Routine signal detection and statistical tools on paediatrics - PowerPoint PPT Presentation

Routine signal detection and statistical tools on paediatrics Paediatric workshop – 28 April 2014 Presented by: Cosimo Zaccaria Signal Management – Pharmacovigilance Department An agency of the European Union

How to im prove m onitoring of drugs in the paediatric population Pharm acovigilance 1 ) Paediatric query 2 ) Paediatric Signal Detection 2 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

• PhV obligations for routine monitoring of ADRs • Stastical and clinical relevance • e-RMR: a tool for signal detection • Medication error in children • Update on future developments • Conclusion 3 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

LEGAL BACKGROUND 4 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Signal definition Signal is an information that arises from one or m ultiple sources (including observations and experiments), which suggests a new potentially causal association, or a new aspect of a know n association , between an intervention and an event or set of related event, either adverse or beneficial, that is judged to be of sufficient likelihood to justify verificatory action. Report of the Council for International Organisations of Medical Sciences WG VIII, Practical Aspects of Signal Detection in Pharmacovigilance (CIOMS, Geneva 2010). 5 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Sources of information–Hypothesis Generating Spontaneous ADR reporting systems Clinical trials data Source of all signals evaluated by EMA in 2 0 1 3 ( n= 2 4 4 9 ) Pharm acovigilance Pharmacoepidemiological studies Scientific literature Non clinical trial data e.g non- interventional studies 6 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Sources of information–Hypothesis Generating 91% of signals evaluated by EMA originate from EudraVigilance 70% of signals validated and forwarded to PRAC by EMA originate from EudraVigilance Clinical trials data Source of all signals evaluated by EMA in 2 0 1 3 ( n= 2 4 4 9 ) 91% Pharm acovigilance 5% 3% 1% EV Literature Other RAs Other 7 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Lim itations of the spontaneous reporting • Under reporting and reporting bias • No drug exposure data • No adverse event natural history ( prevalence, incidence) • No indication epidem iology ( prevalence, incidence) • Data quality and m issing data • Confounders 8 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Signal detection lim itation Fishing in the Fishing w ith the w rong tool w rong place 9 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

DME = designated medical events List of MedDRA term s w ith high m ortality rate and high likelihood to be drug related. The list was introduced in 2012 to serve as safety net in signal detection and ensure that important events requiring review would not be missed. Renal failure Dermatitis exfoliative Acute hepatic failure Renal failure acute Acute respiratory distress Dermatitis exfoliative generalised Respiratory failure syndrome Reye's syndrome Disseminated intravascular coagulation Acute respiratory failure Rhabdomyolysis Drug reaction with eosinophilia and systemic Agranulocytosis Status epilepticus symptoms Anaemia haemolytic autoimmune Stevens-Johnson syndrome Epilepsy Anaphylactic reaction Subacute hepatic failure Erythema multiforme Anaphylactic shock Sudden cardiac death Febrile bone marrow aplasia Anaphylactoid reaction Sudden death Febrile neutropenia Anaphylactoid shock Sudden hearing loss Angioedema Gastrointestinal mucosal necrosis Sudden visual loss Aplasia pure red cell Gastrointestinal necrosis Suicidal behaviour Aplastic anemia Grand mal convulsion Suicidal ideation Asterixis Granulocytopenia Suicide attempt Autoimmune hepatitis Haemoglobinaemia Thrombotic thrombocytopenic Autoimmune neutropenia Haemoglobinuria purpura Autoimmune pancreatitis Haemolysis Torsade de pointes Autoimmune pancytopenia Haemolytic anaemia Toxic epidermal necrolysis Autoimmune thrombocytopenia Haptoglobin decreased 10 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Signal Validation Signal Managem ent Statistical approach: SDRs/ PRR Signal confirm ation Signal Detection Clinical approach: • Temporal association PRAC • Biological plausibility • Dechall.-Rechallenge Signal Assessm ent / Recom m . 11 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Hypothesis = Statistical relevance Dynam ic vs Static PRR overtim e Disproportional Analysis PRR = a/(a+b) c/(c+d) PRR PRR(-) = 1: no reporting difference PRR(-) > 1: there is a difference PRR threshold in EV for a potential signal = Lower bound of the 9 5 % Confidence I nterval of PRR ≥ 1 12 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Hypothesis = Clinical relevance I CSR ( CI OMS I ) • Biological plausibility • Dechallenge / Rechallenge ( + / -) • Tim e to Onset: tem poral association • Confounders: co- suspected / concom itant drugs • Underlying disease 13 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

SI GNAL VALI DATI ON Causality Assessm ent Rational • VERY LIKELY/CERTAIN: plausible TTO, with no alternative explanations • PROBABLE: reasonable TTO, unlikely attribute to alternative explanations • POSSIBLE: reasonable TTO but it could also be attributed to alternative explanation • UNLIKELY: improbable TTO also attribute to underlying disease and concomitant drugs • UNRELATED: incompatible TTO and confounded by underlying disease and concomitant drugs • UNASSESSABLE: insufficient information Rational • ONTOLOGY = the information in the SmPC is unable to convey the risk • FREQUENCY = increasing or of concern in the subsets analysed • SERIOUSNESS = outcome of the ADR is more serious than expected • TREATMENT = specific action that the prescriber would not consider • DIAGNOSIS = specific clinical manifestation or test that HCP would not expect causing a delay • PREVENTION = monitoring the specific subset to enforce the preventability of the ADR 14 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Signal Assessm ent by PRAC 15 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

eRMR ( electronic Reaction Monitoring Report) electronic tool for SD: EV~ 4 5 0 0 I CSRs per day - Save tim e in screening - Sim plify the screening process - Avoid duplication of work tracking all issue - Build-up a knowledge overtime about your product - I m prove reliability matching different sources of information (PSUR, SPC, RMP, US_PI etc.) - Sim plify access to the cases 16 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

1 / 2 eRMR – STRUCTURE & CONTENT Use of agreed term inology MedDRA Structure Active substances Drug x SMQs ( narrow ) Drug x HLGTs Drug x SOCs HLTs PTs Drug x Drug x Drug x Drug x Drug x 17 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

1 / 2 eRMR – Structure: EV data and sub- categories EV Observational studies Spontaneous cases Med. Err./ Abuse Med. Err./ Abuse Paediatric cases Paediatric cases Literature cases Geriatric cases Geriatric cases EV Cases Serious cases PRR( -) / PRR Fatal cases HCP cases EEA cases CT cases 18 eRMR: structure and content and optimal use for signal detection - C. Zaccaria Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Priority 19 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

eRMR – Monitoring and Tracking Structure: interactive functionalities Operational colum ns Several sources of information MedDRA PT code combined in the eRMR: SmPC, PRR, SDRs, Signal Status RMP , PSUR, DME, IME etc. Com m ents reviews etc. I ME/ DME Changes SDR ( n) Priority 20 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Focusing on certain ADRs Screening the eRMR Prioritisation of ADRs Pr1 = D esignated M edical E vents Pr2 = I mportant M edical E vents with a S ignal Of D isproportionate  DME  I ME/ SDR R eporting  Fatal  Literature Pr3 = Important and non Important Medical Events with a Fatal outcome or Paediatric Cases 21 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Type of ADRs in EV: Paediatrics Vs Adults 22 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department

Main differences ... Child vs Adult • PhD and PhK increased vulnerability and different drug interaction profile; • Eccipients increased susceptibility; • Drug induced grow th and develp. Disorders drug exposure at a sensitive point in development (critical window) • Paediatric ADRs specific only to children. In utero exposure is an additional risk factor • Chronic disease increased risk for life-long exposure • Unable to com m unicate ADRs clearly 23 Cosimo Zaccaria - Signal Management – Pharmacovigilance Department