Causality Assessment of Adverse Drug Reactions ADR

CAUSALITY ASSESSMENT OF ADRs Causality assessment is a process of establishing the association between the suspected drug and reaction. Causality assessment of ADRs can be performed by clinicians, academics, the pharmaceutical industry and regulators, and in different settings, including clinical trials etc. It is often difficult to decide an adverse clinical event is an ARD or not, when the underlying disease is capable of producing the same clinical manifestation as caused by the drug. Furthermore even if it is identified as a ADR, establishing the relationship between causative drug is a complex and difficult drug. As the patients will be on multiple drugs for multiple disease condition and there are n number of drug available in a market. Why causality assessment is needed? Once a ADR is identified or suspected, the assessment will start from collecting all the relevant information (patient demographic, medication and medical history, detailed description of the reaction like time of onset, duration of the reaction, complication, any management or care provided for the reaction and the outcome of the care and relevant investigation reports) and using the information for assessment. Adverse reactions are rarely particular for the drug, diagnostic tests are usually absent and a rechallenge is rarely ethically justified. In causality assessment only few adverse reactions are ‘definite’ or ‘unlikely’; most of the adverse reactions are somewhere in between the extremes like, i.e. ‘possible’ or ‘probable’. The assessment is highly subjective, (quite often objective) and based on individual clinician assessment. Therefore it may be possible to on clinical and unlikely to another clinician. The causality assessment is needed for the following a. Causality assessment is having a greater importance in both regulatory environment and the pharmaceutical industries. b. As it directly influence the decision-making by a clinician either to continue or to withdraw the likely drugs (in establishing individual the dosage regimen) for the patient. c. To reduce and prevent the incidence of ADRs, which will improve the patients quality of life. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 1 Advances and limitations of standardised case causality assessment Causality assessment can do Causality assessment cannot do  Reduce the disagreement between the  Can’t give accurate quantitative different assessors measurement of relationship likelihood  Classify relationship likelihood  Can’t Distinguish valid from invalid cases  Mark individual case reports  Can’t prove the connection between drug and event  Improvement of scientific evaluation;  Can’t quantify the contribution of a drug to educational the development of an adverse event  Can’t Change uncertainty into certainty Factors or fundamentals need to be considering in causality assessment:While assessing the causality, all the methods or tools, should follow the basic fundamental principles of identification an ADR: (i) Did the event started before taking drug by the patient? This may seem an obvious consideration, but reports are received in which this has not been taken into account, and a careful check has then revealed that the adverse event occurred before using the suspect medicine and so there were no relationship. (ii) Are there any other possible cause for the event?  Could the adverse event is due to the disease being treated?  Could this adverse event is due to some other co-existent disease?  Could this adverse event is due to some other medicine being used simultaneously? (iii) Is the duration to onset of the event plausible (reasonable)? Is the adverse event likely to have occurred in the time frame in question?  Did the event occur, immediately/quickly to be related to the particular medicine, pharmacological action into consideration?  Did the patient use the medicine for a long time without any problems? (Some reactions lately (after long-term exposure to the drug) and most reactions will occur soon after the patient starts taking the medicine.)  The nature of the adverse event should be taken into consideration, while assessing the significance of the period of exposure, for example: o some events take a long time to develop (e.g. cancer); o some develop quickly (e.g. nausea and headache); advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 2 o allergic reactions to first-time exposure to a drug generally take around 10 days to appear. On repeat exposure they may occur immediately. (iv) Did the event occur after using some other medicine? If the event occurs in a short time period after commencing another medicine, then two possibilities should be considered.  The new medicine might have caused the event.  There might be an interaction between the two drugs and the interaction caused the event. (v) Did the event occur after the onset of some new illness?  If so, the event may be due to the new condition/disease. (vi) What is the response to withdrawal of the medicine (Dechallenge)?  Did the patient recovered?  Did the patient condition improved?  There was no change?  Did the patient condition get worse?  Is the response/answer to dechallenge unknown? If this is the case, then it should always be recorded as unknown. If more than one medicine has been withdrawn, and if rechallenge is considered to be appropriate in this case, then it should be performed with only one medicine each at a time. (vii) What is the response to rechallenge? Conditions for a positive rechallenge are:  The patient recovered on initial withdrawal of medicine;  The patient developed the same event again when re-exposed with the same medicine alone, although it may be of different severity;  The patient recovered again, when the medicine was withdrawn once again;  It should be renowned that it is not always safe to subject the patient to a rechallenge;  If the response/answer to rechallenge is unknown, this should be recorded. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 3 Methods of Causality Assessment There are various methods/approaches/algorithms/tools proposed for the ADR Causality Assessment. All these are approaches or methods grouped into 3 categories:I. II. III. Global introspection method or experts judgement Algorithm’s or standard assessment tools Probabilistic methods or Bayesian approach I. Global Introspection method (or) Expert s Judgement The most common and unstructured method of causality assessment is global introspection method, which is based on the knowledge of a group of clinicians or experts. The experts consider all the clinical information available in relation to suspected drug and ADRs, and judgement is made about the likelihood of the relation between the reaction and suspected drug. Knowledge and experience of the clinician plays a very important role in performing global introspection, but variations in clinical judgement can be happen between the experts because of differences in knowledge and experience of clinicians and any information is not available for assessment. Some of tools developed in this way are, World Health Organisation - Uppsala Monitoring Center scale (WHO–UMC), Swedish regulatory agency method and Visual analogue scale. The basic criteria’s for this kind of causality assessment are knowledge on pharmacology, literature information of ADRs, association of adverse event, biological plausibility and other exclusions. The WHO –UMC causality assessment system The World Health Organisation – Uppsala Monitoring Center in association with Centres participating in the Programme for International Drug Monitoring developed WHO-UMC causality assessment tool and is intended as a practical tool for the assessment of case reports (adverse reactions). It is basically a collective assessment by taking into considerations, both the clinical-pharmacological aspects of the case history and the quality of the documentation of the observation. Pharmacovigilance is particularly concerned with the detection of unknown and unexpected adverse reactions; epidemiological knowledge (data) and statistical opportunity play a less important role in the system. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 4 The WHO -UMC causality assessment scale is most widely acceptable tool, because of their validity/ reliability. The relationship between suspected drug and event grouped into 6 categories like, certain (or definite); probable; possible; unlikely; unclassified (or conditional); unassessable (or unclassifiable). WHO-UMC Causality assessment tool Causality Categories Assessment criteria • Event or laboratory test abnormality, with plausible time relationship to drug intake • Cannot be explained by disease or other drugs • Response to withdrawal plausible (pharmacologically, pathologically) Certain • Event definitive pharmacologically or phenomenologically (i.e. an objective and specific medical disorder or a recognised pharmacological phenomenon) • Rechallenge satisfactory, if necessary • Event or laboratory test abnormality, with reasonable time relationship to drug intake Probable / • Unlikely to be attributed to disease or other drugs Likely • Response to withdrawal clinically reasonable • Rechallenge not required • Event or laboratory test abnormality, with reasonable time relationship to drug intake Possible • Could also be explained by disease or other drugs • Information on drug withdrawal may be lacking or unclear • Event or laboratory test abnormality, with a time to drug intake that makes a relationship improbable (but not impossible) Unlikely • Disease or other drugs provide plausible explanations • Event or laboratory test abnormality Conditional / • More data for proper assessment needed, or Unclassified • Additional data under examination Unassessable • Report suggesting an adverse reaction • Cannot be judged because information is insufficient or contradictory / Unclassifiable • Data cannot be supplemented or verified advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 5 The use of the WHO-UMC system To illustrate how the system works, we suggest to first make a comparison of the criteria and wording of ‘Probable’ and Certain’. First of all there is one more principle in the category ‘Certain’, the fourth: ‘Event definitive pharmacologically or phenomenologically’, i.e. an objective and specific medical disorder or a recognised pharmacological phenomenon (for instance ‘grey baby syndrome’ and Chloramphenicol, or anaphylaxis immediately after the administration of a drug that had been given previously). This means that any other event is automatically excluded and can never qualify for ‘Certain’ (even in the case of a positive rechallenge observation). For ‘Certain’, rechallenge information with a satisfactory outcome is requested (i.e. what has happened when the drug was first stopped and later on resumed), unless the evidence in the report is already convincing without a re-exposure. For ‘Probable’, on the other hand, a rechallenge is not required. To qualify as ‘Certain’ the interval between the start of the drug and the onset of the event must be ‘plausible’; this means that there is in sufficient detail a positive argument in support of the view that the drug is causally involved, pharmacologically or pathologically. For ‘Probable’ the time relationship should be ‘reasonable’; this is a more neutral term covering everything that is not unreasonable. Also, with regard to the second criterion, ‘alternative causes’, the wording is different in ‘Probable’. For ‘Certain’ the occurrence of the event cannot be explained by any disease the patient is known to have or any other drug taken. For ‘Probable’, on the other hand, the event is ‘unlikely’ to be attributable to another cause. Also the dechallenge situations (i.e. what happened after stopping) are different. In a ‘Certain’ case report, the course of events constitutes a positive argument in favour of holding the suspected drug responsible, in pharmacological or pathological respects, whereas in a ‘Probable’ case it is sufficient if it is ‘clinically reasonable’ (i.e. not unreasonable). The essential distinctions between ‘Probable’ and ‘Possible’ are that in the latter case there may be another equally likely explanation for the event and/or there is no information or uncertainty with regard to what has happened after stopping. The criteria that may render the connection ‘Unlikely’ are firstly the time relationship is improbable (with the knowledge at the time), and/or another explanation is more likely. The term ‘Unclassified / Conditional’ is of a preliminary nature and is appropriate when, for a proper assessment, there is more data needed and such data are being sought, or are already advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 6 under examination. Finally when the information in a report is incomplete or contradictory and cannot be complemented or verified, the verdict is ‘Unclassifiable’. Since by far the most frequent categories in case reports are ‘Possible’ and ‘Probable’, the usual approach to using the system is to choose one of these categories (depending on the impression of the assessor) and to test if the various criteria fit with the content of the case report. If the report seems stronger one can go one step ‘higher’ (e.g. from ‘Possible’ to ‘Probable’), if the evidence seems weaker one should try a ‘lower’ category. To see if that category is the right one or if it does again not seem to fit, the next adjacent term is tried. For drug-drug interactions the WHO-UMC system can be used by assessing the actor drug, which influences the kinetics or dynamics of the other drug (which has usually been taken over a longer period), in the medical context of the patient. Requirements for inclusion of an event in a specific category Certain a. The event is a specific clinical or laboratory phenomenon. b. The time elapsed between the administration of the drug and the occurrence of the event is plausible. (Requirement: dates of drug administration and date of onset of the event must be known.) c. The event cannot be explained by concomitant disease or any other drug or chemical. (Requirement: Details of other medicines taken must be known. The report must also state if there were no other medicines in use. If this is unknown, then doubt exists and the event cannot be included in this category.) d. The patient recovered within a plausible length of time following withdrawal of the drug. (Requirement: The date of withdrawal of the drug and the time taken for recovery should be known. If these dates are unknown, then doubt exists and the event cannot be included in this category.) e. The same event recurred following rechallenge with the same drug alone. (Requirement: The report must state the outcome of rechallenge. If this is unknown, then doubt exists and the event cannot be included in this category.) If rechallange is not done or result of it is un-known, one cannot categorise as Certain advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 7 Probable a. The event is a specific clinical or laboratory phenomenon. b. The time elapsed between the administration of the drug and the occurrence of the event is plausible. (The dates of drug administration and date of onset of the event must be known.) c. The event cannot be explained by concurrent disease or any other drug or chemical. (Details of other medicines taken must be known. The report must also state if there were no other medicines in use. If this is unknown, then doubt exists and the event cannot be included in this category.) d. The patient recovered within a plausible length of time following withdrawal of the drug. (The date of drug withdrawal and the time taken for recovery should be known.) e. Rechallenge did not occur, or the result is unknown. If dechallange is not done or result of it is un-known or other possible causes are available, one cannot categorise as probable Possible a. The time elapsed between the administration of the drug and the occurrence of the event is plausible. (The dates of drug administration and date of onset of the event must be known.) b. The outcome of withdrawal of the suspect medicine is not known, and/or the medicine might have been continued and the final outcome is not known; and/or c. there might be no information on withdrawal of the medicine; and/ or d. the event could be explained by concomitant disease or use of other drugs or chemicals; and/or e. there might be no information on the presence or absence of other medicines. f. Deaths cannot be coded as probable because there is no opportunity to see the effect of withdrawal of the drug. If there is a plausible time relationship, a death should be coded as possible. g. In addition to deaths, there is a further group of events that do not fit the relationship assessment process and the coding can vary. Consider the following examples:o Myocardial infarction. Many patients recover from this event as part of the natural history of the disease and, with very few exceptions, recovery is not a response to withdrawal of a drug. Hence the result of “dechallenge” is meaningless. This type of reaction may be coded as “possible”. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 8 o Stroke. Some patients recover fully, some partially, some remain severely disabled and some die. All these outcomes are part of the natural history of the disease and, with very few exceptions, are unrelated to drug withdrawal. Again, the result of “dechallenge” is usually meaningless. This type of reaction may be coded as “possible”. o Acute anaphylaxis immediately following an injection. Here there is an obvious direct relationship, but the usual parameters for establishing relationship, e.g. dechallenge do not apply. In this example, the best category for the relationship is “certain”. Unlikely a. The event occurred with a duration to onset that makes a causal effect improbable with the drug being considered. (The pharmacology of the drug and nature of the event should be considered in arriving at this conclusion.); and/or b. the event commenced before the first administration of the drug; and/ or c. the drug was withdrawn and this made no difference to the event when, clinically, recovery would be expected. (This would not apply for some serious events such as myocardial infarction, or events causing permanent damage.); and/or d. it is strongly suggestive of a non-causal relationship if the drug was continued and the event resolved. Unclassified or conditional a. These are reports with insufficient data to establish a relationship and more data are expected. This is a temporary repository, and the category for these events will be finalized when the new data become available. Unassessable a. An event has occurred in association with a drug, but there are insufficient data to make an assessment. b. Some of the data may be contradictory or inconsistent. c. Details of the report cannot be supplemented or verified. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 9 II. Algorithm’s or standard assessment tools Algorithms are simple and precisely framed set of questions to gather information adequately to perform the causality assessment i.e., the relationship between the suspected drug and reaction. These algorithms have scores, based on the answers to the questions and we will categorise the reaction into definite, probable, possible, or unlikely an ADR. A number of algorithms have developed to reduce the intra-inter-rater variability (assessor’s bias) to make this Causality Assessment Tool (CAT) more reproducible, valid and to standardise the assessment tools. But, no single algorithms have been accepted as a gold standard because of the discrepancy exists between the published and developed assessment tools. Some of the available algorithms are Naranjo’s algorithm, Liverpool ADR causality assessment tool, kramer’s algorithm, Ciba geigy method , Balanced assessment method, Summary time plot , Jones algorithm, Maria and Victorino (M and V) scale ,Roussel Uclaf causality assessment method (RUCAM), Drug Interaction Probability Scale (DIPS) ,Dangaumou’s french method , Begaud algorithm, the Yale algorithm, Karch and Lasagne, Adverse Drug Reactions Advisory Committee guidelines (ADRAC) and French imputation method. Among all Naranjo’s algorithms is one of the most commonly used assessment tools. Algorithms may be less open to the effect of confounding variables (co morbidities or concomitant drugs), but clinical judgment will varies. Naranjo’s ADR probability Scale The Naranjo Algorithm is a questionnaire (set of 10 questions) designed by Naranjo C A et al (1981) for determining the likelihood of whether an ADR is actually due to the suspected drug or other factor. The probabilities are assigned via a score termed definite, probable, possible or doubtful. This probability scale uses questionnaire and points are added and taken away based on the response to each questions. Naranjo’s scale may not be best suitable for assessment in children. After completing the questionnaire, based on the obtained score, the ADR can be categorise into definite, probable, possible and doubtful. If the total score is 9 or more than 9 then ADR is consider as “Definite’, if the total score is in between 5 to 8 then ADR is consider as “Probable’, if the total score is in between 1 to 4 then ADR is consider as “possible’ and if the total score is 0 or less than 0 then ADR is consider as “doubtful’. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 10 Naranjo’s ADR probability Scale To assess the ADR, please answer the following questionnaire and give the pertinent score. Q.No Questions Yes No Don’t Score know 1 Are there previous conclusive reports of this reaction? +1 0 0 Did the adverse event appear after the suspected drug 2 +2 -1 0 was administered? Did the adverse reaction improve when the drug was 3 +1 0 0 discontinued or a specific antagonist was given? Did the adverse reaction appear when the drug was Re4 +2 -1 0 administered? Are there alternative causes (other than the drug) that 5 -1 2 0 Could on their own have caused the reaction? 6 Did the reaction reappear when a placebo was given? -1 +1 0 Was the drug detected in the blood (or other fluids) in 7 +1 0 0 concentrations known to be toxic? Was the reaction more severe when the dose was 8 +1 0 0 increased, or less severe when the dose was decreased? Did the patient have a similar reaction to the same or 9 +1 0 0 similar drugs in any previous exposure? Was the adverse event confirmed by any objective 10 +1 0 0 evidence? Total Score : Naranjo’s Score-Interpretation of scores Total Score ≥9 Definite 5 to 8 probable 1 to 4 possible ≤0 doubtful Interpretation of Scores The reaction (a) followed a reasonable temporal (time) sequence after a drug or toxic drug levels has identified in body fluids / tissues, (b) followed a recognized response to suspected drug, (c) confirmed by dechallange (improve patients symptoms on drug removal i.e., reaction reduced) and rechallange (reappearance of reaction again after administrating the suspected drug ). The reaction (a) followed a reasonable temporal (time) sequence after a drug, (b) followed a recognized response to suspected drug, (c) confirmed by dechallange and rechallange not done or don’t know the outcomes, (d) cannot reasonably explained by the patients clinical condition. The reaction (a) followed a reasonable temporal (time) sequence after a drug, (b) possible followed a recognized pattern to suspected drug, and (c) can be explained by the patients clinical condition. The reaction is likely related to factors other than a suspected drug. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 11 Liverpool ADR Causality Assessment Tool (LCAT) Liverpool ADR Causality Assessment Tool was established for use in paediatrics mostly and generalised for use in other settings also. Ruairi M Gallagher et al, (2013) clinical researcher fellows from university of Liverpool developed this tool, by considering the adaptive Hallas scale as a base. Note: i) Unassessable here means a condition where the drug is administered at one time only (e.g. vaccine), the patient received intermittent therapy (e.g. chemotherapy), or he/she on medication which cannot be stopped (e.g. immunosuppressant drugs), ii) Examples of positive objective evidence: positive lab investigations of the causal ADR mechanism (not those just confirm the adverse reaction), supratherapeutic drug levels and highquality indication of a dose-dependent association with toxicity in the patient. iii) An independent panel with extensive expertise in Pharmacovigilance and statistics (the Adverse Drug Reactions in Children (ADRIC) Steering Group) was asked to review the tool upon completion of the internal evaluation. For Online Liverpool ADR Causality http://www.pmidcalc.org/index.php# Assessment Tool use following advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs link- 12 III. Probabilistic methods or Bayesian approach Bayesian methods use the epidemiological data (literature) of the drug and the present clinical case information for the causality assessment of the suspected drug and the adverse reaction. This Bayesian method has been developed to overcome the limitation of the other methods or tools of causality assessment. This Bayesian method can be assessable both through offline (Bayesian Adverse Reaction Diagnostic Instrument - BARDI) and online (MacBARDI). Bayesian method has been applied to various adverse reactions and has been cross-validated with objective tests. The merits of Bayesian method are:a. Flexibility- can add all clinical relevant information to see all possible cause of the reaction; b. Repeatability - same results will be obtained when you provide same information; c. Transparency- clear representation of how the calculations are made; d. Explicitness- assumptions and information used for the calculations are made clear; e. Completeness-any clinical relevant information of differential diagnostic incorporated; f. Etiologic Balancing- clinical data is assessed for how compatible the findings are with drug versus nondrug causation; g. No Arbitrary Constraints-the effect of information on the assessment is not limited a priori. Limitations are the time constrains and the complexity of calculations may limit the use. Attempts are made to overcome limitations by developing the computerized instruments. But, no simple Bayesian-based causality assessment method is widely available. Also, model development for BARD1 is time-consuming and requires multidisciplinary input. Bayesian Adverse Reaction Diagnostic Instrument (BARDI) BARDI have been proposed to provide a formal inferential framework for causality assessment through a complex statistical approach. This method is unambiguous, in using information and explaining how each piece of data is weighed for Bayesian statistics to combine factors coherently. BARDI was developed to overcome the numerous limitations associated with expert judgements and algorithm tools of casualty assessment. This method is for judgmental analysis that uses conditional odds to assess the probability that a drug was responsible for causing the adverse drug event. This Bayesian probability approach (Posterior advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 13 odds-PsO) is calculated on the basis of information from two sources: epidemiologic (literature) and clinical trial data, which are considered in the prior odds (PrO) ; and a complete clinical data, in which each risk factor is measured individually in the likelihood ratios. Bayesian Adverse Reaction Diagnostic Instrument is used to calculate the odds in favour of a particular drug compared with an alternative cause of adverse reactions. These odds are referred as posterior odds (PsO). The prior odds (PrO) factor is the ratio of the expected drugattributable risk and the background information of population sharing similar characteristics with the patient being considered (such as medical condition). BARDI uses the rules of prescriptive probability theory to break down the posterior odds into six subsets: one deal with the epidemiologic or clinical trials information (the prior odds) and the other five deal with individual patient related case specific information (the likelihood ratios). The five likelihood ratios (LRs) deal with any information of differential diagnostic value under the categories of patient history (Hi); timing of the adverse event with respect to drug administration (Ti); characteristics of the adverse event (Ch); drug dechallenge (De), which refers to any signs, symptoms, or occurrences after drug withdrawal; and drug rechallenge (Re) or re-administration of the suspected causal drug(s). Detailed information of this is odds given below: a. History (Hi) it includes all those conditions of the case that occur before the drug administration in question that may have a bearing on the differential diagnosis. For example, if a drug-induced event is known to be three times more common in females (and the nondrug-induced event was equal by age), the likelihood ratio for history would be equal to three if the patient is female. b. Timing (Ti) it explains the temporal relationship between the beginning of drug administration and the onset of the adverse event. The actual time taken to show the reactions in this patient will be correlated with the expected time of onset for the drug in question and alternative causes. Expected times of onset of reactions are collected from epidemiological data and plotted in histograms so the probability of onset at a certain time can be calculated. c. Characteristics (Ch) it include any diagnostic tests or distinct clinical features of the adverse event that can be used to change the probability in favour of or against one of the causes. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 14 d. Dechallenge (De) it includes the series of changes (symptoms and signs) seen in patient condition after the suspected drug is discontinued. For example, if the patients symptoms (reductions in adverse event) improved after drug discontinuation , it can be of diagnostic significance. e. Rechallenge (Re) If the adverse event reappeared after the drug readministration of the same suspected, this information may have differential diagnostic value These six components are multiplied to obtain the posterior odds (PsO): PsO = PrO x LR(Hi) x LR(Ti) x LR(Ch) x LR(De) x LR(Re) The PsO is frequently expressed as a posterior probability (PsP) in favor of the drug in question: PSP = PsO/(PsO + 1) The Bayesian probabilistic approach can be applied as a spreadsheet programme on both paper and computer. It calculates and provides the instant results in both numerical value and graphical representation as soon as the suspected drug and adverse reaction is evaluated. Case reports will be studied, relevant data that fits to assess the prior probability and distinguished potential causes are also considered, noted. These are now converted to easy computer-aided programs such as Mac-BARDI Q and A. MacBARDI Spreadsheet MacBARDI-Q&A is based on an earlier computerized spreadsheet. This spreadsheet (MacBARDI) used for cases of pulmonary fibrosis associated with antiarrhythmics, cutaneous reactions associated with sulphonamides and anticonvulsants, fetal alcohol syndrome and benzodiazepine withdrawal. The MacBARDI computerized spreadsheet requires five types of information for estimation: (i) pure information lines that describe the input needed; (ii) input lines that are the parameters used to calculate each of the six factors; (iii) assumption lines, which are built-in inputs used in the calculations; (iv) calculation lines that calculate and show the value of each term in the assessment; (v) output lines, which show the value of each factor necessary to calculate the posterior odds and the posterior odds itself. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 15 MacBARDI facilitates updating new information for case analyses and has all the criteria necessary for a good causality assessment method. One should have the knowledge of BARDI for using the Using the spreadsheet, however, and it did not have automatic database lookup, the assessor should find relevant information on the database and enter it on the spreadsheet. MacBARDI Q&A is a simple and easy to use, contains question and answer format. These programmes provides access to specific data and list of assumptions which allows the user to judge the medical and pharmacological information based on which the calculations are made. This programme has 3 components: the interface (Q&A); the spreadsheet (MacBARDI) and a database of pertinent information (Q&A knowledge base). The Q&A knowledge base will collect all the epidemiological, medical and pharmacological information which is encoded in the database. This provide the information to the MacBARDI Spreadsheet on drug and nondrug attributable risk of the reaction, the magnitude of risk factor in patient history, expected time of onset of onset of overall reactions (drug and nondrug induced), any change in likelihood of dechallange and rechallange. All this data, how the assessment is done will be displayed on screen to learn and crosscheck. If the patient has taken some concomitant drugs that is not there in the knowledge base, than the particular information from literature has to been gained and added to the knowledge base for proper assessment. SPREADSHEET MacBARDI Q&A Accesses database Perform calculation Summarise analysis KNOWLEDGE BASE Based on cases published & experts opinion Provide information to spreadsheet The three components of MacBARDI-Q&A - developed for sulphonamides and aromatic anticonvulsants (carbamazepine, Phenobarbital, and Phenytoin) associated hypersensitivity skin reactions. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 16 The first four screen of MacBARDI –Q&A shows, where the prior odds will be calculated. It asks some relevant questions such as the adverse reaction experienced by the patient and the drug taken by the patient etc. For example if patient experienced exanthematous eruption (EE) while taking or after shortly taken a course of sulphonamide (D1), MacBARDI-Q&A uses this two piece of information and assess with the knowledge base, report it results. The risk of a sulfonamide induced rash is 1.5% and the background risk is 1.0%. MacBARDI spreadsheet calculates the prior odds 1.5:1 in favour of the drug (sulfonamide). Correspondence to 60% probability in favour of drug and 40% other causes. The accessor can verify the entire process of calculation by taking printout of MacBARDI spreadsheet. The impact of missing information can be assessed and the missing information can be added to the spreadsheet and can take repeat calculation again to get the exact assessment. As we discussed this BARDI (Bayesian system) has the opportunity to add up all the possible drugs and can check one by one, the system gives adjusted posterior probability of each drug. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 17 MacBARDI-Q&A WELCOME TO MacBARDI HYPERSENSITIVITY This differential diagnostic aid will assess the probability that an individual case of hypersensitivity rash was caused by a given drug. Its outputs is subject to the physician judgement in every case Let’s begin by determining what adverse event the individual in question experienced. Answer the following about the reaction: TYPE of RASH: EE= Exanthematous Eruption (enter this general rash if exact rash is unknown) UR= Urticarial Reaction VA= Vasculitis EM= Erythema Multiform ST= Stevens-Johnson Syndrome or toxin epidermal necrolysis ZZ= other RASH= ? (enter only 1st 2 letters of code in the box) Now, Lets know the possible drug of the adverse reaction: EE Which of the following drugs the patient was taking either at the time of reaction or within 5 days before? If the patient was taking >1choose 1 to run 1st. DRUG CLASS EVENTS IN KNOWLEDGE BASE D1= SULFONAMIDE EE.EM,UR and VA D2= PHENOBARBITAL EE,UR D3= CEFACIOR EE D4= ERYTHROMYCIN EE,UR and VA D5= AMOXICILLIN EM D6= CARBAMAZEPINE EE.EM,ST D7= PHENYTOIN EE,EM and ST D89= OTHERS DRUG = ? (enter the 2 digits code below box) Calculation of the PRIOR ODDS OK- at this point we can calculate the Prior Odds. The Prior Odds is the drug-attributable risk of the event divided by the background or nondrug risk of having the event: And has taken drug: D2 The risk of a drug-induced rash occurring is: 1.5% And the background risk of this event is: 1.0% Therefore the prior odds is: This means that the odds in favour of the drug in question causing this adverse event are: 1.5 to 1 in favour of the drug. This corresponds to a probability of: 60% in favour of the drug for the typical patient. This is the actual printout from MacBARDI-Q&A for the above given example. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 18 Processes for Establishing the Relationship To ensure a methodological approach for establishing the relationship, (cohort event monitoring) CEMFLOW can be used. CemFlow is a preferred method for clinical reviewer to enter the data and ensures a systematic and standard approach to reviewing the events reported on the questionnaires. Pharmacovigilance center head and CemFlow clinical coordinator can approach UMC for user registration.(email: info@who-umc.org).. CemFlow is available online and offline coding sheet (following should be recorded systematically) Result of dechallenge (Select the most suitable outcome) Event Resolved; Resolving; Resolved With Sequelae; Not Resolved; Worse; Death; Unknown; No Dechallenge (medicine continued). Result of rechallenge with the same medicine by itself (Select the most suitable outcome) Recurrence Of Problem; No Recurrence; Unknown; No Rechallenge. Outcome of the event (Select the most suitable term) Resolved; Resolving (As At A Recorded Date); Not Resolved; Worse; Patient Died; Permanent Disability; Unknown. Clinical details (collect most appropriate details) Concomitant disease(s); Relevant patient history, e.g. liver disease or renal disease; Previous exposure to same medicine(s): yes or no? (or) any reaction to previous exposure – yes or no? or if “yes”, record the reaction term(s) for previous reaction(s). Logic check As it is very difficult to assess the drug and reaction relationship. When doing final step use logic considerations like; a) If there has been no rechallenge, or the outcome of rechallenge is unknown, it is not ‘Definite/Certain’ b) If there has been no dechallenge or the result of dechallenge is unknown, it is not ‘Probable’ c) If the outcome of the event is unknown or if there are other possible causes of the event, then it is not a ‘Probable’ Reactions and incidents It is useful for analytical purposes to divide the events into 2 groups (reactions and incidents). i) Events which can be consider as definite, probable or possible can be aggregated and referred to as “reactions” because the event is likely to be related to the medicine(s). ii) Events which can be consider as unlikely, can be referred to as “incidents” because they appear to be incidental to the use of the medicine. iii) The following equation simplifies the concept and is consistent with the definition of “events”: events = reactions + incidents advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 19 Bibliography:1. Clinical pharmacy and therapeutics by Roger Walker and Cate Whittlesea: 5 th edition;chapter 5;Churchill livingstone Elsevier; 2012; pg.no 2. Stephens’ Detection of New Adverse Drug Reactions; john Talbot and Patric Waller; chapter 2;john wiley & amp; sons,ltd;2004; pn.no 3. A text book of clinical pharmacology and therapeutics by James M Ritter , Lionel D Lewis, Timothy GK Mant and Albert Ferro; 5 th edition; chapter 12: hodder Arnold – a part of Hachette liver UK;2008; pg.no 4. A text book of clinical pharmacy practice: G. Parthasarathi, K.N. Hansen, M.C. Nahata-Ed-2 nd ; chapter 9: universities press; 2012;pg.no 5. Clinical Pharmacy by HP Tippins and Amrita Bajaj; 2 nd edition; chapter 9;career publications;2011; pg.no 6. Textbook of Pharmacovigilance : concept and practice by GP Mohanta and PK Manna;chapter 4;pharmamed press;216; pg.no 7. Adverse Drug Reaction by By Stephanie N. Schatz and Robert J. Weber in PSAP 2015 •CNS/Pharmacy Practice;2015; Pg.no 8. Adverse drug reaction – causality assessment by srinivas R and ramya G; ijrpc 2011.1(3) 9. A Text Book of Pharmacy Practice by K.G. Ravi Kumar ; 2nd edition; chapter 7; pharma carrier publication; 2018; pg.no 10. The Australian method of drug-event assessment (special workshop-regulatory) by Maurice L, Mashford, MD; drug information journal, 1984, vol 18, pp271-273. 11. Microcomputer-assisted bayesian differential diagnosis of severe adverse reactions to new drugs: a 4-year experience by; Claudio A. Naranjo, , Krista L. Lanctot; drug information journal, vol, 25, pp,243-250,1991. 12. Computer-Assisted Evaluation of Adverse Events Using a Bayesian Approach by; Krista L. Lanctot, and Claudio A. Naranjo, J clin Pharmacol 1994;34:142-147 13. Comparison of the Bayesian approach and a simple algorithm for assessment of adverse drug events by; Krista L. Lanctk, and Claudio A. Naranjo; clinical pharmacology & therapeutics; vol 58, number 6; pg.no 692 - 698 14. Causality Assessment in Pharmacovigilance: A Step Towards Quality Care by; Dr. Rohan C. Hire, Dr. Parag J. Kinage, Dr. Nilesh N. Gaikwad; Scholars Journal of Applied Medical Sciences., 2013; 1(5):386-392. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 20 15. Bracken LE, Nunn AJ, Kirkham JJ, Peak M, Arnott J, Smyth RL, et al. (2017) Development of the Liverpool Adverse DrugReaction Avoidability Assessment Tool. PLoSONE 12(1): e0169393. doi:10.1371/journal.pone.0169393 16. A Practical Handbook On The Pharmacovigilance Of Antiretroviral Medicines:world health medicine; Relationship/causality assessment; 57-64 17. ADRIC: Adverse Drug Reactions In Children – a programme of research using mixed methods by; Rosalind L Smyth, Matthew Peak; national institute for health research ; chapter 5 ; vol 2 (3), june 2014, pg.no-78. 18. The use of the WHO-UMC system for standardised case causality assessment by; UMC. 19. Pande S. Causality or Relatedness assessment in adverse drug reaction and its relevance in dermatology. Indian J Dermatol 2018;63:18-21. 20. A method of estimating the probability of adverse drug reaction by; C. A. Naranjo, M.D., U.edward MS ,Busto; clinical pharmacology and therapeutics; dec 2015; pg.no 239-40. advclinpharmaedu.com | Pharmacovigilance and ADRs | Causality Assessment of ADRs 21