Methods for causality assessment of adverse drug reactions: a systematic review

Deduplicating the FDA adverse event reporting system with a novel application of network-based grouping

OBJECTIVE

To improve the reliability of data mining for product safety concerns in the Food and Drug Administration's (FDA) Adverse Event Reporting System (FAERS) by robustly identifying duplicate reports describing the same patient experience.

MATERIALS AND METHODS

A duplicate detection algorithm based on a probabilistic record linkage algorithm, including features extracted from report narratives, and designed to support FAERS case safety review as part of the Information Visualization Platform (InfoViP) has been upgraded into a full deduplication pipeline for the entire FAERS database. The pipeline contains several new and updated components, including a network analysis-based community detection routine for breaking up sparsely connected groups of duplicates constructed from chains of pairwise comparisons. The pipeline was applied to all 29 million FAERS reports to assemble groups of duplicate cases.

RESULTS

The pipeline was evaluated on 12 human expert adjudicated data sets with a total of 2300 reports and was found to have better overall performance than the current tool used at the FDA for labeling duplicates on 10 of them, with F1 scores ranging from 0.36 to 0.93, with half above 0.75. Because minimizing false discovery increases human expert review efficiency, the improved deduplication pipeline was applied to all historic and daily incoming FAERS reports at FDA and identified about 5 million reports as duplicates.

CONCLUSIONS

The InfoViP deduplication pipeline is operating at FDA to identify duplicate case reports in FAERS and provide deduplicated input for improved efficiency and accuracy of safety review operations like adverse event data mining calculations.

Causality assessment of adverse events by healthcare professionals in an academic hospital setting: a descriptive retrospective study

KEY POINTS

Adverse events among drug users are frequent during hospital stay. Causality assessment of adverse events was poorly documented by healthcare professionals in a hospital setting.

Interplay of Spontaneous Reporting and Longitudinal Healthcare Databases for Signal Management: Position Statement from the Real-World Evidence and Big Data Special Interest Group of the International Society of Pharmacovigilance

Signal management, defined as the set of activities from signal detection to recommendations for action, is conducted using different data sources and leveraging data from spontaneous reporting databases (SRDs), which represent the cornerstone of pharmacovigilance. However, the exponentially increasing generation and availability of real-world data collected in longitudinal healthcare databases (LHDs), along with the rapid evolution of artificial intelligence-based algorithms and other advanced analytical methods, offers a wide range of opportunities to complement SRDs throughout all stages of signal management, especially signal detection. Integrating information derived from SRDs and LHDs may reduce their respective limitations, thus potentially enhancing post-marketing surveillance. The aim of this position statement is to critically evaluate the complementary role of SRDs and LHDs in signal management, exploring the potential benefits and challenges in integrating information coming from these two data sources. Furthermore, we presented successful cases of the interplay between SRDs and LHDs for signal management, along with future opportunities and directions to improve such interplay.

Interactions Between Direct Oral Anticoagulants and Drugs Involving CYP-Enzymes and P-gp Transporters Detected by a Clinical Decision Support System: A Retrospective Cohort Study

PURPOSE

Physicians can be supported by electronic clinical decision support systems (CDSS) for drug-drug interaction (DDI) management of DDIs between direct oral anticoagulants (DOACs) and CYP3 A4/P-glycoprotein (P-gp) inhibitors and inducers, to prevent adverse drug events (ADEs).

METHODS

A retrospective cohort study was performed studying DDI alerts for DOAC-CYP3 A4/P-gp influencing drug combination prescriptions in patients admitted to a tertiary care hospital. Patient-, DDI-, and ADE-related data were analyzed to explore CDSS performance and real-world DDI management. A multidisciplinary panel conducted an ADE analysis using the Drug Interaction Probability Scale.

RESULTS

Out of 15,201 triggered CDS alerts, 166 individual alerts were included. Primary indication for DOAC therapy was atrial fibrillation (86.1%). The most involved DOAC was dabigatran (63%). DDI exposure was median 3 days (IQR = 2-7 days). CYP3 A4/P-gp inhibitor DDIs were most prevalent (n = 121), with amiodarone being most prevalently prescribed (71%). Inducers were mainly antiepileptic drugs (n = 31). Thirty-four CDS alerts (20%) were actively shown to the physician upon prescribing (acceptance rate 50%). Eighteen ADEs were identified (11%, 14 bleeding; 4 thromboembolic events), 15 having a possible and 3 a probable probability of being DDI associated.

CONCLUSION

Despite a low number of observed, potentially associated ADEs, CDSS for DDI management aids physicians to optimize DOAC treatment as the described DDIs can cause significant patient harm. Increased ADE monitoring and larger real-world studies are needed to refine CDS tools and further optimize patient safety.

Graph neural network-based subgraph analysis for predicting adverse drug events

PURPOSE

Adverse drug events (ADEs) are a significant global public health concern, and they have resulted in high rates of hospital admissions, morbidity, and mortality. Prior to the use of machine learning and deep learning methods, ADEs may not become well recognized until long after a drug has been approved and is widely used, which poses a significant challenge for ensuring patient safety. Consequently, there is a need to develop computational approaches for earlier identification of ADEs not detected during pre-registration clinical trials.

METHODS

This paper presents a state-of-the-art network-based approach that models patients as subgraphs composed of nodes of International Classification of Diseases (ICD) codes and directed edges illustrating disease progression. Four Graph Neural Network (GNN) variants were employed to make sub-graph level predictions that answer three Research Questions (RQ): 1) whether ADE(s) would occur given a patient's prior diagnoses history, 2) when an ADE would occur, and 3) which ADE would occur. The first and second RQs were addressed using a binary classification approach. The third RQ was addressed using a multi-label classification model.

RESULTS

The proposed network-based approach demonstrated superior performance in predicting ADEs, with the GraphSage model exhibiting the highest accuracy for both RQ 1 (0.8863) and RQ 3 (0.9367), while the Graph Attention Networks (GAT) model was found to perform best for RQ 2 (0.8769). Furthermore, an analysis segmented by ADE classification revealed that while RQs 1 and 3 exhibited minimal variance across different ADE categories, a distinct advantage was observed for categories B, C, and E in the context of RQ 2 when applying this sub-graph method.

CONCLUSION

The network-based approach demonstrates the potential of GNNs in supporting the early detection and prevention of ADEs. Accurately predicting ADEs could enable healthcare professionals to make informed clinical decisions, take preventive measures and adjust medication regimens before serious adverse events occur. The proposed prediction method could also lead to optimized usage of healthcare resources by preventing hospital admissions and reducing the overall burden of adverse drug events on the healthcare systems.

A scoping review of the clinical utility of adverse drug reaction causality analysis tools for use in the hospital setting

INTRODUCTION

Identification and monitoring of adverse drug reactions (ADRs) and interventions to reduce ADRs are essential for patient safety in hospitals. Causality analysis (CA) is an approach that helps to determine a causal link between medication and patient harm (i.e. an ADR). While numerous CA tools exist, there is no gold standard.

AREAS COVERED

Five online databases were searched to identify studies that evaluated the potential clinical utility of CA tools for ADRs. CA tools were mapped against the Bradford Hill (BH) criteria and included if they adhered to the first seven criteria proposed by BH. Upon the database search, 550 studies were identified, with 41 studies being selected that looked at tools mapped to BH. Thirty-four different CA tools were identified in the included studies.

EXPERT OPINION

Naranjo and WHO-UMC were the most reported CA tools for studies examining inter-rater and intra-rater reliability. Naranjo commonly received a 'fair' agreement level while WHO-UMC received a 'substantial' agreement level between raters. Along with kappa statistics, time using the CA tool was also analyzed, with WHO-UMC being the most time-efficient. There does not appear to be one CA tool that can be applied universally to pharmacovigilance efforts in hospital in-patient settings.

ID Consultant: Laboratory Monitoring During Long-Term Use of Oral Antimicrobials in Pediatric Patients

Oral antimicrobials remain the mainstay of long-term treatment for many infections. Meanwhile, the use of oral agents is becoming commonplace for the treatment of several pediatric infections once managed exclusively with parenteral therapies. Unfortunately, antimicrobials are associated with several laboratory toxicities, particularly when high doses or combination therapies are used, but there is a paucity of data on optimal laboratory monitoring strategies. In this ID Consultant article, we offer a summary of the 3 most common laboratory-based toxicities seen with long-term use of oral antimicrobials-drug-induced kidney injury, liver injury, and hematological toxicities-and we provide our recommended approach to monitoring.

Algorithmovigilance, lessons from pharmacovigilance

Artificial Intelligence (AI) systems are increasingly being deployed across various high-risk applications, especially in healthcare. Despite significant attention to evaluating these systems, post-deployment incidents are not uncommon, and effective mitigation strategies remain challenging. Drug safety has a well-established history of assessing, monitoring, understanding, and preventing adverse effects in real-world usage, known as pharmacovigilance. Drawing inspiration from pharmacovigilance methods, we discuss concepts that can be adapted for monitoring AI systems in healthcare. This discussion aims to improve responses to adverse effects and potential incidents and risks associated with AI deployment in healthcare but also beyond.

Development of a text mining algorithm for identifying adverse drug reactions in electronic health records

Objective

Adverse drug reactions (ADRs) are a significant healthcare concern. They are often documented as free text in electronic health records (EHRs), making them challenging to use in clinical decision support systems (CDSS). The study aimed to develop a text mining algorithm to identify ADRs in free text of Dutch EHRs.

Materials and Methods

In Phase I, our previously developed CDSS algorithm was recoded and improved upon with the same relatively large dataset of 35 000 notes (Step A), using R to identify possible ADRs with Medical Dictionary for Regulatory Activities (MedDRA) terms and the related Systematized Nomenclature of Medicine Clinical Terms (SNOMED-CT) (Step B). In Phase II, 6 existing text-mining R-scripts were used to detect and present unique ADRs, and positive predictive value (PPV) and sensitivity were observed.

Results

In Phase IA, the recoded algorithm performed better than the previously developed CDSS algorithm, resulting in a PPV of 13% and a sensitivity of 93%. For The sensitivity for serious ADRs was 95%. The algorithm identified 58 additional possible ADRs. In Phase IB, the algorithm achieved a PPV of 10%, a sensitivity of 86%, and an F-measure of 0.18. In Phase II, four R-scripts enhanced the sensitivity and PPV of the algorithm, resulting in a PPV of 70%, a sensitivity of 73%, an F-measure of 0.71, and a 63% sensitivity for serious ADRs.

Discussion and Conclusion

The recoded Dutch algorithm effectively identifies ADRs from free-text Dutch EHRs using R-scripts and MedDRA/SNOMED-CT. The study details its limitations, highlighting the algorithm's potential and significant improvements.

The Association of Dementia With Incident Adverse Drug Reactions in Hospitalized Older Adults

BACKGROUND

Older adults with dementia commonly receive multiple medications and have higher hospitalization rates, elevating the risk of potentially inappropriate prescribing and in-hospital adverse drug reactions (ADRs). There is limited evidence examining ADRs in older adults with dementia during hospitalization.

OBJECTIVES

Our aim was to assess the association between dementia and incidence of ADRs during hospitalization and to identify prevalent types of ADRs and medications linked to ADRs.

DESIGN

Secondary analysis of the SENATOR trial database, which was a randomized controlled trial of an intervention to reduce ADRs in older inpatients with multimorbidity.

SETTING AND PARTICIPANTS

A total of 1537 patients (47.2% females) with a mean age of 78.1 years were recruited from 6 European hospitals.

METHODS

Sociodemographic data, functional status, cognitive status, clinical information, and ADR-related outcomes were extracted from the SENATOR database. Inpatients with dementia were identified based on prior International Classification of Diseases, Tenth Revision (ICD-10), dementia diagnosis, receiving acetylcholinesterase inhibitors or memantine, or a Mini-Mental State Examination score ≤24 at admission without concurrent delirium.

RESULTS

Among participants, 392 (25.5%) were identified as having dementia. The proportion of patients with probable or certain incident in-hospital ADRs was similar between the groups with and without dementia (22.4% vs 25.4%, P > .05). However, in-hospital rates of probable or certain ADRs from 12 common categories were less frequently identified in patients with dementia compared to those without (19.4% vs 23%, P = .025). Major constipation (6.4% vs 9.9%, P = .03) and acute dyspepsia, nausea, or vomiting (2.8% vs 5%, P = .03) were less commonly observed ADRs in patients with dementia.

CONCLUSIONS AND IMPLICATIONS

We did not observe an increased risk of in-hospital ADRs among inpatients with dementia. However, ADRs related to the gastrointestinal tract and identified by subjective symptoms were less frequently identified in this group. This study lays the groundwork for developing new tools for ADR diagnosis for older patients with dementia.

Rash caused by lurasidone in old chinese patient with bipolar disorder: case-based review

BACKGROUND

Rash is one of common adverse drug reaction and which have been reported in typical and atypical antipsychotics. Reports of lurasidone induced skin reactions are sparse. In this study, we report a case of rash caused by lurasidone.

CASE PRESENTATION

A 63-year-old man with bipolar disorder (BD) who is treated by lurasidone. However, the patient presents a rash all over after lurasidone dose increasing from 40 mg/day to 60 mg/day. With the diagnosis of drug induced rash, lurasidone was discontinued, and the rash complete disappears within 2 weeks. In addition, all case reports about antipsychotics associated rash were reviewed by searching English and Chinese database including Pubmed, Embase, Cochrane Library, CNKI and Wanfang database. A total of 139 articles contained 172 patients were included in our study. The literature review and our case suggest that the cutaneous adverse events caused by antipsychotic drugs should not be ignored, particularly for the patient who was first use or at dose increasing of antipsychotic.

CONCLUSIONS

In conclusion, we report a case of lurasidone related rash and review rash caused by antipsychotics. Psychiatrists should be alert to the possibility of the rash caused by antipsychotics, especially the patient was first use of antipsychotics or the antipsychotic dose was increasing.

The International Working Group on New Developments in Pharmacovigilance: Advancing Methods and Communication in Pharmacovigilance

PURPOSE

In 2019, the International Working Group (IWG), focusing on New Developments in Pharmacovigilance, was established. This group is coordinated by the Drug Safety Research Unit in the United Kingdom, and the mission of the IWG is to progress pharmacovigilance methodologies and promote the safe and effective use of medicines and vaccines, thereby further protecting patients. Novel therapeutics are continuously being developed to alleviate medical conditions, but with advancing technologies, innovative pharmacovigilance methodologies need to be developed to effectively monitor the use and safety of these products. With reduced timelines proposed for premarketing clinical trials and increased application of real-world evidence supporting regulatory approvals, products may be used in real-world clinical practice in shorter timeframes than before. Therefore, the need for effective methods of monitoring medicines and collecting safety data in real-time is of paramount importance to public health.

METHODS

The IWG aims to advance existing methodologies used in the detection, monitoring, and analysis of safety data in pharmacovigilance and to communicate best practice proposals to support decision making in health care. The IWG will identify areas requiring review of current processes or methodologic research and will communicate the output of the IWG through peer-reviewed publications, reports, and presentation of findings at relevant conferences and scientific meetings.

FINDINGS

The IWG is currently reviewing two areas in pharmacovigilance; case-level causality assessment and the strengths and limitations of data sources. The IWG is advancing these areas by producing two scoping reviews which will be easily accessible to regulatory agencies, industry, academia, and interested persons or organizations.

IMPLICATIONS

The scoping reviews comply with the IWGs mission to progress pharmacovigilance methodologies and promote the safe and effective use of medicines and vaccines. The present article shares details of the objectives of the IWG and provides an overview on the status of IWG activities.

Artificial intelligence in pharmacovigilance - Opportunities and challenges

Pharmacovigilance (PV) is a data-driven process to identify medicine safety issues at the earliest by processing suspected adverse event (AE) reports and extraction of health data. The PV case processing cycle starts with data collection, data entry, initial checking completeness and validity, coding, medical assessment for causality, expectedness, severity, and seriousness, subsequently submitting report, quality checking followed by data storage and maintenance. This requires a workforce and technical expertise and therefore, is expensive and time-consuming. There has been exponential growth in the number of suspected AE reports in the PV database due to smart collection and reporting of individual case safety reports, widening the base by increased awareness and participation by health-care professionals and patients. Processing of the enormous volume and variety of data, making its sensible use and separating "needles from haystack," is a challenge for key stakeholders such as pharmaceutical firms, regulatory authorities, medical and PV experts, and National Pharmacovigilance Program managers. Artificial intelligence (AI) in health care has been very impressive in specialties that rely heavily on the interpretation of medical images. Similarly, there has been a growing interest to adopt AI tools to complement and automate the PV process. The advanced technology can certainly complement the routine, repetitive, manual task of case processing, and boost efficiency; however, its implementation across the PV lifecycle and practical impact raises several questions and challenges. Full automation of PV system is a double-edged sword and needs to consider two aspects - people and processes. The focus should be a collaborative approach of technical expertise (people) combined with intelligent technology (processes) to augment human talent that meets the objective of the PV system and benefit all stakeholders. AI technology should enhance human intelligence rather than substitute human experts. What is important is to emphasize and ensure that AI brings more benefits to PV rather than challenges. This review describes the benefits and the outstanding scientific, technological, and policy issues, and the maturity of AI tools for full automation in the context to the Indian health-care system.

Trust but Verify: Lessons Learned for the Application of AI to Case-Based Clinical Decision-Making From Postmarketing Drug Safety Assessment at the US Food and Drug Administration

Adverse drug reactions are a common cause of morbidity in health care. The US Food and Drug Administration (FDA) evaluates individual case safety reports of adverse events (AEs) after submission to the FDA Adverse Event Reporting System as part of its surveillance activities. Over the past decade, the FDA has explored the application of artificial intelligence (AI) to evaluate these reports to improve the efficiency and scientific rigor of the process. However, a gap remains between AI algorithm development and deployment. This viewpoint aims to describe the lessons learned from our experience and research needed to address both general issues in case-based reasoning using AI and specific needs for individual case safety report assessment. Beginning with the recognition that the trustworthiness of the AI algorithm is the main determinant of its acceptance by human experts, we apply the Diffusion of Innovations theory to help explain why certain algorithms for evaluating AEs at the FDA were accepted by safety reviewers and others were not. This analysis reveals that the process by which clinicians decide from case reports whether a drug is likely to cause an AE is not well defined beyond general principles. This makes the development of high performing, transparent, and explainable AI algorithms challenging, leading to a lack of trust by the safety reviewers. Even accounting for the introduction of large language models, the pharmacovigilance community needs an improved understanding of causal inference and of the cognitive framework for determining the causal relationship between a drug and an AE. We describe specific future research directions that underpin facilitating implementation and trust in AI for drug safety applications, including improved methods for measuring and controlling of algorithmic uncertainty, computational reproducibility, and clear articulation of a cognitive framework for causal inference in case-based reasoning.

Comparison of WHO-UMC and Naranjo Scales for Causality Assessment of Reported Adverse Drug Reactions

OBJECTIVES

The 2 most commonly used scales worldwide are the World Health Organization-Uppsala Monitoring Centre (WHO-UMC) and the Naranjo scales. The present study was planned to assess the degree of agreement between the 2 scales when the same adverse drug reactions (ADR) were assessed by 5 raters independently.

METHODS

One hundred individual case safety reports were selected randomly from the ADR database of our institute and the details were emailed to 5 different experts (raters), who were DM Clinical Pharmacology residents from different institutes in India. An independent causality assessment of these ADRs was performed independently by these raters using both the WHO-UMC and Naranjo causality assessment scales. The agreement between the 2 scales was assessed for each rater using Cohen κ, and the overall interrater agreement was assessed using Fleiss κ.

RESULTS

The Cohen κ level of agreement between the 2 scales for the 5 raters were substantial, fair, substantial, moderate, and substantial, respectively. The most common causality assessment category as per WHO-UMC scale was "possible" but varied among the raters on the Naranjo scale. No ADR was categorized as "certain" by any rater on the Naranjo scale. The Fleiss κ value for agreement among the 5 raters was found to be 0.2 (slight) for the WHO-UMC scale and 0.297 (fair) for the Naranjo scale.

CONCLUSIONS

A moderate level of agreement was observed in this study between the WHO-UMC and Naranjo scales. The level of agreement among these 5 raters was found to be similar for the WHO-UMC and the Naranjo scales, indicating a similar degree of subjectivity for the 2 scales. Hence, more robust and less subjective scales are required for causality assessment.

Causality Assessment Between Drugs and Fatal Cerebral Haemorrhage Using Electronic Medical Records: Comparative Evaluation of Disease-Specific and Conventional Methods

INTRODUCTION

A new algorithm for causality assessment of drugs and fatal cerebral haemorrhage (ACAD-FCH) was published in 2021. However, its use in clinical practice has not been verified.

OBJECTIVES

This study aimed to explore the practical value of the ACAD-FCH when applying information available in clinical practice.

METHODS

The medical records of patients who died at the University of Tokyo Hospital in 2020 were reviewed, and cases with intracranial haemorrhage were selected. Two evaluators independently assessed these cases using three methods (the ACAD-FCH, Naranjo algorithm, and WHO-UMC scale). The number of 'Yes', 'No', and 'No information/Do not know' responses to each question by both evaluators were summed and compared. Inter-rater reliability was evaluated for each method using agreement rates and kappa coefficients with 95% confidence intervals (CI).

RESULTS

Among 316 deaths, 24 cases with intracranial haemorrhage were evaluated. The proportion of ‛No information/Do not know' responses for each question was 35.6% (95% CI 31.4-40.6%) for the ACAD-FCH and 66.9% (95% CI 62.5-71.1%) for the Naranjo algorithm. The respective agreement rates and kappa coefficients were 0.917 (0.798-1.00) and 0.867 (0.675-1.00) for the ACAD-FCH, 0.708 (0.512-0.904) and 0.139 (-0.236 to 0.513) for the Naranjo algorithm, and 0.50 (0.284-0.716) and 0.326 (0.110-0.541) for the WHO-UMC scale, respectively.

CONCLUSION

Our findings suggest the utility of the ACAD-FCH when assessing death cases with intracranial haemorrhage. However, larger studies including intra-rater assessments are warranted for further validation of this algorithm.

Causality, Severity, Preventability and Predictability Assessments Scales for Adverse Drug Reactions: A Review

The pharmacovigilance program of India (PvPI), after its inception, has been reliably acquiring force in bringing issues to light among the masses, healthcare professionals, the pharma industry, and clinical staff at hospitals. Adverse drug reactions are unintended events that occur after exposure to a drug, biological product, or medical device, and they may result in morbidity and mortality. It is critical to monitor the safety of drugs during the post-marketing phase to find long-term and rare ADRs, as well as ADRs in special populations and patients with co-morbidities that are not usually included during clinical trials. The definitive objective of pharmacovigilance is to collate data and analyze it. Assessing the causality between ADRs and drugs is necessary to decrease the occurrence of ADRs and to reduce the risk of drug-related ADRs. ADRs may lead to increased morbidity, increased hospital stays, and increased cost of treatment, resulting in compromised patient safety. Causality assessment is the evaluation of the likelihood that a particular treatment is the cause of an observed adverse event and establishing a causal association between a drug and a drug reaction is necessary to prevent further recurrences. Numerous methods available for establishing a causal association between the drug and adverse events have been broadly classified into clinical judgment or global introspection, algorithms, and probabilistic methods. These include the Swedish method, World Health Organization-Uppsala Monitoring Centre (WHO-UMC) scale, Naranjo's algorithm, Kramer algorithm, Jones algorithm, Karch algorithm, Bégaud algorithm, Adverse Drug Reactions Advisory Committee guidelines, Bayesian Adverse Reaction Diagnostic Instrument, and so on. Despite various methods available, none of the causality assessment tools have been universally accepted as the gold standard. Naranjo's algorithm and WHO-UMC scales are, however, the most commonly used. Similarly, for preventability and severity assessment of ADRs, the Schumock and Thornton scale and Hartwig and Siegel's scale are most commonly used. Hence, we reviewed different tools and methods available to assess the causality, preventability, and severity of ADRs.

Inter-rater agreement between WHO- Uppsala Monitoring Centre system and Naranjo algorithm for causality assessment of adverse drug reactions

Determining the causality of Adverse Drug Reactions (ADRs) is essential for management and prevention of future occurrences. The WHO-Uppsala Monitoring Centre (UMC) system is recommended under the Pharmacovigilance Program of India whereas Naranjo's algorithm is commonly utilized by clinicians, but their agreement remains a subject of investigation. This study aims to compare the inter-rater agreement between these two scales for causality assessment of ADRs. In this cross-sectional study, two groups of pharmacovigilance experts were given a set of total 399 anonymized individual case safety reports, collected over six months. The raters were blinded to each other's assessments and applied the WHO-UMC system and Naranjo algorithm to each case independently. Inter-rater agreement was then evaluated utilizing Cohen's kappa. The suspected ADRs were also comprehensively analysed on parameters like age, sex, route of administration, speciality, organ system affected, most common drug categories and individual drugs, outcome of ADRs. Analysis of 399 suspected ADRs revealed that mean age of patients was 36.8 ± 18.0 years, females were more frequently affected, highest proportion of reports were from psychiatry inpatients, seen with antipsychotic drugs, involved the central nervous system, with oral administration, and 91% resolved. On causality assessment by the WHO-UMC system, 53.3% were "Certain" whereas Naranjo's algorithm categorized 96.74% of ADRs as "Probable". Cohen's kappa showed a "Minimal" agreement (0.22) between WHO-UMC and Naranjo system of causality assessment. The considerable lack of agreement between the two commonly employed systems of causality assessment of ADRs warrants further investigation into specific factors influencing the disagreement to improve the accuracy of causality assessments.

A retrospective study of the efficacy and safety of levofloxacin in children with severe infection

Objectives

Levofloxacin is widely used because of its broad-spectrum antimicrobial activity and convenient dosing schedule. However, the relevance of its use in children remains to be investigated. The purpose of this study is to investigate the efficacy and safety of levofloxacin use in children with severe infections.

Methods

We conducted a retrospective observational study of patients <18 years of age who received levofloxacin intravenously in the Pediatric Intensive Care Unit (PICU) of our hospital during the period between 2021 and 2022. Patient demographics, course characteristics, clinical effectiveness, and adverse event correlations were extracted through a retrospective tabular review.

Results

We included 25 patients treated with 28 courses of levofloxacin. The mean age of these children treated with levofloxacin was 4.41 years. Conversion of pathogenic microbiological test results to negative after levofloxacin treatment was detected in 11 courses (39.29%). A decrease in inflammatory markers, white blood cell or C-reactive protein counts, was detected in 18 courses (64.29%). A total of 57 adverse events occurred during the treatment period, of which 21 were possibly related to levofloxacin and no adverse events were probably related to levofloxacin.

Conclusion

The effectiveness of levofloxacin use in children with serious infections is promising, especially for the treatment of multidrug-resistant bacteria. Adverse events occurring during the initiation of levofloxacin therapy in children are reported to be relatively common, but in this study, only a small percentage of them were possibly related to levofloxacin, and none of them were highly possibly related to levofloxacin.

Virus-Induced Epilepsy vs. Epilepsy Patients Acquiring Viral Infection: Unravelling the Complex Relationship for Precision Treatment

The intricate relationship between viruses and epilepsy involves a bidirectional interaction. Certain viruses can induce epilepsy by infecting the brain, leading to inflammation, damage, or abnormal electrical activity. Conversely, epilepsy patients may be more susceptible to viral infections due to factors, such as compromised immune systems, anticonvulsant drugs, or surgical interventions. Neuroinflammation, a common factor in both scenarios, exhibits onset, duration, intensity, and consequence variations. It can modulate epileptogenesis, increase seizure susceptibility, and impact anticonvulsant drug pharmacokinetics, immune system function, and brain physiology. Viral infections significantly impact the clinical management of epilepsy patients, necessitating a multidisciplinary approach encompassing diagnosis, prevention, and treatment of both conditions. We delved into the dual dynamics of viruses inducing epilepsy and epilepsy patients acquiring viruses, examining the unique features of each case. For virus-induced epilepsy, we specify virus types, elucidate mechanisms of epilepsy induction, emphasize neuroinflammation's impact, and analyze its effects on anticonvulsant drug pharmacokinetics. Conversely, in epilepsy patients acquiring viruses, we detail the acquired virus, its interaction with existing epilepsy, neuroinflammation effects, and changes in anticonvulsant drug pharmacokinetics. Understanding this interplay advances precision therapies for epilepsy during viral infections, providing mechanistic insights, identifying biomarkers and therapeutic targets, and supporting optimized dosing regimens. However, further studies are crucial to validate tools, discover new biomarkers and therapeutic targets, and evaluate targeted therapy safety and efficacy in diverse epilepsy and viral infection scenarios.

Navigating Drug-Induced Adversities: A Python-Based Console Application for Causality Assessment Using the Naranjo Algorithm

The timely and accurate adverse drug reactions (ADR) assessment is vital for effective patient management and healthcare delivery. The Naranjo Algorithm is a widely recognized tool for determining the probability that a drug induces a given ADR. However, the process can be time-consuming and susceptible to human error. This study introduces a Python-based console application (Python Software Foundation, Wilmington, Delaware, United States) designed to automate the Naranjo Algorithm for ADR causality assessment. The application was developed using Python 3.11.4 on a Windows 11 system (Microsoft Corporation, Redmond, Washington, United States) and compiled in Notepad (Microsoft Corporation), a basic text editor, highlighting its accessibility and ease of use in various settings. User input is solicited for each question in the Naranjo Algorithm, validated for acceptable entries, and subsequently scored. The final score categorizes the reaction into Doubtful, Possible, Probable, or Definite ADR, facilitating rapid clinical decision-making. Preliminary validation shows promising reliability and effectiveness, making it a valuable asset in research and clinical settings for assessment.

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.

Intensive care drug therapy and its potential adverse effects on blood pressure and heart rate in critically ill children

BACKGROUND

Owing to complex treatment, critically ill children may experience alterations in their vital parameters. We investigated whether such hemodynamic alterations were temporally and causally related to drug therapy.

METHODS

In a university pediatric intensive care unit, we retrospectively analyzed hemodynamic alterations defined as values exceeding the limits set for heart rate (HR) and blood pressure (BP). For causality assessment, we used the World Health Organization-Uppsala Monitoring Center (WHO-UMC) system, which categorizes the probability of causality as "certain," "probable," "possible," and "unlikely."

RESULTS

Of 315 analyzed patients with 43,200 drug prescriptions, 59.7% experienced at least one hemodynamic alteration; 39.0% were affected by increased HR, 19.0% by decreased HR, 18.1% by increased BP, and 16.2% by decreased BP. According to drug information databases, 83.9% of administered drugs potentially lead to hemodynamic alterations. Overall, 88.3% of the observed hemodynamic alterations had a temporal relation to the administration of drugs; in 80.2%, more than one drug was involved. Based on the WHO-UMC system, a drug was rated as a "probable" causing factor for only 1.4% of hemodynamic alterations. For the remaining alterations, the probability ratings were lower because of multiple potential causes, e.g., several drugs.

CONCLUSIONS

Critically ill children were frequently affected by hemodynamic alterations. The administration of drugs with potentially adverse effects on hemodynamic parameters is often temporally related to hemodynamic alterations. Hemodynamic alterations are often multifactorial, e.g., due to administering multiple drugs in rapid succession; thus, the influence of individual drugs cannot easily be captured with the WHO-UMC system.

Causality assessment of adverse drug reactions in neonates: a comparative study between Naranjo's algorithm and Du's tool

BACKGROUND

Algorithms for causality assessment of adverse drug reactions (ADRs) in a neonatal intensive care unit (NICU) are important in the management of adverse events, however, it is inconclusive which tool best suits pharmacovigilance in neonates.

AIM

To compare the performance of the algorithms of Du and Naranjo in determining causality in cases of ADRs in neonates in a NICU.

METHOD

This observational and prospective study was conducted in a NICU of a Brazilian maternity school between January 2019 and December 2020. Independently, three clinical pharmacists used the algorithms of Naranjo and Du in 79 cases of ADRs in 57 neonates. The algorithms were evaluated for inter-rater and inter-tool agreement using Cohen's kappa coefficient (k).

RESULTS

The Du algorithm showed greater ability to identify definite ADRs (≈ 60%), but had low reproducibility (overall k = 0.108; 95% CI 0.064-0.149). In contrast, the Naranjo algorithm showed a lower proportion of definite ADRs (< 4%), but had good reproducibility (overall k = 0.402; 95% CI 0.379-0.429). The tools showed no significant correlation regarding ADR causality classification (overall k = - 0.031; 95% CI - 0.049 to 0.065).

CONCLUSION

Although the Du algorithm has a lower reproducibility compared to the Naranjo, this tool showed good sensitivity for classifying ADRs as definite, proving to be a more suitable tool for neonatal clinical routine.

Causality assessment of adverse drug reaction: A narrative review to find the most exhaustive and easy-to-use tool in post-authorization settings

BACKGROUND

The core motive of pharmacovigilance is the detection and prevention of adverse drug reactions (ADRs), to improve the risk-benefit balance of the drug. However, the causality assessment of ADRs remains a major challenge among clinicians, and none of the available tools of causality assessment used for assessing ADRs have been universally accepted.

OBJECTIVE

To provide an up-to-date overview of the different causality assessment tools.

METHODS

We conducted electronic searches in MEDLINE, EMBASE, and the Cochrane database. The eligibility of each tool was screened by three reviewers. Each eligible tool was then scrutinized for its domains (the reported specific set of questions/areas used for calculating the likelihood of cause-and-effect relation of an ADR) to discover the most comprehensive tool. Finally, we subjectively assessed the tool's ease-of-use in a Canadian, Indian, Hungarian, and Brazilian clinical context.

RESULTS

Twenty-one eligible causality assessment tools were retrieved. Naranjo's tool and De Boer's tool appeared the most comprehensive among all the tools, covering 10 domains each. Regarding "ease-of-use" in a clinical setting, we judged that many tools were hard to implement in a clinical context because of their complexity and/or lengthiness. Naranjo's tool, Jones's tool, Danan and Benichou's tool, and Hsu and Stoll's tool appeared to be the easiest to implement into various clinical contexts.

CONCLUSION

Among the many tools identified, 1981 Naranjo's scale remains the most comprehensive and easy to use for performing causality assessment of ADRs. Upcoming analysis should compare the performance of each ADR tool in clinical settings.

Important Considerations for Signal Detection and Evaluation

Safety clinicians have a wealth of resources describing how to perform signal detection. Nevertheless, there are some nuances concerning approaches taken by regulatory authorities and statistical considerations that should be appreciated. New approaches, such as the FDA Medical Queries, illustrate the value of considering medical concepts over individual adverse events. One area which would benefit from further clarity is how safety signals may be evaluated for evidence of a causal relationship to the drug of interest. Just as such safety signals can take many forms, the types of tools and methods required to interrogate these signals are equally as diverse. An understanding of the complexity of this process can aid the safety reviewer in successfully characterizing the emerging safety profile of a drug during the pre-marketing phase of development.

Recurring Fatigue After Biologic Administration: Patient-Reported Data from the Dutch Biologic Monitor

BACKGROUND

Fatigue is a common problem in immune-mediated inflammatory disease (IMID) patients, significantly impacting their quality of life.

OBJECTIVES

In this study, we describe the pattern and characteristics of fatigue as a patient-reported adverse drug reaction (ADR) of biologics, and compared patient and treatment characteristics with patients reporting other ADRs or no ADRs.

METHODS

In this cohort event monitoring study, the description and characteristics of fatigue reported as a possible ADR in the Dutch Biologic Monitor were assessed and analysed for commonly recurring themes or patterns. Baseline and treatment characteristics of patients with fatigue and patients reporting other ADRs or no ADRs were compared.

RESULTS

Of 1382 participating patients, 108 patients (8%) reported fatigue as an ADR of a biologic. Almost half of these patients (50 patients, 46%) described episodes of fatigue during or shortly after biologic injection, which often recurred following subsequent injections. Patients with fatigue were significantly younger than patients with other ADRs or patients without ADRs (median age for patients with fatigue, 52 years; median age for patients with other ADRs, 56 years; and median age for patients without ADRs, 58 years); significantly more often smoked (25% vs. 16% and 15%); used infliximab (22% vs. 9% and 13%), rituximab (9% vs. 3% and 1%) or vedolizumab (6% vs. 2% and 1%); and significantly more often had Crohn's disease (28% vs. 13% and 13%) and other comorbidities (31% vs. 20% and 15%). Patients with fatigue significantly less frequently used etanercept (12% vs. 29% and 34%) or had rheumatoid arthritis (30% vs. 45% and 43%).

CONCLUSIONS

IMID patients may experience fatigue as a postdosing effect of biologics.

Medication-related Medical Emergency Team activations: a case review study of frequency and preventability

OBJECTIVES

Despite recognition of clinical deterioration and medication-related harm as patient safety risks, the frequency of medication-related Rapid Response System activations is undefined. We aimed to estimate the incidence and preventability of medication-related Medical Emergency Team (MET) activations and describe the associated adverse medication events.

METHODS

A case review study of consecutive MET activations at two acute, academic teaching hospitals in Melbourne, Australia with mature Rapid Response Systems was conducted. All MET activations during a 3-week study period were assessed for a medication cause including identification of the contributing adverse medication event and its preventability, using validated tools and recognised classification systems.

RESULTS

There were 9439 admissions and 628 MET activations during the study period. Of these, 146 (23.2%) MET activations were medication related: an incidence of 15.5 medication-related MET activation per 1000 admissions. Medication-related MET activations occurred a median of 46.6 hours earlier (IQR 22-165) in an admission than non-medication-related activations (p=0.001). Furthermore, this group also had more repeat MET activations during their admission (p=0.021, OR=1.68, 95% CI 1.09 to 2.59). A total of 92 of 146 (63%) medication-related MET activations were potentially preventable. Tachycardia due to omission of beta-blocking agents (10.9%, n=10 of 92) and hypotension due to cumulative toxicity (9.8%, n=9 of 92) or inappropriate use (10.9%, n=10 of 92) of antihypertensives were the most common adverse medication events leading to potentially preventable medication-related MET activations.

CONCLUSIONS

Medications contributed to almost a quarter of MET activations, often early in a patient's admission. One in seven MET activations were due to potentially preventable adverse medication events. The most common of these were omission of beta-blockers and clinically inappropriate antihypertensive use. Strategies to prevent these events would increase patient safety and reduce burden on the MET.

Does a steroid dementia syndrome really exist? A brief narrative review of what the literature highlights about the relationship between glucocorticoids and cognition

Glucocorticoids (GCs) may cause cognitive impairment through complex pathways involving specific receptors. In the human brain, hippocampal CA1 neurons exhibit the highest level of GCs receptors. Even the elderly expressed these receptors. The purpose of this brief review is to concentrate on the relationship between GCs and cognition in order to discuss the effects of the so-called steroid dementia in routine clinical practice.

Adverse drug reactions in an ageing PopulaTion (ADAPT) study: Prevalence and risk factors associated with adverse drug reaction-related hospital admissions in older patients

Background: Older people experience greater morbidity with a corresponding increase in medication use resulting in a potentially higher risk of adverse drug reactions (ADRs). Objectives: The aim of this study was to; 1) determine the prevalence and characteristics of ADR-related hospital admissions among older patients (≥65 years) in Ireland; and 2) identify the risk factors associated with ADR-related hospital admissions. Methods: A cross-sectional study of ADR prevalence in patients aged ≥65 years admitted acutely to hospital in Ireland over a 8 month period (November 2016- June 2017). A multifaceted review of each hospital admission was undertaken to assess the likelihood of an ADR being a reason for admission (cause of admission or contributing to admission) in the context of the patient's medication, clinical conditions, comorbidities and investigations. A number of decision aids were applied by two independent reviewers to assess ADR causality, avoidability and severity. A random sample of patients, determined not to have a suspected ADR on screening, were assigned to a non-ADR control group. Multivariable logistic regression was used to assess the association between potential risk factors for ADR-related admissions compared with non-ADR-related admissions. Results: In total, 3,760 hospital admission episodes (in 3,091 patients) were screened and 377 admissions were considered ADR-related (10.0%, 95% CI 9.1%, 11.0%). 219 (58.1%) ADR-related admissions were caused by an ADR, while ADRs contributed to 158 (41.9%) admissions. 268 (71.1%) of all ADR-related admissions were deemed definitely or possibly preventable/avoidable. 350 (92.8%) ADRs were classified as being of moderate severity, with 27 (7.2%) classified as severe. Antithrombotic agents, mainly aspirin and warfarin, were the drugs most frequently associated with ADR-related admissions (gastrointestinal and vascular haemorrhagic disorders). In multivariable analysis, immobility, frailty, having delirium or ulcer disease and taking anticoagulant and antiplatelet medication on admission were significantly associated with an ADR-related hospital admission. Conclusion: One in ten hospital admissions, among those aged 65 + years, were considered ADR-related, with approximately 70% potentially avoidable. Reliable and validated ADR detection and prediction tools are needed to develop prevention strategies.

Nonsteroidal anti-inflammatory drug exposure and the risk of microscopic colitis

BACKGROUND

Medication consumption has been suggested as a risk factor for microscopic colitis (MC), but studies of varying design have yielded inconsistent results. Our aim was to evaluate the association between medications and MC.

METHODS

A hybrid cohort of prospectively identified patients undergoing colonoscopy with biopsies for suspicion of MC (N = 144) and patients with MC enrolled within three months of diagnosis into an MC registry (N = 59) were surveyed on medication use. Medication use was compared between patients with and without diagnosis of MC by chi-squared test and binomial logistic regression adjusted for known risk factors of MC: age and gender.

RESULTS

In total, 80 patients with MC (21 new, 59 registry) were enrolled. Patients with MC were more likely to be older (p = 0.03) and female (p = 0.01) compared to those without MC. Aspirin and other non-steroidal anti-inflammatory drugs were more commonly used among patients who developed MC (p < 0.01). After controlling for age and gender, these medications remained independent predictors of MC with odds ratio for any non-steroidal anti-inflammatory drug use of 3.04 (95% CI: 1.65-5.69). No association between MC and other previously implicated medications including proton pump inhibitors and selective serotonin reuptake inhibitors was found.

CONCLUSIONS

In this cohort of patients with chronic diarrhea, we found use of aspirin and non-steroidal anti-inflammatory drugs, but not other implicated medications to be associated with the development of MC. Whether these drugs trigger colonic inflammation in predisposed hosts or worsen diarrhea in undiagnosed patients is unclear. However, we feel that these findings are sufficient to discuss potential non-steroidal anti-inflammatory drug cessation in patients newly diagnosed with MC.

Shortcomings of Administrative Data to Derive Preventive Strategies for Inhospital Drug-Induced Acute Kidney Failure—Insights from Patient Record Analysis

Structured analyses of hospital administrative data may detect potentially preventable adverse drug events (ADE) and therefore are considered promising sources to prevent future harm and estimate cost savings. Whether results of these analyses indeed correspond to ADE that may be preventable in clinical routines needs to be verified. We exemplarily screened all adult inpatients admitted to a German University Hospital (n = 54,032) for International Classification of Diseases-10th revision (ICD-10) diagnoses coding for drug-induced kidney injury (AKI). In a retrospective chart review, we checked the coded adverse events (AE) for inhospital occurrence, causality to drug exposure, and preventability in all identified cases and calculated positive predictive values (ppv). We identified 69 inpatient cases of whom 41 cases (59.4%) experienced the AE in the hospital (ppv-range 0.43–0.80). Causality assessment revealed a rather likely causal relationship between AE and drug exposure in 11 cases (15.9, 11/69, ppv-range 0.17–0.22) whereby preventability measures could be postulated for seven cases (10.1%, 7/69). Focusing on drug-induced AKI, this study exemplarily underlines that ICD-10-code-based ADE prevention efforts are quite limited due to the small identification rate and its high proportion of primarily outpatient events. Furthermore, causality assessment revealed that cases are often too complex to benefit from generic prevention strategies. Thus, ICD-10-code-based calculations might overestimate patient harm and economic losses.

Prevalence of iatrogenic CNS inflammation at a tertiary neuroimmunology clinic

BACKGROUND

Various vaccines, tumor-necrosis-factor-alpha inhibitors (TNFAIs), immune-checkpoint inhibitors (ICIs), and other immunomodulators have been linked to inflammatory CNS events. The prevalence of iatrogenic events in the neuroimmunology clinic is unknown.

OBJECTIVE

To evaluate the prevalence and clinical characteristics of iatrogenic CNS inflammation in a tertiary neuroimmunology clinic.

METHODS

We analyzed 422 consecutive patients seen over five years at a tertiary neuroimmunology clinic who were systematically screened for exposure to vaccines, TNFAIs, ICIs, or other immunomodulators. In patients with suspected iatrogenic events, the Naranjo Adverse Drug Reaction Probability Scale was used to score the probability of iatrogenicity.

RESULTS

In total, 27 potential iatrogenic events were observed, accounting for 6.4% of all new referrals. The average Naranjo score was 5.78 +/- 1.65 with 74% of the cases scored as probable and 26% scored as possible. The clinical phenotypes included MS relapses (37%); autoimmune encephalitis (30%); NMOSD attacks (15%); transverse myelitis (11%); optic neuritis (4%); and MOGAD attacks (4%). A monophasic course was observed in 44% of cases while 41% had a relapsing course. All patients stopped or interrupted treatment with the offending agent. In addition, 41% of the iatrogenic events were fully responsive to corticosteroids; 22% were partially responsive; and 15% resolved spontaneously. The most common potential triggers were vaccines (37%) followed by TNFAIs (33%) then ICIs (26%). A significantly higher number of probable iatrogenic events were observed among the ICI and vaccine groups compared to a higher number of possible events among the TNFAI group. The latter group also had a significantly longer interval since exposure. The ICI group was more likely to present with monophasic autoimmune encephalitis.

CONCLUSION

Iatrogenic CNS inflammation is rare and typically involves steroid-responsive monophasic events. A subset of iatrogenic events can unmask or worsen relapsing disorders. The probability of iatrogenicity was higher in vaccine and ICI-related events compared to TNFAI-related events.

Recurrent pneumothorax in a patient with liposarcoma as either a complication of lung micrometastasis or a potential adverse event from antibiotic therapy: A case report

Spontaneous pneumothorax (PNX) is an infrequent manifestation of primary lung cancer, soft tissue sarcoma and metastasis. There are no easily accessible data in the literature regarding the correlation between PNX and antibiotics, whereas cases of PNX following chemotherapy have been observed. Only 1-10% of treatment-related adverse events are estimated to be reported to the Food and Drug Administration. The present study described a case of PNX of the left lung in a 70-year-old treatment-naive patient with retroperitoneal liposarcoma. The PNX developed after 8 days of treatment with levofloxacin and after 6 days of piperacillin/tazobactam treatment for a suspicious inflammatory area in the right lung detected by an FDG-PET scan before the patient started chemotherapy. A chest CT scan confirmed the presence of metastasis in the right lung, but neither FDG-PET/CT nor CT showed metastatic disease in the left lung. A total of 14 days after the end of the third cycle of doxorubicin (2 months after the initial diagnosis of PNX), the patient manifested a massive PNX of the right lung. In conclusion, these findings indicated that spontaneous PNX could be linked to the use of some antibiotics.

Toward Kidney-Specific Causality Assessment Tool

PURPOSE

Current nonspecific causality assessment tools lack the assessment of drug-induced acute kidney injury (DIAKI). We recently published an editorial letter for developing a specific causality assessment tool for DIAKI. The purpose of the present review was to suggest the possible required parameters and outline the path to developing a kidney-specific causality assessment tool (KSCAT).

METHODS

A stepwise approach for developing a KSCAT is important as this will be first version of this new tool. Thus, as a first step, we performed a screening of previously published articles on nonspecific and liver-specific causality assessment tools to define possible parameters. The suggested parameters for KSCAT fall into 3 categories: (1) drug-related; (2) kidney-related; and (3) terminology. A tri-polar method was then created that consists of definitive adverse drug reactions (ADRS), terminology, and without ADRS to suggest that the new KSCAT be efficient, specific, user friendly, and less time-consuming. Finally, a pyramid model is suggested to offer the perspectives of experts in the fields of pharmacovigilance, pharmacoepidemiology, and nephrology, as well as decision makers, while developing a KSCAT.

FINDINGS

Causality assessment tools, either nonspecific or organ-specific, fall into 3 categories: (1) expert judgment; (2) algorithms; and (3) probabilistic methods. None of the current causality assessment tools is sufficient for assessing the causality of kidney-related ADRs and for screening the expanded definition of ADR included in European Union Directive 2010/84/EU.

IMPLICATIONS

The causal relationship between drug(s) and DIAKI may be difficult and may not be assessed appropriately with the use of nonspecific tools or approaches. The aim of this article was to reiterate the need for KSCAT development and to propose the associated steps by stating the main principles: namely, the definition of ADR, suggested parameters to be included in the KSCAT, and integration of technology. Our ultimate desire is to invite experts to develop this new tool using an interdisciplinary approach and to benefit from our review in pursuing the next steps. The development of a KSCAT should start with regular and interdisciplinary consortium meetings of experts; the tool should then be tested for its usability, specificity, and practicality; and, finally, it should be used in real-life pharmacovigilance practices, as well as in research by health authorities, regulators, decision-makers, scientists, and clinicians. A KSCAT would support the provision of reliable and reproducible measures of the relationship likelihood in suspected cases of ADR to overcome uncertainty and provide a standardized approach. (Clin Ther. 2022;44:XXX-XXX) © 2022 Elsevier HS Journals, Inc.

Actual drug-related harms in residential aged care facilities: a narrative review

INTRODUCTION

Older people in residential aged care facilities (RACFs) have a high risk of safety issues and concerns about the potential quality of care received. This narrative review investigates the types of actual drug-related harms, their prevalence, reporting of any standard definitions for these harms, and their identification methods.

AREAS COVERED

The authors conducted a systematic search on Ovid Embase, Ovid Medline, and PubMed from March 2001 to March 2021. This narrative review included all types of studies targeting aged care residents aged 65 years and above with actual drug-related harms.

EXPERT OPINION

The prevalence of actual drug-related harms in residents ranged from 0.07% to 63.0%. Falls, drug-drug interactions, neuropsychiatric symptoms, anaphylaxis, urinary tract infection, hypoglycemia, hypokalaemia, and acute kidney injury are the most common drug-related harms in older residents. Psychotropic drugs are the most common drug class implicated in these harms. Evidence related to the association between individual psychotropic drugs and injury, or harm is also lacking. Due to the variation in study duration, reported prevalence, identification methods, and absence of a definition for actual drug-related harms in most studies, further research is mandated to understand the prevalence and clinical implications of drug-related harms in older residents.

Adverse Drug Reactions at Nonelective Hospital Admission in Children and Adolescents: Comparison of 4 Causality Assessment Methods

OBJECTIVES

This study aimed to compare assessment methods to determine adverse drug reactions (ADRs) at nonelective hospital admission in pediatric patients, to investigate the interrater reliability of assessment methods in pediatric care, and to analyze symptoms related to ADRs and (suicidal) drug intoxications.

METHODS

For 1 year, the medical records of nonelective patients admitted to a university pediatric department were evaluated for potential ADRs using 4 assessments methods by 1 experienced rater. Krippendorff α was calculated from a sample of 14 patients evaluated by 4 experienced raters to determine interrater reliability.

RESULTS

In 1831 nonelective hospital admissions, 63.4% (1161 of 1831) of patients had received at least one drug before admission. We found a potential causal relationship between drugs and symptoms documented at admission and thus potential ADRs according to Naranjo in 23.3% (271 of 1161) of those patients, World Health Organization - Uppsala Monitoring Centre (WHO-UMC) in 22.5% (261 of 1161), Koh in 21.7% (252 of 1161), and Begaud in 16.5% (192 of 1161). The probability rating of the potential causal relationships varied considerably between the methods (Naranjo-Begaud, P < 0.01; Naranjo-Koh, P < 0.001; Koh-Begaud, P < 0.01; Begaud-WHO-UMC, P < 0.01). Acceptable interrater reliability (α ≥ 0.667) was only obtained for WHO-UMC (α = 0.7092). The most frequently identified definite ADR was sedation in 1.5% of all nonelective patients with medication before hospital admission. In 1.2% (22 of 1831) of all nonelective admissions, we found drug intoxications with suicidal intent.

CONCLUSIONS

The assessment methods showed a high variability in the determination of a potential causal relationship between drug and documented symptom, in the classification of the probability of ADRs, and suboptimal interrater reliability. Thus, their feasibility in pediatric patients is limited.

Performance of a trigger tool for detecting adverse drug reactions in patients with polypharmacy acutely admitted to the geriatric ward

Aim

To investigate the performance of an adverse drug reaction (ADR) trigger tool in patients with polypharmacy acutely admitted to our geriatric ward.

Findings

The ADR trigger tool had a positive predictive value (PPV) of 41.8%. Usual care recognised 83.5% of ADRs considered as possible, probable or certain, increasing to 97.1% when restricted to probable and certain ADRs.

Message

It is unlikely that implementation of the ADR trigger tool will improve detection of unrecognised ADRs in older patients acutely admitted to our geriatric ward.

Supplementary Information

The online version contains supplementary material available at 10.1007/s41999-022-00649-x.

Purpose

Adverse drug reactions (ADRs) account for 10% of acute hospital admissions in older people, often under-recognised by physicians. The Dutch geriatric guideline recommends screening all acutely admitted older patients with polypharmacy with an ADR trigger tool comprising ten triggers and associated drugs frequently causing ADRs. This study investigated the performance of this tool and the recognition by usual care of ADRs detected with the tool.

Methods

A cross-sectional study was performed in patients ≥ 70 years with polypharmacy acutely admitted to the geriatric ward of the University Medical Centre Utrecht. Electronic health records (EHRs) were screened for trigger–drug combinations listed in the ADR trigger tool. Two independent appraisers assessed causal probability with the WHO-UMC algorithm and screened EHRs for recognition of ADRs by attending physicians. Performance of the tool was defined as the positive predictive value (PPV) for ADRs with a possible, probable or certain causal relation.

Results

In total, 941 trigger–drug combinations were present in 73% (n = 253/345) of the patients. The triggers fall, delirium, renal insufficiency and hyponatraemia covered 86% (n = 810/941) of all trigger–drug combinations. The overall PPV was 41.8% (n = 393/941), but the PPV for individual triggers was highly variable ranging from 0 to 100%. Usual care recognised the majority of ADRs (83.5%), increasing to 97.1% when restricted to possible and certain ADRs.

Conclusion

The ADR trigger tool has predictive value; however, its implementation is unlikely to improve the detection of unrecognised ADRs in older patients acutely admitted to our geriatric ward. Future research is needed to investigate the tool’s clinical value when applied to older patients acutely admitted to non-geriatric wards.

Supplementary Information

The online version contains supplementary material available at 10.1007/s41999-022-00649-x.

Lab-Based Culprit Drug Identification Methods for Cutaneous Drug Eruptions: A Scoping Review

BACKGROUND

Identification of culprit drugs when managing cutaneous drug eruptions is essential. Causality assessment methods (CAMs) have been proposed, including lab-based techniques. However, no consensus guidelines exist.

OBJECTIVES

To identify and map the functionality and feasibility of lab-based CAMs.

METHODS

A scoping review was conducted to identify culprit drug identification methods. Publications on lab-based methods were analyzed. Medline, Embase, and Cochrane Central Register of Controlled Trials databases were searched.

RESULTS

Twenty-five publications met inclusion criteria. Nine lab-based CAMs were studied, including lymphocyte transformation test, cytokine measurement (ELISpot, ELISA, beads array assay), modified IFN-ɣ ELISpot, CellScan, histamine release, granzyme B-ELISpot, intracellular granulysin, lymphocyte toxicity assay, and HLA allele genotyping. Diagnostic accuracy was reported for 8/9 methods. Clinical assessment and operational algorithms were commonly used as validation benchmarks. Lab-based methods were assessed at different phases of a drug eruption including in the acute (18.1%), recovery (27.3%), acute and recovery (27.3%), or an unspecified phase (27.3%). Lymphocyte transformation test (specificity 30% to 100%, sensitivity 27% to 73%) and cytokine measurement (specificity 76% to 100%, sensitivity 20% to 84%) were the most common methods studied.

CONCLUSIONS

Lab-based CAMs can be low-risk, effective, and complementary of clinical methods. High-quality studies are needed to adequately develop and validate these tools for clinical practice.

Quality of Spontaneous Reports of Adverse Drug Reactions Sent to a Regional Pharmacovigilance Unit

Spontaneous reports (SRs) of adverse drug reactions (ADRs) remain the basis of pharmacovigilance systems. The main objective of this study was to evaluate the quality of SRs received by the Pharmacovigilance Unit of Beira Interior, in Central Portugal. The second objective was to identify factors associated with complete SRs. SRs received between 1 January 2017 and 31 October 2019 were analyzed. SR information was classified as “mandatory” or “recommended” criteria. SR were then grouped into three categories (well, slightly, and poorly documented). Association between “well documented” SR and confounding variables was estimated using a multiple logistic regression model. The results showed 22.4% of SRs are “well documented”, and 41.2% are “poorly documented”. Most of the complete SRs correspond to non-serious ADRs (55.8%), with a negative association between complete SRs and serious ADRs (OR = 0.595, [95% CI 0.362–0.977], p = 0.040). There is also a significant association between complete SRs and e-mail notification (OR = 1.876, [95% CI 1.060–3.321], p = 0.002). The results highlight the small amount of SR documentation sent to pharmacovigilance systems. There is an association between non-serious ADRs and complete SRs. These results reinforce the need for training for notification of ADRs and that these SRs include as much information as possible for an effective drug risk management.

Drug-related hospital admissions in older adults: comparison of the Naranjo algorithm and an adjusted version of the Kramer algorithm

PURPOSE

Drug-related admissions (DRAs) are an important cause of preventable harm in older adults. Multiple algorithms exist to assess causality of adverse drug reactions, including the Naranjo algorithm and an adjusted version of the Kramer algorithm. The performance of these tools in assessing DRA causality has not been robustly shown. This study aimed to evaluate the ability of the adjusted Kramer algorithm to adjudicate DRA causality in geriatric inpatients.

METHODS

DRAs were assessed in a convenience sample of patients admitted to the acute geriatric wards of an academic hospital. DRAs were identified by expert consensus and causality was evaluated using the Naranjo and the adjusted Kramer algorithms. Positive agreement with expert consensus was calculated for both algorithms. A multivariable logistic regression analysis was performed to explore determinants for a DRA.

RESULTS

A total of 218 geriatric inpatients was included of whom 65 (29.8%) experienced a DRA. Positive agreement was 72.3% (95% confidence interval (CI), 59.6-82.3%) and 100% (95% CI, 93.0-100%) for the Naranjo and the adjusted Kramer algorithm, respectively. Diuretics were the main culprits and most DRAs were attributed to a fall (n = 18; 27.7%). A fall-related principal diagnosis was independently associated with a DRA (odds ratio 20.11; 95% CI, 5.60-72.24).

CONCLUSION

The adjusted Kramer algorithm demonstrated a higher positive agreement with expert consensus in assessing DRA causality in geriatric inpatients compared to the Naranjo algorithm. Our results further support implementation of the adjusted Kramer algorithm as part of a standardized DRA assessment in older adults.

Methods for Identifying Culprit Drugs in Cutaneous Drug Eruptions: A Scoping Review

BACKGROUND

Cutaneous drug eruptions are a significant source of morbidity, mortality, and cost to the healthcare system. Identifying the culprit drug is essential; however, despite numerous methods being published, there are no consensus guidelines.

OBJECTIVES

Conduct a scoping review to identify all published methods of culprit drug identification for cutaneous drug eruptions, compare the methods, and generate hypotheses for future causality assessment studies.

ELIGIBILITY CRITERIA

Peer-reviewed publications involving culprit drug identification methods.

SOURCES OF EVIDENCE

Medline, Embase, and Cochrane Central Register of Controlled Trials.

CHARTING METHODS

Registered PRISMA-ScR format protocol on Open Science Forum.

RESULTS

In total, 109 studies and 26 reviews were included comprising 656,635 adverse drug events, most of which were cutaneous. There were 54 methods of culprit drug identification published, categorized as algorithms, probabilistic approaches, and expert judgment. Algorithms had higher sensitivity and positive predictive value, but lower specificity and negative predictive value. Probabilistic approaches had lower sensitivity and positive predictive value, but higher specificity and negative predictive value. Expert judgment was subjective, less reproducible, but the most frequently used to validate other methods. Studies suggest that greater accuracy may be achieved by specifically assessing cutaneous drug eruptions and using combinations of causality assessment categories.

CONCLUSIONS

Culprit drug identification for adverse drug reactions remains a challenge. Many methods have been published, but there are no consensus guidelines. Using causality assessment methods specifically for cutaneous drug eruptions and combining aspects of the different causality assessment categories may improve efficacy. Further studies are needed to validate this hypothesis.

Sports nutrition supplements and adverse events - a meta-epidemiological study of case reports specifically addressing causality assessment

PURPOSE

This meta-epidemiological study aimed to systematically review case reports regarding sports nutrition supplements and adverse events (AEs), specifically addressing the issue of causality assessments.

METHODS

Through a systematic literature search we identified all published case reports of AEs associated with sports nutrition supplements between 1 January 2008 and 1 March 2019. Data regarding AEs, suspected supplements, relevant causality assessment factors and the reporting of clinical reasoning and/or systematic causality assessment methods were extracted.

RESULTS

In all, 72 publications were included, reporting 134 supplements and 37 different AEs in 97 patients (85% males; median age: 30 years [range: 14-60]). Information regarding previous health and regular prescription drugs was not presented in 30% (29/97) and 46% (45/97) of cases, respectively. In 23% (22/97) of the cases, no alternative cause was mentioned. Clinical reasoning was identified in 63% (61/97), and in 13% (8/61) of these, a systematic causality assessment method was applied. In cases with clinical reasoning, a theoretic rationale (92% vs 78%, P = 0.05), a description of previous cases (90% vs 72%, P = 0.021) and body fluid analysis (18% vs 3%, P = 0.027) were reported to a greater extent. Among cases with clinical reasoning, the application of a systematic causality assessment method captured additional important aspects: use of medication (100% vs 55%, P = 0.015), alcohol use (88% vs 43%, P = 0.020) and illicit drug use (88% vs 40%, P = 0.011).

CONCLUSIONS

In published case reports where sports nutrition supplements were suspected to have caused AEs, essential factors for causality assessment were left out in a non-negligible proportion. Clinical reasoning was identified in most cases whereas a systematic causality assessment method was applied in a minority. Factors of importance for causality assessment were reported to a greater extent in cases including clinical reasoning, and the application of a systematic causality assessment method captured additional aspects of importance.

"Artificial Intelligence" for Pharmacovigilance: Ready for Prime Time?

There is great interest in the application of 'artificial intelligence' (AI) to pharmacovigilance (PV). Although US FDA is broadly exploring the use of AI for PV, we focus on the application of AI to the processing and evaluation of Individual Case Safety Reports (ICSRs) submitted to the FDA Adverse Event Reporting System (FAERS). We describe a general framework for considering the readiness of AI for PV, followed by some examples of the application of AI to ICSR processing and evaluation in industry and FDA. We conclude that AI can usefully be applied to some aspects of ICSR processing and evaluation, but the performance of current AI algorithms requires a 'human-in-the-loop' to ensure good quality. We identify outstanding scientific and policy issues to be addressed before the full potential of AI can be exploited for ICSR processing and evaluation, including approaches to quality assurance of 'human-in-the-loop' AI systems, large-scale, publicly available training datasets, a well-defined and computable 'cognitive framework', a formal sociotechnical framework for applying AI to PV, and development of best practices for applying AI to PV. Practical experience with stepwise implementation of AI for ICSR processing and evaluation will likely provide important lessons that will inform the necessary policy and regulatory framework to facilitate widespread adoption and provide a foundation for further development of AI approaches to other aspects of PV.