The goal of PVomega is to detect drug-drug-event statistical patterns supporting hypotheses generation on drug-drug interactions in databases of adverse event reports.
Installation
You can install the development version of PVomega like so:
install.packages("pak")
pak::pak("Uppsala-Monitoring-Centre/PVomega")and load the library like so:
What Omega measures
Omega (Noren et al. 2008) is a disproportionality measure for drug-drug-event combination. It compares the observed number of reports listing two drugs and an adverse event with the number expected if the two drugs did not interact.
“No interaction” is defined by an additive model: the reporting rate of the event under both drugs is the background reporting rate plus the contribution attributable to each drug. This is the key difference from logistic regression, whose interaction term assumes multiplicative effects of separate drugs and misses well-established interactions such as cerivastatin with gemfibrozil.
The steps are:
- Count the reports in each of the four exposure strata: neither drug, D1 only, D2 only, both drugs.
- Compute the relative reporting rate of the event in each stratum:
f00,f10,f01,f11. - Estimate the expected rate under both drugs,
g11, fromf00,f10andf01only (eq. 16 of the paper). - Compute the expected count
E111 = g11 * n11.. - Compute the shrunk ratio
omega = log2((n111 + 0.5) / (E111 + 0.5))and its Gamma credibility interval.
A potential interaction is highlighted when omega_lower is above 0.
Starting from counts
omega_from_counts() works with any data source. The example below is the first case study of the paper: delayed bleeding with itraconazole and oral contraceptives.
res <- omega_from_counts(
n111 = 10, n11. = 23, n1.1 = 10, n.11 = 19,
n1.. = 39, n.1. = 1489, n..1 = 39, n... = 5503
)
res[, .(f00, f10, f01, f11, g11, E111, omega, omega_lower, omega_upper)]
#> f00 f10 f01 f11 g11 E111 omega
#> <num> <num> <num> <num> <num> <num> <num>
#> 1: 0.005002501 0 0.006139154 0.4347826 0.006139154 0.1412005 4.03347
#> omega_lower omega_upper
#> <num> <num>
#> 1: 3.003327 4.790041Starting from DiAna FAERS data
With the DiAna data and package imported, the DiAna function omega_analysis() computes all counts and returns one row per drug1-drug2-event combination (script not run).
library(DiAna)
FAERS_version <- "24Q4"
import("DEMO")
import("DRUG")
import("REAC")
res <- omega_analysis(
drug1_selected = "miconazole",
drug2_selected = "warfarin",
reac_selected = c("international normalised ratio increased", "haemorrhage"),
restriction = Demo[!RB_duplicates_only_susp]$primaryid
)
resGroups of terms can be collapsed with lists, as in DiAna:
Methodological choices to report
Three choices for the parameters in the DiAna function omega_analysis() change the counts, and should be stated in any publication:
-
Drug roles (
drug_roles). By default only primary suspect, secondary suspect and interacting drugs count as exposure, as at Uppsala Monitoring Centre. Reports where a drug is only concomitant fall into the background stratum. -
Deduplication (
restriction). FAERS contains duplicate reports. Pass the primary IDs retained by one of DiAna’s deduplication algorithms. -
Report universe. The total
n...is the number of distinct reports intemp_reac(withinrestriction). Every count refers to this population.
Interpreting results
As recommended in the original paper, do not read Omega in isolation. The columns f00, f10, f01 and f11 show why a combination was highlighted. For example, a large f10 means D1 alone already explains much of the reporting.
References
Noren et al. 2007
Noren GN, Sundberg R, Bate A, Edwards IR. A statistical methodology for drug-drug interaction surveillance. Research Report. 2007;2007:6. Mathematical Statistics, Stockholm University.
Noren et al. 2008
Noren GN, Sundberg R, Bate A, Edwards IR. A statistical methodology for drug-drug interaction surveillance. Stat Med. 2008;27(16):3057-70.
Hult et al. 2020
Hult S, Sartori D, Bergvall T, et al. A feasibility study of drug-drug interaction signal detection in regular pharmacovigilance. Drug Saf. 2020;43:775-85.
