TABLE 1.
Parameter | Explanation | Range of values |
---|---|---|
X | Dichotomous exposure | 0 or 1 |
Y | Count outcome | ≥0 |
Ci | Ten dichotomous confounders | 0 or 1 |
P(Ci) | Prevalence of confounder i | 5% to 25% in increments of 5% |
β0 | Baseline exposure prevalence | 25% or 50% |
βCi | Strength of the confounder-exposure association | log(1.5) to log(4.0) in increments of 0.5 |
βPS−INT | Strength of the C1 by C2 within-PS interaction | log(1.0) log(3.0) to log(5.0) in increments of 0.5 |
γX | Treatment effect | log(1.0) to log(5.0) in increments of 1.0 |
γ0 | Baseline outcome rate | 0.01, 0.10, 0.25 or 0.50 events per unit of person-time |
γCi | Strength of the confounder-outcome association | log(1.5) to log(4.0) in increments of 0.5 |
γTX−INT | Strength of the C1 by treatment interaction | log(1.0) log(3.0) to log(5.0) in increments of 0.5 |
PSCOH | Propensity score estimated in the entire cohort | 0 to 1 |
PSSS | Propensity score estimated in the subgroup | 0 to 1 |
a | Number of exposed outcomes in the entire cohort | As observed |
aC1=1 | Number of exposed outcomes in the subgroup of C1 =1 | As observed |