PV Calculation Engine
Pharmacovigilance computation tools: signal detection (PRR/ROR/IC/EBGM), causality assessment (Naranjo, WHO-UMC), seriousness classification (ICH E2A), benefit-risk analysis, and reporting rate calcul
17 toolscalculate
Compute Prr
Compute Proportional Reporting Ratio (PRR) for a drug-event pair from a 2x2 contingency table. PRR > 2 with chi-squared
abcd
Compute Ror
Compute Reporting Odds Ratio (ROR) for a drug-event pair from a 2x2 contingency table. ROR > 1 with lower 95% CI > 1 ind
abcd
Compute Ic
Compute Information Component (IC/BCPNN) for a drug-event pair. IC > 0 with IC025 (lower credibility interval) > 0 indic
abcd
Compute Ebgm
Compute Empirical Bayesian Geometric Mean (EBGM/GPS) for a drug-event pair. EBGM > 2 with EB05 (lower bound) > 1 indicat
abcd
Assess Naranjo Causality
Compute Naranjo adverse drug reaction probability score (0-13). Categories: Definite (>=9), Probable (5-8), Possible (1-
previous_reportsafter_drugimproved_on_withdrawalreappeared_on_rechallengealternative_causesplacebo_reactiondrug_detecteddose_relatedprevious_exposureobjective_evidence
Classify Seriousness
Classify adverse event seriousness per ICH E2A criteria. An event is serious if it meets ANY criterion: death, life-thre
resulted_in_deathlife_threateningrequired_hospitalizationresulted_in_disabilitycongenital_anomalymedically_important
Compute Benefit Risk
Compute a quantitative benefit-risk ratio using the NexVigilant QBR framework. Scores benefit (efficacy × population imp
efficacy_scorepopulation_impactrisk_severityrisk_frequencyrisk_detectability
Compute Disproportionality Table
Compute ALL four disproportionality measures (PRR, ROR, IC, EBGM) simultaneously from a single 2x2 contingency table. Re
abcd
Assess Who Umc Causality
Assess drug-event causality using the WHO-UMC system. Categories: Certain, Probable/Likely, Possible, Unlikely, Conditio
temporal_relationshipknown_responsedechallenge_positiverechallenge_positivealternative_explanationsufficient_information
Compute Reporting Rate
Compute adverse event reporting rate per unit exposure. Normalizes raw case counts by exposure denominator (prescription
case_countexposure_denominatordenominator_unit
Compute Signal Half Life
Compute signal persistence using exponential decay model. Estimates how long a safety signal remains actionable based on
initial_signal_strengthdecay_rate
Compute Expectedness
Assess whether an adverse event is expected (listed) or unexpected (unlisted) based on reference safety information. Une
event_termdrug_name
Compute Time To Onset
Analyze time-to-onset patterns using Weibull distribution. Shape parameter k indicates: k<1 early hazard (direct pharmac
onset_days
Score Case Completeness
Score ICSR completeness against E2B(R3) minimum data elements. Checks required fields (patient, reporter, drug, event) a
patient_identifierreporter_identifiersuspect_drugadverse_event
Compute Number Needed Harm
Compute Number Needed to Harm (NNH) from exposed and unexposed incidence rates. NNH = 1/ARI where ARI = |risk_exposed -
risk_exposedrisk_unexposed
Compute Confidence Interval
Compute Wilson score confidence interval for a proportion. More reliable than Wald CI for small samples and extreme prop
successestotal
Compute Signal Trend
Detect trend direction in time-series safety signal scores using linear regression. Reports slope, R-squared, and whethe
observations