Pharmacovigilance Signal Detection: PRR, ROR, Chi-Square & Case Counts

What you'll build

What you will be able to do

Each lab ends with a concrete result that you build yourself and the checker confirms, cell by cell.

• Turn a list of 50 spontaneous reports into the 2 × 2 table behind every disproportionality statistic, with one COUNTIFS per cell, and reconcile it with the list.
• Calculate the PRR for Velorantib and pancreatitis, its standard error on the log scale and the 95% confidence limits.
• Screen eight drug–event pairs at once: the ROR, its log-scale standard error and the 95% limits, then count the pairs whose lower limit is above 1.
• Test whether Corimide and hepatitis are reported together more often than chance: expected count, χ² with and without Yates' continuity correction, and the p-value.
• Apply the classic three-part screening rule to ten drug–event pairs, count how many criteria each pair meets, flag the signals and colour their rows.
• Calculate the time to onset for 30 Velorantib–pancreatitis cases, summarise it with the median and quartiles, bin it into onset windows and chart the distribution.
• Follow pancreatitis reports for Velorantib over eight quarters: the share of all reports, the cumulative count, the jump after a media story, and a combo chart.
• Summarise the 24 well-documented Velorantib–pancreatitis cases the way a signal assessment report does: share of serious cases, median age, outcomes and dechallenge.