF). However, there are concerns that their use may be associated with hepatotoxic effects.

OBJECTIVES

The purpose of this study was to determine whether the use of NOACs is associated with an increased risk of serious liver injury compared with the use of vitamin K antagonists (VKAs) in NVAF patients with and without prior liver disease.

METHODS

Using the administrative databases of the Canadian province of Quebec's health insurances, the authors conducted a cohort study among patients newly diagnosed with NVAF between January 2011 and December 2014. Adjusted hazard ratios (HRs) and 95% confidence intervals (CIs) of serious liver injury (defined as either a hospitalization or related death) were estimated using time-dependent Cox proportional hazards models, comparing current use of NOACs to current use of VKAs separately among patients with or without prior liver disease.

RESULTS

The cohort comprised 51,887 patients, including 3,778 with prior liver disease. During 68,739 person-years of follow-up, 585 patients experienced a serious liver injury. Compared with current use of VKAs, current use of NOACs was not associated with an increased risk of serious liver injury in patients without or with prior liver disease (adjusted HR: 0.99; 95% CI: 0.68 to 1.45; and adjusted HR: 0.68; 95% CI: 0.33 to 1.37, respectively).

CONCLUSIONS

Compared with VKAs, NOACs were not associated with an increased risk of serious liver injury irrespective of baseline liver status. Overall, these results provide reassurance regarding the hepatic safety of NOACs.

Evaluation of Acute Kidney Injury Associated With Anticancer Drugs Used in Gastric Cancer in the Japanese Adverse Drug Event Report Database

Background: Development of acute kidney injury (AKI) depends on the severity of renal dysfunction, clinical setting, comorbid factors, and geographical location. Gastric cancer is one of the deadliest malignancies worldwide, and its incidence is significantly high in Japan. Objective: We analyzed the rank-order of the association of anticancer agents for gastric cancer with AKI using a spontaneous reporting system database, the Japanese Adverse Drug Event Report database. Methods: We performed a retrospective pharmacovigilance disproportionality analysis using the adverse event reports submitted to the Pharmaceuticals and Medical Devices Agency between April 2004 and March 2017. Results: Anticancer drug-related AKI was common in patients in their 60s and 70s (39.2% and 43.2%, respectively). AKI occurred most frequently within 1 month after anticancer drug administration. The signals of AKI were reported after treatment with S-1 (tegafur/gimeracil/oteracil), cisplatin (CDDP), and capecitabine, with significant adjusted reporting odds ratios (95% CI) of 1.50 (1.09-2.07), 3.43 (2.48-4.74), and 1.82 (1.15-2.90), respectively. CDDP-induced AKI was more likely to occur in patients who were male, hypertension, or diabetes mellitus. Conclusion and Relevance: This study showed that most AKI cases were related to S-1 and/or CDDP adjuvant chemotherapy for gastric cancer treatment. The data also clarified that AKIs occurred within 1 month and that their clinical outcomes were more severe than previous reports of drug-induced AKI in general medicine. Our study provides useful information to minimize the risks of administration to patients at high risk for S-1 and/or CDDP containing chemotherapy-induced AKI.

A systematic literature review and classification of knowledge discovery in traditional medicine

INTRODUCTION AND OBJECTIVE

Despite the importance of machine learning methods application in traditional medicine there is a no systematic literature review and a classification for this field. This is the first comprehensive literature review of the application of data mining methods in traditional medicine.

METHOD

We reviewed 5 database between 2000 to 2017 based on the Kitchenham systematic review methodology. 502 articles were identified and reviewed for their relevance to application of machine learning methods in traditional medicine, 42 selected papers were classified and categorized on four dimension; 1) application domain of data mining techniques in traditional medicine; 2) the data mining methods most frequently used in traditional medicine; 3) main strength and limitation of data mining techniques in traditional medicine; 4) the performance evaluation methods in data mining methods in traditional medicine.

RESULT

The result obtained showed that main application domain of data mining techniques in traditional medicine was related to syndrome differentiation. Bayesian Networks (BNs), Artificial Neural Networks (ANNs) and Support Vector Machines (SVMs) were recognized as being the methods most frequently applied in traditional medicine. Furthermore, each data mining techniques has its own strength and limitations when applied in traditional medicine. Single scaler methods were frequently used for performance evaluation of data mining methods.

CONCLUSION

Machine learning methods have become an important research field in traditional medicine. Our research provides information about this methods by examining the related articles.

Lessons from meta-analyses of randomized clinical trials for analysis of distributed networks of observational databases

Networks of constellations of longitudinal observational databases, often electronic medical records or transactional insurance claims or both, are increasingly being used for studying the effects of medicinal products in real-world use. Such databases are frequently configured as distributed networks. That is, patient-level data are kept behind firewalls and not communicated outside of the data vendor other than in aggregate form. Instead, data are standardized across the network, and queries of the network are executed locally by data partners, and summary results provided to a central research partner(s) for amalgamation, aggregation, and summarization. Such networks can be huge covering years of data on upwards of 100 million patients. Examples of such networks include the FDA Sentinel Network, ASPEN, CNODES, and EU-ADR. As this is a new emerging field, we note in this paper the conceptual similarities and differences between the analysis of distributed networks and the now well-established field of meta-analysis of randomized clinical trials (RCTs). We recommend, wherever appropriate, to apply learnings from meta-analysis to help guide the development of distributed network analyses of longitudinal observational databases.

Using social media in safety signal management: is it reliable?

Social media use is growing globally, with a reported 3 billion active users in 2017. This medium is used increasingly in a health setting by patients (and to a limited extent, healthcare professionals) to share experiences and ask advice on medical conditions as well as pharmaceutical products. In recent years, attention has turned to this huge, generally untapped, source of potential health information as a possible tool for pharmacovigilance, and in particular signal detection. In this article we explore some of the challenges of utilizing social media for safety signal detection and look at some of the pilot studies conducted to date in order to weigh the evidence for and against the utility of social media data in safety signal detection. After doing so we can conclude that the analysis of social media datasets has demonstrated a limited contribution to the signal detection and signal management process. The data available in social media can complement blind spots in traditional pharmacovigilance datasets and provide significant value for targeted investigations and studies such as those relating to abuse, misuse, use in pregnancy, and patient sentiments.

Masking by vaccines in pediatric drug safety signal detection in the EudraVigilance database

PURPOSE

Postmarketing drug safety surveillance relies upon measures of disproportionate reporting in spontaneous reporting systems. It has been hypothesized that products or events reported frequently may "mask" signals.

METHODS

We analyzed the masking effect of vaccines in pediatrics in the EudraVigilance database by conducting disproportionality analysis in the full database (containing vaccine exposures) and in a restricted set (excluding vaccine exposures). We measured performance of the reporting odds ratio (ROR) in both data sets using a pediatric-specific drug reference set and in the absence of a reference set. We assessed masking effects across age groups and conducted a classification tree (CART) analysis.

RESULTS

Removal of vaccines decreased the ROR values both in negative and positive controls. Exceptions were drug-event combinations including outcomes frequent in vaccine reports. When restricted to positive control associations, removal of vaccine-related events resulted in increased ROR values for events commonly reported following vaccination. For events rarely associated with vaccination, ROR values decreased for all age groups, especially infants. Analysis in the absence of a reference set showed decrease in ROR following vaccine removal and CART revealed that change in ROR with vaccine removal depended upon age and proportion of reports including a vaccine.

CONCLUSIONS

Removal of vaccines for signal detection in a pediatric population has an impact on ROR, dependent upon the reporting frequency of the event of interest in combination with vaccines. We recommend stratification by age and removal of vaccine exposures if the investigated adverse drug reactions include those typically reported in association with vaccines for the age strata.

Early anomaly detection in smart home: A causal association rule-based approach

As the world's population grows older, an increasing number of people are facing health issues. For the elderly, living alone can be difficult and dangerous. Consequently, smart homes are becoming increasingly popular. A sensor-rich environment can be exploited for healthcare applications, in particular, anomaly detection (AD). The literature review for this paper showed that few works consider environmental factors to detect anomalies. Instead, the focus is on user activity and checking whether it is abnormal, i.e., does not conform to expected behavior. Furthermore, reducing the number of anomalies using early detection is a major issue in many applications. In this context, anomaly-cause discovery may be helpful in recommending actions that may prevent risk. In this paper, we present a novel approach for detecting the risk of anomalies occurring in the environment regarding user activities. The method relies on anomaly-cause extraction from a given dataset using causal association rules mining. These anomaly causes are utilized afterward for real-time analysis to detect the risk of anomalies using the Markov logic network machine learning method. The detected risk allows the method to recommend suitable actions to perform in order to avoid the occurrence of an actual anomaly. The proposed approach is implemented, tested, and evaluated for each contribution using real data obtained from an intelligent environment platform and real data from a clinical datasets. Experimental results prove our approach to be efficient in terms of recognition rate.

Signals of gastroesophageal reflux disease caused by incretin-based drugs: a disproportionality analysis using the Japanese adverse drug event report database

Background

Incretin-based drugs are important in the treatment of type 2 diabetes. However, among the incretin-based drugs, glucagon-like peptide-1 receptor agonists (GLP-1-RAs) have been reported to cause gastroesophageal reflux disease (GERD)-like symptoms making it difficult to continue treatment. Therefore, with the aim of clarifying the relationship between incretin-based drugs and GERD-like symptoms, we conducted a pharmacoepidemiological study using the Japanese adverse drug event report database (JADER).

Methods

Dipeptidyl peptidase-4 inhibitors (DPP-4-Is) and GLP-1-RAs were set as the incretin-based target drugs. The reporting odds ratio (ROR) and the information component (IC) was used for the detection of quantitative signals. Furthermore, we also compared the time to onset of GERD-like symptoms by log-rank test.

Results

GERD-like symptoms were reported in 36 GLP-1-RAs cases (ROR: 5.61, 95% confidence interval (95% CI): 3.95–7.96 and IC: 2.17, 95% CI: 1.66–2.67) and GLP-1-RAs were detected in the signal. In contrast, DPP-4-Is were not detected in the signal.

There was no sex difference with regard to the expression time of GERD-like symptoms by GLP-1-RAs (log-rank test, p = 0.5381). However, the expression time of GERD-like symptoms from GLP-1-RAs was shorter in patients older than 70 years of age than that in those younger than 70 years of age (log-rank test, p < 0.0001).

Conclusions

The administration of GLP-1-RA had a higher incidence of GERD-like symptoms earlier than the administration of DPP-4-Is. In this study, although we think that further investigation is necessary, and suggest that patients older than 70 years of age who have been administered GLP-1-RAs need earlier attention to address GERD-like symptoms than younger patients.

Botulinum Toxin Type A Overdoses: Analysis of the FDA Adverse Event Reporting System Database

INTRODUCTION

Published literature on overdoses related to botulinum toxin A (BtxA) agents is scarce.

OBJECTIVE

The aim of this study was to assess the BtxA drug class' respective agents for associations with overdose.

METHODS

United States Food and Drug Administration (FDA) adverse event reporting system (FAERS) database was utilized to search for overdoses. The analysis was conducted on data between second quarter 2014 and third quarter 2017. BtxA cases were included when they were considered the "Primary Suspect" drug. Overdose was defined as presence of 'overdose' being reported as an adverse event. Primary outcome was incidence of 'overdose' compared within the respective agents. Additionally, a disproportionality analysis was conducted utilizing reporting odds ratio (ROR) versus onabotulinumtoxinA as a referent while controlling for confounding variables.

RESULTS

A total of 3,837,406 unique adverse events were reported during the study period for all drugs in the FAERS database. Of which, 13,078 were BtxA cases. The rate of adverse events involving overdose for abobotulinumtoxinA (20.2%; 215/1065) was significantly higher than both onabotulinumtoxinA (0.4%; 48/11,323; p < 0.0001) and incobotulinumtoxinA (0.1%; 1/690; p < 0.0001). In the regression analysis, abobotulinumtoxinA (ROR 73.26; 95% CI 51.17-104.90) had a significant association with overdose, whereas incobotulinumtoxinA (ROR 0.73; 95% CI 0.10-5.36) did not, versus the referent onabotulinumtoxinA.

CONCLUSION

The present analysis showed adverse events of abobotulinumtoxinA were significantly associated with overdose versus the other two BtxA agents. Overdose can be difficult to research, particularly for in-clinic administered drugs. Future studies should venture to confirm these results in new and novel ways.

Methodological Considerations for Comparison of Brand Versus Generic Versus Authorized Generic Adverse Event Reports in the US Food and Drug Administration Adverse Event Reporting System (FAERS)

BACKGROUND

The US Food and Drug Administration Adverse Event Reporting System (FAERS), a post-marketing safety database, can be used to differentiate brand versus generic safety signals.

OBJECTIVE

To explore the methods for identifying and analyzing brand versus generic adverse event (AE) reports.

METHODS

Public release FAERS data from January 2004 to March 2015 were analyzed using alendronate and carbamazepine as examples. Reports were classified as brand, generic, and authorized generic (AG). Disproportionality analyses compared reporting odds ratios (RORs) of selected known labeled serious adverse events stratifying by brand, generic, and AG. The homogeneity of these RORs was compared using the Breslow-Day test. The AG versus generic was the primary focus since the AG is identical to brand but marketed as a generic, therefore minimizing generic perception bias. Sensitivity analyses explored how methodological approach influenced results.

RESULTS

Based on 17,521 US event reports involving alendronate and 3733 US event reports involving carbamazepine (immediate and extended release), no consistently significant differences were observed across RORs for the AGs versus generics. Similar results were obtained when comparing reporting patterns over all time and just after generic entry. The most restrictive approach for classifying AE reports yielded smaller report counts but similar results.

CONCLUSION

Differentiation of FAERS reports as brand versus generic requires careful attention to risk of product misclassification, but the relative stability of findings across varying assumptions supports the utility of these approaches for potential signal detection.

Development of an automated assessment tool for MedWatch reports in the FDA adverse event reporting system

Objective

As the US Food and Drug Administration (FDA) receives over a million adverse event reports associated with medication use every year, a system is needed to aid FDA safety evaluators in identifying reports most likely to demonstrate causal relationships to the suspect medications. We combined text mining with machine learning to construct and evaluate such a system to identify medication-related adverse event reports.

Methods

FDA safety evaluators assessed 326 reports for medication-related causality. We engineered features from these reports and constructed random forest, L1 regularized logistic regression, and support vector machine models. We evaluated model accuracy and further assessed utility by generating report rankings that represented a prioritized report review process.

Results

Our random forest model showed the best performance in report ranking and accuracy, with an area under the receiver operating characteristic curve of 0.66. The generated report ordering assigns reports with a higher probability of medication-related causality a higher rank and is significantly correlated to a perfect report ordering, with a Kendall's tau of 0.24 ( P  = .002).

Conclusion

Our models produced prioritized report orderings that enable FDA safety evaluators to focus on reports that are more likely to contain valuable medication-related adverse event information. Applying our models to all FDA adverse event reports has the potential to streamline the manual review process and greatly reduce reviewer workload.

Drug Safety Monitoring in Children: Performance of Signal Detection Algorithms and Impact of Age Stratification

INTRODUCTION

Spontaneous reports of suspected adverse drug reactions (ADRs) can be analyzed to yield additional drug safety evidence for the pediatric population. Signal detection algorithms (SDAs) are required for these analyses; however, the performance of SDAs in the pediatric population specifically is unknown. We tested the performance of two SDAs on pediatric data from the US FDA Adverse Event Reporting System (FAERS) and investigated the impact of age stratification and age adjustment on the performance of SDAs.

METHODS

We tested the performance of two established SDAs: the proportional reporting ratio (PRR) and the empirical Bayes geometric mean (EBGM) on a pediatric dataset from FAERS (2004-2012). We compared the performance of the SDAs with a published pediatric-specific reference set by calculating diagnostic test-related statistics, including the area under the curve (AUC) of receiver operating characteristics. Impact of age stratification and age-adjustment on the performance of the SDAs was assessed. Age adjustment was performed by pooling (Mantel-Hanszel) stratum-specific estimates.

RESULTS

A total of 115,674 pediatric reports (patients aged 0-18 years) comprising 893,587 drug-event combinations (DECs) were analysed. Crude values of the AUC were similar for both SDAs: 0.731 (PRR) and 0.745 (EBGM). Stratification unmasked four DECs, e.g., 'ibuprofen and thrombocytopenia'. Age adjustment did not improve performance.

CONCLUSION

The performance of the two tested SDAs was similar in the pediatric population. Age adjustment does not improve performance and is therefore not recommended to be performed routinely. Stratification can reveal new associations, and therefore is recommended when either drug use is age-specific or when an age-specific risk is suspected.

Mixed Approach Retrospective Analyses of Suicide and Suicidal Ideation for Brand Compared with Generic Central Nervous System Drugs

INTRODUCTION

Several different types of drugs acting on the central nervous system (CNS) have previously been associated with an increased risk of suicide and suicidal ideation (broadly referred to as suicide). However, a differential association between brand and generic CNS drugs and suicide has not been reported.

OBJECTIVES

This study compares suicide adverse event rates for brand versus generic CNS drugs using multiple sources of data.

METHODS

Selected examples of CNS drugs (sertraline, gabapentin, zolpidem, and methylphenidate) were evaluated via the US FDA Adverse Event Reporting System (FAERS) for a hypothesis-generating study, and then via administrative claims and electronic health record (EHR) data for a more rigorous retrospective cohort study. Disproportionality analyses with reporting odds ratios and 95% confidence intervals (CIs) were used in the FAERS analyses to quantify the association between each drug and reported suicide. For the cohort studies, Cox proportional hazards models were used, controlling for demographic and clinical characteristics as well as the background risk of suicide in the insured population.

RESULTS

The FAERS analyses found significantly lower suicide reporting rates for brands compared with generics for all four studied products (Breslow-Day P < 0.05). In the claims- and EHR-based cohort study, the adjusted hazard ratio (HR) was statistically significant only for sertraline (HR 0.58; 95% CI 0.38-0.88).

CONCLUSION

Suicide reporting rates were disproportionately larger for generic than for brand CNS drugs in FAERS and adjusted retrospective cohort analyses remained significant only for sertraline. However, even for sertraline, temporal confounding related to the close proximity of black box warnings and generic availability is possible. Additional analyses in larger data sources with additional drugs are needed.

Adverse drug reaction early warning using user search data

Purpose

The purpose of this paper is to propose a framework to detect adverse drug reactions (ADRs) using internet user search data, so that ADR events can be identified early. Empirical investigation of Avandia, a type II diabetes treatment, is conducted to illustrate how to implement the proposed framework.

Design/methodology/approach

Typical ADR identification measures and time series processing techniques are used in the proposed framework. Google Trends Data are employed to represent user searches. The baseline model is a disproportionality analysis using official drug reaction reporting data from the US Food and Drug Administration’s Adverse Event Reporting System.

Findings

Results show that Google Trends series of Avandia side effects search reveal a significant early warning signal for the side effect emergence of Avandia. The proposed approach of using user search data to detect ADRs is proved to have a longer leading time than traditional drug reaction discovery methods. Three more drugs with known adverse reactions are investigated using the selected approach, and two are successfully identified.

Research limitations/implications

Validation of Google Trends data’s representativeness of user search is yet to be explored. In future research, user search in other search engines and in healthcare web forums can be incorporated to obtain a more comprehensive ADR early warning mechanism.

Practical implications

Using internet data in drug safety management with a proper early warning mechanism may serve as an earlier signal than traditional drug adverse reaction. This has great potential in public health emergency management.

Originality/value

The research work proposes a novel framework of using user search data in ADR identification. User search is a voluntary drug adverse reaction exploration behavior. Furthermore, user search data series are more concise and accurate than text mining in forums. The proposed methods as well as the empirical results will shed some light on incorporating user search data as a new source in pharmacovigilance.

Comparison of brand versus generic antiepileptic drug adverse event reporting rates in the U.S. Food and Drug Administration Adverse Event Reporting System (FAERS)

OBJECTIVE

Despite the cost saving role of generic anti-epileptic drugs (AEDs), debate exists as to whether generic substitution of branded AEDs may lead to therapeutic failure and increased toxicity. This study compared adverse event (AE) reporting rates for brand vs. authorized generic (AG) vs. generic AEDs. Since AGs are pharmaceutically identical to brand but perceived as generics, the generic vs. AG comparison minimized potential bias against generics.

METHODS

Events reported to the U.S. Food and Drug Administration Adverse Event Reporting System between January 2004 to March 2015 with lamotrigine, carbamazepine, and oxcarbazepine listed as primary or secondary suspect were classified as brand, generic, or AG based on the manufacturer. Disproportionality analyses using the reporting odds ratio (ROR) assessed the relative rate of reporting of labeled AEs compared to reporting these events with all other drugs. The Breslow-Day statistic compared RORs across brand, AG, and other generics using a Bonferroni-corrected P<0.01.

RESULTS

A total of 27,150 events with lamotrigine, 13,950 events with carbamazepine, and 5077 events with oxcarbazepine were reported, with generics accounting for 27%, 41%, and 32% of reports, respectively. Although RORs for the majority of known AEs were different between brand and generics for all three drugs of interest (Breslow-Day P<0.001), RORs generally were similar for AG and generic comparisons. Generic lamotrigine and carbamazepine were more commonly involved in reports of suicide or suicidal ideation compared with the respective AGs based on a multiple comparison-adjusted P<0.01.

SIGNIFICANCE

Similar AED reporting rates were observed for the AG and generic comparisons for most outcomes and drugs, suggesting that brands and generics have similar reporting rates after accounting for generic perception biases. Disproportional suicide reporting was observed for generics compared with AGs and brand, although this finding needs further study.

Can Disproportionality Analysis of Post-marketing Case Reports be Used for Comparison of Drug Safety Profiles?

Clinical trials usually do not have the power to detect rare adverse drug reactions. Spontaneous adverse reaction reports as for example available in post-marketing safety databases such as the FDA Adverse Event Reporting System (FAERS) are therefore a valuable source of information to detect new safety signals early. To screen such large data-volumes for safety signals, data-mining algorithms based on the concept of disproportionality have been developed. Because disproportionality analysis is based on spontaneous reports submitted for a large number of drugs and adverse event types, one might consider using these data to compare safety profiles across drugs. In fact, recent publications have promoted this practice, claiming to provide guidance on treatment decisions to healthcare decision makers. In this article we investigate the validity of this approach. We argue that disproportionality cannot be used for comparative drug safety analysis beyond basic hypothesis generation because measures of disproportionality are: (1) missing the incidence denominators, (2) subject to severe reporting bias, and (3) not adjusted for confounding. Hypotheses generated by disproportionality analyses must be investigated by more robust methods before they can be allowed to influence clinical decisions.