Skip to main content
. 2021 Sep 7;30(12):1703–1715. doi: 10.1002/pds.5346

TABLE 5.

Impact of different values of nonadherence in comparative effectiveness research

Truth Observed data
Paroxetine use versus nonusers Drug P1 Drug P0 Drug P1 Drug P0
Y = 1 150 9900 217 9833
Y = 0 850 89 100 1450 88 500
Total 1000 99 000 1667 98 333
Risk 0.15 0.10 0.13 0.10
RR 1.5 1.4
Escitalopram versus nonusers Drug E1 Drug E0 Drug E1 Drug E0
Y = 1 75 9950 87.5 9937.5
Y = 0 425 89 550 537.5 89437.5
Total 500 99 500 625 99 375
Risk 0.15 0.10 0.14 0.10
RR 1.5 1.8
Escitalopram versus paroxetine Drug P1 Drug E1 Drug P1 Drug E1
Y = 1 150 75 217 87.5
Y = 0 850 425 145 537.5
Total 1000 500 1667 625
Risk 0.15 0.15 0.13 0.14
RRE−P 1.0 1.08

Note: Values used: baseline risk: 0.1; RR: paroxetine 1.5, escitalopram 1.5; exposure prevalence: paroxetine 0.01, escitalopram 0.005; data capture: paroxetine 1.0, escitalopram 1.0; adherence: paroxetine 0.6, escitalopram 0.8.

Abbreviations: P1: exposed to paroxetine; E1 exposed to escitalopram; P0 and E0 nonexposed to paroxetine or escitalopram; Y = outcome.