Table 5.

Mean Average Precision at K.

	Baselines		Statistical Models				Causal Models
Adverse Event Type [+ ctrls, − ctrls]	ROR	χ 2	${\beta }_{sql}^5$	${\beta }_{psi}^5$	${\beta }_{sql}^{10}$	${\beta }_{psi}^{10}$	${\mathrm{\Delta }}_{sql}^5$	${\mathrm{\Delta }}_{psi}^5$	${\mathrm{\Delta }}_{sql}^{10}$	${\mathrm{\Delta }}_{psi}^{10}$
K = 10
Acute kidney failure [10+, 9−]	0.4756	0.4921	0.4106	0.3068	0.4768	0.4135	0.5378	0.7032 †	0.4756	0.5653
Acute liver failure [43+, 11−]	0.8955	0.8917	0.8389	0.7579	0.8803	0.725	0.7972	0.8951	0.6815	0.9627 †
Acute myocardial infarction [15+, 23−]	0.4889	0.5981	0.4974	0.4021	0.4911	0.4155	0.5193	0.8529 †	0.7296	0.5883
Gastrointestinal hemorrhage [20+, 32−]	0.5556	NA	NA	NA	0.1	0.125	NA	0.325	0.125	0.2
K = 25
Acute kidney failure [10+, 9−]	0.4866	0.515	0.4751	0.4166	0.5057	0.4665	0.5422	0.7277 †	0.5229	0.5477
Acute liver failure [43+, 11−]	0.7993	0.7863	0.7767	0.7132	0.8449	0.7067	0.8163	0.8764	0.7085	0.9051 †
Acute myocardial infarction [15+, 23−]	0.3795	0.6056	0.54	0.4227	0.455	0.4261	0.5092	0.6715 †	0.6595	0.5618
Gastrointestinal hemorrhage [20+, 32−]	0.4258†	0.2093	0.224	0.1937	0.255	0.1857	0.3025	0.3739	0.2361	0.2626

ROR = reporting odds ratio. χ² = chi squared. β = generalized linear models (multiple variable logistic regression). Δ = average treatment effect using graphical causal models. Results exceeding best baseline performance are in boldface.

^†indicates best performance for a side effect.