Table 1. Targeted population, weights,... | American Society of Hematology

Table 1.

Targeted population, weights, and estimands

Method for controlling confounders	Weights for treated, untreated	Target population	Estimand
Inverse weighting	$\frac{1}{e (x)}$ , $\frac{1}{(1 - e (x))}$	Combined from the treated and untreated	ATE
	1, $\frac{e (x)}{(1 - e (x))}$	Treated population	ATT
	$\frac{1 - e (x)}{e (x)}$ ,1	Control population	ATC
	1-e $(x)$ , e $(x)$	Overlapping population	ATO
	$\frac{1 (a < e (x) < 1 - a)}{e (x)}$ , $\frac{1 (a < e (x) < 1 - a)}{(1 - e (x))}$	Trimming population	Not specified
Matching	$\frac{Min (e (x), 1 - e (x))}{e (x)}$ , $\frac{Min (e (x), 1 - e (x))}{(1 - e (x))}$	Matching population	ATT
Matching-adjusted indirect comparison	$\frac{1 - e (x)}{e (x)}$ ,1	Control population	ATC

Method for controlling confounders	Weights for treated, untreated	Target population	Estimand
Inverse weighting	$\frac{1}{e (x)}$ , $\frac{1}{(1 - e (x))}$	Combined from the treated and untreated	ATE
	1, $\frac{e (x)}{(1 - e (x))}$	Treated population	ATT
	$\frac{1 - e (x)}{e (x)}$ ,1	Control population	ATC
	1-e $(x)$ , e $(x)$	Overlapping population	ATO
	$\frac{1 (a < e (x) < 1 - a)}{e (x)}$ , $\frac{1 (a < e (x) < 1 - a)}{(1 - e (x))}$	Trimming population	Not specified
Matching	$\frac{Min (e (x), 1 - e (x))}{e (x)}$ , $\frac{Min (e (x), 1 - e (x))}{(1 - e (x))}$	Matching population	ATT
Matching-adjusted indirect comparison	$\frac{1 - e (x)}{e (x)}$ ,1	Control population	ATC

$e (x) = P S = \Pr (T = 1 | V)$ is the propensity score, where T = 1 for the single-arm treatment group, T = 0 for the external control group, and V is the set of observed confounders in both groups.

ATO, average treatment effect in the overlap population.

View Large

This Feature Is Available To Subscribers Only