Disproportionality methods are largely used to identify statistical associations between products and events in their respective databases of safety reports. They should be distinguished from ‘potential signals’ that can originate from individual case analysis and formal epidemiological studies.
Proportional Reporting Ratio (PRR): The PRR is a simple way to get a measure of how common an adverse event for a particular drug is compared to how common the event is in the overall database.
It is used to measure the strength of the statistical association between a risk factor (specific drug) and a condition (specific adverse event).
The PRR allows the comparison of frequencies of reporting, in order to determine if there is a disproportionate reporting of a specific adverse event with a specific drug compared to other adverse events and other drugs.
This concept of disproportionality may be displayed by means of a 2 x 2 contingency table where
- “a” is all reports for a specific adverse event (“Event Y”) for Product (e.g. a drug) X,
- “b” is all reports for all other adverse events for Product X,
- “a + b” are all the reports for Product X,
- “c” is all reports for all other products for Event Y,
- “d” is all reports for all other products for all other adverse event, and
- “c + d” is all reports for all other products.
The PRR = [a/(a+b)] / [c/(c+d)]
If the ratio of [a/(a+b)] is greater than the ratio of [c/(c+d)], then Event Y is “disproportionately reported” for Product X, when the rest of the database is considered as a background “expected.”
A PRR value higher than 1 indicates a higher probability that the event under consideration occurs in treated patients with the considered product compared to this event occurring with other products, while a PRR of 2 for a drug event combination indicates that the proportion of reports for the drug-event combination is twice the proportion of the event in the overall database.
Example 1:
- Proportion of individual cases of nausea involving a medicinal product ‘Trade Name’ = 5% (e.g. 5 reports of nausea amongst a total of 100 reports reported with medicinal product ‘Trade Name’).
- Proportion of reports of nausea involving all the other medicinal products in a database (but medicinal product ‘Trade Name’) = 5% (e.g. 5000 reports of nausea amongst 100,000 reports reported with all other medicinal products). Therefore, the PRR is equal to 1 (0.05/0.05).
Example 2:
- Proportion of individual cases of nausea involving medicinal product ‘Trade Name’ = 15% (e.g. 15 reports of nausea amongst a total of 100 reports reported with medicinal product ‘Trade Name’).
- Proportion of individual cases of nausea involving all other medicinal products in a database (but medicinal product ‘Trade Name’) = 5% (e.g. 5000 reports of nausea amongst 100,000 reports reported with all other medicinal products). Therefore, the PRR is equal to 3 (0.15/0.05).
The PRR measures a reporting relationship between a medicinal product and an adverse event on the basis of a relative increase of the proportion of individual cases related to an adverse event. This does not necessarily imply a causal relationship between the administered medicinal product and the occurrence of the adverse event. Consequently, the signals identified with quantitative methods should always be medically assessed.
Below elements influence the value of the PRR and may induce masking effects. Alternatively they may exaggerate the importance of a medicinal product-adverse event statistical association. Hence for PRR interpretation below elements should take into account:
- The type of medicinal products included in the database
- The medical terminology(ies) applied
- The coding practices
- The date of the creation of the database
- The source of ICSR (all unsolicited reports)
Leave a Reply