Causality Assessment in postmarketing adverse events

1
Causality Assessment in postmarketing adverse
events
Anshu Vashishtha MD PhD Watson Pharmaceuticals
2
Evaluation of causality (Question 3)

• Pros and Cons
• Criteria to consider
• Examples of methods of assessment- WHO Index
  cases, PRR, Global Clinical Assessment, broad
  categorization
• Going forward

3
Evaluation of causality

• Individual level
• Case Collection
• Importance of Case definition (confirmation of
  diagnosis) CIOMS , MSSO
• Importance of quality
• Necessity reported adverse event maybe related
  to concomitant medication or coexisting disease

4
Individual adverse event causality assessment
-Pros

• Better case quality and follow up
• Surfaces confounding factors and follow- up needs
  while case is fresh
• Allows ongoing signal assessment by number of
  reports considered probably, possibly or unlikely
  to be related

5
Individual adverse event causality assessment
-Pros

• More meaningful Package insert
• If reported cases are not considered causally
  related, may keep those events out of package
  insert
• If assessed as causally related, allows true
  risks to be communicated better

6
Individual adverse event causality assessment
-Cons

• Takes time and resources - some approaches more
  conducive to large scale application
• Limited by lack of all relevant data in several
  instances
• Legal liability of sponsor?

7
Criteria to consider for causality assessment

• positive rechallenge
• positive dechallenge (resolution upon stopping
  suspect drug, in absence of other intervention or
  treatment)
• known class effect
• biological plausibility
• lack of alternative explanation - concomitant
  drug or disease
• typical adverse drug reaction (low background
  rate)
• Reference Guidelines for preparing Core Clinical
  Safety Information on Drugs
• Report of CIOMS Working Group III, 1995

8
Criteria to consider for causality assessment -2

• Dose response
• Lack of concomitant factors (clean subject eg.
  child)
• Consistency of time to onset eg. early for
  immediate hypersensitivity or long term for
  tumorigenesis
• High frequency of reported cases
• Similar findings in toxicity studies
• Reference Guidelines for preparing Core Clinical
  Safety Information on Drugs
• Report of CIOMS Working Group III, 1995

9
Criteria to consider for causality assessment -3

• Positive in vitro test eg. IgE antibodies to
  allergen and elevated serum serum tryptase in
  anaphylaxis
• positive in vivo test eg. Intradermal or prick
  test for immediate hypersensitivity or patch test
  for delayed
• Identified subset at risk or predisposing factor
• Reference Guidelines for preparing Core Clinical
  Safety Information on Drugs
• Report of CIOMS Working Group III, 1995

10
Criteria to consider for causality assessment -4

• Protopathic bias drug given to treat early
  symptoms may appear temporally associated with
  the subsequent illness
• Stimulated Reporting- recall bias after publicity

11
Methods for causality assessment

• Evaluating Clinical Differential Diagnosis or
  algorithms
• Factors to judge (Lane and Hutchison , 1986)
• Repeatability
• Explicitness
• Explanatory culpability
• Completeness
• Biological Balancing
• No a priori constraints
• From Detection of New Adverse Drug Reactions
  Editor, MDB Stephens, JCC Talbot, PA Routledge,
  4th edition

12
Methods for causality assessment WHO index cases

• WHO (Edwards et al , 1990) Three index cases
• Substantial Index case has information on all
  eleven major items and lack of confounding
  variables (feasible on first six)
• Information on source of case, identification of
  case, description of reaction, name of drug,
  treatment dates, reaction date,
• age, sex, all drugs with doses and dates,
  indication for treatment/ underlying disease,
  outcome
• 1 index case 2 substantial or 4 feasible cases
• From Detection of New Adverse Drug Reactions
  Editor, MDB Stephens, JCC Talbot, PA Routledge,
  4th edition

13
Probability of 3 cases occurring with 300,000
treated patients (Begaud et al, 1995)
14
Proportional reporting rate

• PRR AD/ BC
• Signal (PRR gt3) (Chi -squared gt5) (Number of
  casesgt3)
• Remains one aspect of causality assessment
  subject to usual biases inherent in spontaneous
  reporting (Waller 1998)

15
Evaluation of causality- broad categorization

• Example
• Category A Good reasons and sufficient
  documentation to assume causal relationship
• Category B connection uncertain and even
  doubtful, eg. Because of missing data
• Category O Not assessable because of missing or
  conflicted data
• Causality classification at Pharmacovigilance
  centers in the European Community Meyboom RHB
  and Royer RJ, Pharmacoepidemiology and Drug
  Safety , Vol 1 87-97(1992)

16
Evaluation of causality- way forward

• Individual level- research European agencies
  experience (eg. French require causality
  assessment by sponsors)
• Case Collection- necessary for signal evaluation-
  WHO, algorithms or global introspection as a
  guide
• Importance of Case definition (confirmation of
  diagnosis), good case quality
• Grades of causal likelihood- role of algorithms
  needs evaluation

17
Evaluation of causality- way forward- 2

• Need to
• evaluate existing methods and experience on
  worldwide basis
• agree on approach to categorization into broad
  zones of likely relationship of reported
  adverse events
• conduct retrospective evaluation for utility of
  causal assessment - initially using existing data
  from FDA database
• consider pilot on certain drugs
• determine overall guidance based on above results