A new algorithm to identify the causality of adverse drug reactions

A Bayesian generating function approach to adverse drug reaction screening

Determining causality of an adverse drug reaction (ADR) requires a multifactor assessment. The classic Naranjo algorithm is still the dominant assessment tool used to determine causality. But, in spite of its effectiveness, the Naranjo algorithm is manually intensive and impractical for assessing very many ADRs and drug combinations. Thus, over the years, many "automated" algorithms have been developed in an attempt to determine causality. By-and-large, these algorithms are either regression-based or Bayesian. In general, the automatic algorithms have several major drawbacks that preclude fully automated causality assessment. Therefore, signal detection (or causality screening) plays a role in a "first pass" of large ADR databases to limit the number of ADR/drug combinations a skilled human further assesses. In this work a Bayesian signal detector based on analytic combinatorics is developed from a point of view commonly adopted by engineers in the field of radar and sonar signal processing. The algorithm developed herein addresses the commonly encountered issues of misreported data and unreported data. In the framework of signal processing, misreported ADRs are identified as "clutter" (unwanted data) and unreported ADRs are identified as "missed detections". Including the aforementioned parameters provides a more complete probabilistic description of ADR data.

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.

Drug-induced acanthosis nigricans: A systematic review and new classification

Drug-induced acanthosis nigricans is an uncommon subtype of acanthosis nigricans and the data on this topic is not well understood by clinicians as it is presently limited in the literature. Previous reports of drug-induced acanthosis nigricans have simply consisted of a list of drugs possibly implicated in causing acanthosis nigricans. Several drugs listed are based on single case reports without biopsy confirmation, report of clearing on stopping the drug or reporting on whether acanthosis nigricans recurred with drug rechallenge. A comprehensive literature search was conducted using PubMed, EMBASE(Ovid), Cochrane Library, Scopus, and Web of Science electronic databases. The authors screened the initial result of the search strategy by title and abstract using the following inclusion criteria: eligible studies included those with patients who developed acanthosis nigricans secondary to a drug. This study is the first to comprehensively review the drugs that have been implicated in the development of acanthosis nigricans. A total of 38 studies were included in the systematic review, and a total of 13 acanthosis nigricans inducing drugs were identified. Nicotinic acid and insulin were the two most significant drugs that were reported to cause acanthosis nigricans. By using the results of this study, we created a revised classification system of drug-induced acanthosis nigricans which can be used as a concise framework for clinicians to refer to.

Determination of good pharmacovigilance reporting practices in Quebec hospital pharmacies using a modified Delphi method

PURPOSE

Many published guidelines are available for health care providers describing the best way to manage patient's adverse drug reactions (ADRs). However, there is a lack of guidance on the best way to promote and manage ADR reporting within hospitals. The goal of this study was to develop good pharmacovigilance reporting practices (GPRPs).

METHODS

This descriptive study used a modified Delphi method. The research team developed 41 statements, according to a modified Specific Measurable Attainable Realistic Timely (SMART) method and grouped them in six categories: organization (n = 12 statements), pharmacovigilance committee (n = 4), database (n = 5), training (n = 5), tools (n = 3), and quality (n = 12). The Delphi consultation (two online rounds, conducted in 2018) involved directors of pharmacy in Quebec hospitals.

RESULTS

Of 30 directors of pharmacy invited to participate in the first round, 27 (90%) did so. Following this round, the wording of five statements was modified according to pre-established rules. Twenty-five (93%) of the original 27 participants responded during the second round. Of the initial 41 statements, 37 were selected (average score ≥ 7); the other four were eliminated. Of the 37 statements selected, 22 had a "must do" formulation, 12 had a "should do" formulation, and three had a "may do" formulation.

CONCLUSION

Using a modified Delphi method, we established a set of GPRPs for hospital pharmacy based on 37 statements. To our knowledge, these are the first GPRPs published in the hospital pharmacy literature.

Can causality assessment fulfill the new European definition of adverse drug reaction? A review of methods used in spontaneous reporting

Causality assessment is a fundamental biomedical technique for the signal detection performed by Pharmacovigilance centers in a Spontaneous reporting system. Moreover, it is a crucial and important practice for detecting preventable adverse drug reactions. Among different methods for causality assessment, algorithms (such as the Naranjo, or Begaud Methods) seem for their operational procedure and easier applicability one of the most commonly used methods. With the upcoming of the new European Pharmacovigilance legislation including in the definition of the adverse event also effects resulting from abuse, misuse and medication error, all well-known preventable causes of ADRs, there was an emerging need to evaluate whether algorithms could fulfill this new definition. In this review, twenty-two algorithmic methods were identified and none of them seemed to fulfill perfectly the new criteria of adverse event although some of them come close. In fact, several issues were arisen in applying causality assessment algorithms to these new definitions as for example the impossibility to answer the rechallenge question in case of medication error or AEFI (Adverse Event Following Immunization). Moreover, the exact conditions at which events occurred, as for example dosage or mode of administration should be considered to better assess causality in conditions of abuse/overdose, or misuse as well as in conditions of lack of expected efficacy reports for biotechnological drugs and adverse event occurring after mixing of vaccines. Therefore, this review highlights the need of updating algorithmic methods to allow a perfect applicability in all possible clinical scenarios accordingly or not with the terms of marketing authorization.

Proposal for a New Tool to Evaluate Drug Interaction Cases

The assessment of causation for a potential drug interaction requires thoughtful consideration of the properties of both the object and precipitant drugs, patient-specific factors, and the possible contribution of other drugs that the patient may be taking. The Naranjo nomogram was designed to evaluate single-drug adverse events, not drug–drug interactions. Several of the questions on the Naranjo nomogram do not apply to potential drug–drug interactions, while others do not specify object or precipitant drug. Nevertheless, it has been inappropriately used to evaluate drug–drug interactions. The Drug Interaction Probability Scale (DIPS) was developed to provide a guide to evaluating drug interaction causation in a specific patient. It is intended to be used to assist practitioners in the assessment of drug interaction–induced adverse outcomes. The DIPS uses a series of questions relating to the potential drug interaction to estimate a probability score. An accurate assessment using the DIPS requires knowledge of the pharmacologic properties of both the object and precipitant drugs. Inadequate knowledge of either the drugs involved or the basic mechanisms of interaction will be a limitation for some users. The DIPS can also serve as a guide in the preparation of articles describing case reports of drug interactions, as well as in the evaluation of published case reports.

Bridging islands of information to establish an integrated knowledge base of drugs and health outcomes of interest

The entire drug safety enterprise has a need to search, retrieve, evaluate, and synthesize scientific evidence more efficiently. This discovery and synthesis process would be greatly accelerated through access to a common framework that brings all relevant information sources together within a standardized structure. This presents an opportunity to establish an open-source community effort to develop a global knowledge base, one that brings together and standardizes all available information for all drugs and all health outcomes of interest (HOIs) from all electronic sources pertinent to drug safety. To make this vision a reality, we have established a workgroup within the Observational Health Data Sciences and Informatics (OHDSI, http://ohdsi.org) collaborative. The workgroup’s mission is to develop an open-source standardized knowledge base for the effects of medical products and an efficient procedure for maintaining and expanding it. The knowledge base will make it simpler for practitioners to access, retrieve, and synthesize evidence so that they can reach a rigorous and accurate assessment of causal relationships between a given drug and HOI. Development of the knowledge base will proceed with the measureable goal of supporting an efficient and thorough evidence-based assessment of the effects of 1,000 active ingredients across 100 HOIs. This non-trivial task will result in a high-quality and generally applicable drug safety knowledge base. It will also yield a reference standard of drug–HOI pairs that will enable more advanced methodological research that empirically evaluates the performance of drug safety analysis methods.

Dilemmas of the causality assessment tools in the diagnosis of adverse drug reactions

Importance: Basic essence of Pharmacovigilance is prevention of ADRs and its precise diagnosis is crucially a primary step, which still remains a challenge among clinicians. Objective: This study is undertaken with the objective to scrutinize and offer a notion of commonly used as well as recently developed methods of causality assessment tools for the diagnosis of adverse drug reactions and discuss their pros and cons. Evidence review: Overall 49 studies were recognized for all assessment methods with five major decisive factors of causality evaluation, all the information regarding reasons allocating causality, the advantages and limitations of the appraisal methods were extracted and scrutinized. Findings: From epidemiological information a past prospect is designed and subsequent possibility merged this background information with a clue in the individual case to crop up with an approximation of causation. Expert judgment is typically based on the decisive factor on which algorithms are based, nevertheless in imprecise manner. The probabilistic methods use the similar principle; however connect probabilities to each measure. Such approaches are quite skeptical and liable to generate cloudy causation results. Causation is quite intricate to ascertain than correlation in Pharmacovigilance due to numerous inherent shortcomings in causality assessment tools. Conclusions and relevance: We suggest that there is a need to develop a high quality assessment tool which can meticulously establish suitable diagnostic criteria for ADRs with universal acceptance to improvise the fundamental aspect of drug safety and evade the impending ADRs with the motive to convert Pharmacovigilance into a state of art.

ADRIC: Adverse Drug Reactions In Children – a programme of research using mixed methods

Aims

To comprehensively investigate the incidence, nature and risk factors of adverse drug reactions (ADRs) in a hospital-based population of children, with rigorous assessment of causality, severity and avoidability, and to assess the consequent impact on children and families. We aimed to improve the assessment of ADRs by development of new tools to assess causality and avoidability, and to minimise the impact on families by developing better strategies for communication.

Review methods

Two prospective observational studies, each over 1 year, were conducted to assess ADRs in children associated with admission to hospital, and those occurring in children who were in hospital for longer than 48 hours. We conducted a comprehensive systematic review of ADRs in children. We used the findings from these studies to develop and validate tools to assess causality and avoidability of ADRs, and conducted interviews with parents and children who had experienced ADRs, using these findings to develop a leaflet for parents to inform a communication strategy about ADRs.

Results

The estimated incidence of ADRs detected in children on admission to hospital was 2.9% [95% confidence interval (CI) 2.5% to 3.3%]. Of the reactions, 22.1% (95% CI 17% to 28%) were either definitely or possibly avoidable. Prescriptions originating in the community accounted for 44 out of 249 (17.7%) of ADRs, the remainder originating from hospital. A total of 120 out of 249 (48.2%) reactions resulted from treatment for malignancies. Off-label and/or unlicensed (OLUL) medicines were more likely to be implicated in an ADR than authorised medicines [relative risk (RR) 1.67, 95% CI 1.38 to 2.02;p < 0.001]. When medicines used for the treatment of oncology patients were excluded, OLUL medicines were not more likely to be implicated in an ADR than authorised medicines (RR 1.03, 95% CI 0.72 to 1.48;p = 0.830). For children who had been in hospital for > 48 hours, the overall incidence of definite and probable ADRs based on all admissions was 15.9% (95% CI 15.0 to 16.8). Opiate analgesic drugs and drugs used in general anaesthesia (GA) accounted for > 50% of all drugs implicated in ADRs. The odds ratio of an OLUL drug being implicated in an ADR compared with an authorised drug was 2.25 (95% CI 1.95 to 2.59;p < 0.001). Risk factors identified were exposure to a GA, age, oncology treatment and number of medicines. The systematic review estimated that the incidence rates for ADRs causing hospital admission ranged from 0.4% to 10.3% of all children [pooled estimate of 2.9% (95% CI 2.6% to 3.1%)] and from 0.6% to 16.8% of all children exposed to a drug during hospital stay. New tools to assess causality and avoidability of ADRs have been developed and validated. Many parents described being dissatisfied with clinician communication about ADRs, whereas parents of children with cancer emphasised confidence in clinician management of ADRs and the way clinicians communicated about medicines. The accounts of children and young people largely reflected parents’ accounts. Clinicians described using all of the features of communication that parents wanted to see, but made active decisions about when and what to communicate to families about suspected ADRs, which meant that communication may not always match families’ needs and expectations. We developed a leaflet to assist clinicians in communicating ADRs to parents.

Conclusion

The Adverse Drug Reactions In Children (ADRIC) programme has provided the most comprehensive assessment, to date, of the size and nature of ADRs in children presenting to, and cared for in, hospital, and the outputs that have resulted will improve the management and understanding of ADRs in children and adults within the NHS. Recommendations for future research: assess the values that parents and children place on the use of different medicines and the risks that they will find acceptable within these contexts; focusing on high-risk drugs identified in ADRIC, determine the optimum drug dose for children through the development of a gold standard practice for the extrapolation of adult drug doses, alongside targeted pharmacokinetic/pharmacodynamic studies; assess the research and clinical applications of the Liverpool Causality Assessment Tool and the Liverpool Avoidability Assessment Tool; evaluate, in more detail, morbidities associated with anaesthesia and surgery in children, including follow-up in the community and in the home setting and an assessment of the most appropriate treatment regimens to prevent pain, vomiting and other postoperative complications; further evaluate strategies for communication with families, children and young people about ADRs; and quantify ADRs in other settings, for example critical care and neonatology.

Funding

The National Institute for Health Research Programme Grants for Applied Research programme.

An updated method improved the assessment of adverse drug reaction in routine pharmacovigilance

OBJECTIVE

Updating a logistic causality assessment method to improve its agreement with consensual expert judgment (CEJ).

STUDY DESIGN AND SETTING

A random sample of 53 drug-event pairs from a pharmacovigilance database were evaluated independently by CEJ and by a group of experts in pharmacovigilance using the logistic method. Causes of disagreement between both approaches were analyzed, and changes in the assessment of some criteria of the logistic method were proposed and tested in models. The model giving results closest to the CEJ was retained and compared with the initial version on another set of drug-event pairs.

RESULTS

Finally, only the criterion "Search for nondrug cause" was changed into "Search for other causes." The assessment not investigated, possible other cause decreased the probability of drug causation instead of being neutral, whereas the assessment not applicable, not required remained neutral. This new version presents much improved specificity (0.56 vs. 0.33), relatively good sensitivity (0.96), and positive and negative predictive values (0.92 and 0.71).

CONCLUSION

The updated logistic method presented here improves the initial version that had poor specificity and tended to overestimate drug causation. This new version presents satisfactory characteristics to be used in routine pharmacovigilance.

Development and inter-rater reliability of the Liverpool adverse drug reaction causality assessment tool

Aim

To develop and test a new adverse drug reaction (ADR) causality assessment tool (CAT).

Methods

A comparison between seven assessors of a new CAT, formulated by an expert focus group, compared with the Naranjo CAT in 80 cases from a prospective observational study and 37 published ADR case reports (819 causality assessments in total).

Main Outcome Measures

Utilisation of causality categories, measure of disagreements, inter-rater reliability (IRR).

Results

The Liverpool ADR CAT, using 40 cases from an observational study, showed causality categories of 1 unlikely, 62 possible, 92 probable and 125 definite (1, 62, 92, 125) and ‘moderate’ IRR (kappa 0.48), compared to Naranjo (0, 100, 172, 8) with ‘moderate’ IRR (kappa 0.45). In a further 40 cases, the Liverpool tool (0, 66, 81, 133) showed ‘good’ IRR (kappa 0.6) while Naranjo (1, 90, 185, 4) remained ‘moderate’.

Conclusion

The Liverpool tool assigns the full range of causality categories and shows good IRR. Further assessment by different investigators in different settings is needed to fully assess the utility of this tool.

Development of a combined system for identification and classification of adverse drug reactions: Alerts Based on ADR Causality and Severity (ABACUS)

Currently, adverse drug reaction (ADR) causality and severity are assessed using different systems but there is no standard method to combine the results. In this work, a combined ADR causality and severity assessment system, including an online version, was developed. Logical rules were defined to translate the score obtained from the system into three alert zones: green, amber, and red. The alert zones are useful for triaging ADR cases as they help define the seriousness of the ADR and the urgency of the responses required. This new scoring system may be useful for clinicians, investigators, and regulators seeking information on the likelihood of a drug causing an adverse reaction, and whether an adverse reaction is sufficiently dangerous for the drug to be withheld or undergo further investigation.

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

Numerous methods for causality assessment of adverse drug reactions (ADRs) have been published. The aim of this review is to provide an overview of these methods and discuss their strengths and weaknesses. We conducted electronic searches in MEDLINE (via PubMed), EMBASE and the Cochrane databases to find all assessment methods. Thirty-four different methods were found, falling into three broad categories: expert judgement/global introspection, algorithms and probabilistic methods (Bayesian approaches). Expert judgements are individual assessments based on previous knowledge and experience in the field using no standardized tool to arrive at conclusions regarding causality. Algorithms are sets of specific questions with associated scores for calculating the likelihood of a cause-effect relationship. Bayesian approaches use specific findings in a case to transform the prior estimate of probability into a posterior estimate of probability of drug causation. The prior probability is calculated from epidemiological information and the posterior probability combines this background information with the evidence in the individual case to come up with an estimate of causation. As a result of problems of reproducibility and validity, no single method is universally accepted. Different causality categories are adopted in each method, and the categories are assessed using different criteria. Because assessment methods are also not entirely devoid of individual judgements, inter-rater reliability can be low. In conclusion, there is still no method universally accepted for causality assessment of ADRs.

Gold standards in pharmacovigilance: the use of definitive anecdotal reports of adverse drug reactions as pure gold and high-grade ore

Anecdotal reports of adverse drug reactions are generally regarded as being of poor evidential quality. This is especially relevant for postmarketing drug safety surveillance, which relies heavily on spontaneous anecdotal reports. The numerous limitations of spontaneous reports cannot be overemphasised, but there is another side to the story: these datasets also contain anecdotal reports that can be considered to describe definitive adverse reactions, without the need for further formal verification. We have previously defined four categories of such adverse reactions: (i) extracellular or intracellular tissue deposition of the drug or a metabolite; (ii) a specific anatomical location or pattern of injury; (iii) physiological dysfunction or direct tissue damage demonstrable by physicochemical testing; and (iv) infection, as a result of the administration of an infective agent as the therapeutic substance or because of demonstrable contamination. In this article, we discuss the implications of these definitive ('between-the-eyes') adverse effects for pharmacovigilance.

A quantitative approach of using genetic algorithm in designing a probability scoring system of an adverse drug reaction assessment system

BACKGROUND

The detrimental effects of adverse drug reactions (ADRs) are well established. Hence, precise and accurate assessment of ADRs' causality which can differentiate signal from noise is crucial in screening, management and minimisation of ADRs.

OBJECTIVE

The current study reported our attempt to improve the scoring system of a previously published algorithm of ADR assessment by our group using a genetic algorithm approach so that the final score can measure the probability of ADR causality.

DESIGN

Using ADR cases obtained from the Centre for Drug Administration, the national centre for pharmacovigilance in Singapore, with known causality probability values as reference points, rules were developed to define possible combinations of criteria for 'Definite' ADR cases and 'Probable' ADR cases. A new scoring system was developed using these parameters with the help of genetic algorithm, and tested on 37 'Definite' and 431 'Not Definite' ADR cases. In addition, sensitivity and specificity analysis were performed to allow a comparison of performance between our algorithm and that used by the Adverse Drug Reaction Advisory Committee in Australia (ADRAC).

RESULTS

The new scoring system is able to provide a probability of the causality of an ADR by a suspected drug. When applied to the 'Definite' and 'Not Definite' ADR reports, the new algorithm gave a sensitivity of 83.8% and specificity of 71.0%.

CONCLUSIONS

Using a quantitative method of assessing causality in the new algorithm allows rare and new ADRs to be more readily identified since a quantitative score can give a more precise degree of ADR causality. This scoring system that provides a probability score would help to make this algorithm more informative and assistive for clinicians, regulatory agencies or pharmaceutical companies to generate ADR alerts. The higher sensitivity value displayed by our algorithm also shows that it would be a good ADR screening tool.