Individualized Decision Making Under Partial Identification: ThreePerspectives, Two Optimality Results, and One Paradox

Cui, Yifan

doi:10.1162/99608f92.d07b8d16

Cited by 5 publications

(4 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“… 47 Importantly, bounds on the additive effect can be used in formal decision theoretic approaches, even if these bounds are wide or cover null effects. 48 , 49 Furthermore, if the investigator is willing to invoke assumptions about the probability of exposure, the bounds will be narrower, as we describe in Section External Data and Sensitivity Analysis.…”

Section: Discussionmentioning

confidence: 99%

Identification of Vaccine Effects When Exposure Status Is Unknown

Stensrud

Smith

2023

Epidemiology

View full text Add to dashboard Cite

Results from randomized controlled trials (RCTs) help determine vaccination strategies and related public health policies. However, defining and identifying estimands that can guide policies in infectious disease settings is difficult, even in an RCT. The effects of vaccination critically depend on characteristics of the population of interest, such as the prevalence of infection, the number of vaccinated, and social behaviors. To mitigate the dependence on such characteristics, estimands, and study designs, that require conditioning or intervening on exposure to the infectious agent have been advocated. But a fundamental problem for both RCTs and observational studies is that exposure status is often unavailable or difficult to measure, which has made it impossible to apply existing methodology to study vaccine effects that account for exposure status. In this study, we present new results on this type of vaccine effects. Under plausible conditions, we show that point identification of certain relative effects is possible even when the exposure status is unknown. Furthermore, we derive sharp bounds on the corresponding absolute effects. We apply these results to estimate the effects of the ChAdOx1 nCoV-19 vaccine on SARS-CoV-2 disease (COVID-19) conditional on postvaccine exposure to the virus, using data from a large RCT.

show abstract

Section: Discussionmentioning

confidence: 99%

Identification of Vaccine Effects When Exposure Status Is Unknown

Stensrud

Smith

2023

Epidemiology

View full text Add to dashboard Cite

show abstract

“…In comparison to Equation ( 6), we consider a general alternative policy rather than the only the bestpossible one. As we show below, the choice of this alternative policy will lead to different objectives and optimal solutions (see Cui, 2021, for a recent general discussion).…”

Section: Partial Identification and Minimizing Worst-case Regretmentioning

confidence: 99%

“…The maximum regret objective in Theorem 1 is a weighted average of the expected potential outcomes under treatment and no treatment plus a proxy for the cost. The choice of alternative policy determines these weights c 0 (•), c 1 (•) and cost c (•), all of which potentially vary with the covariates X (see Cui, 2021, for additional discussion on the choice of comparison policy determining the form of objective in instrumental variables settings). The second line of Equation ( 11) shows how to write the worst-case regret R sup (π, ) in terms of observable data using the scoring functions Γ w (either IPW or DR) discussed in Section 2.3.…”

Section: Worst Case Regret Relative To Different Alternative Policiesmentioning

confidence: 99%

“…Instead, these approaches use ideas from robust optimization (see Bertsimas et al, 2011, for a review) to minimize the worst case regret relative to a comparison policy. As Cui (2021) discusses, building on classical ideas in decision theory (Manski, 2005), the choice of comparison policy is key to determining the minimax optimal policy. We will explore this below in the setting where the utilities explicitly depend on the joint potential outcomes or principal strata.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Policy learning with asymmetric utilities

Ben‐Michael¹,

Imai²,

Jiang³

2022

Preprint

View full text Add to dashboard Cite

Data-driven decision making plays an important role even in high stakes settings like medicine and public policy. Learning optimal policies from observed data requires a careful formulation of the utility function whose expected value is maximized across a population. Although researchers typically use utilities that depend on observed outcomes alone, in many settings the decision maker's utility function is more properly characterized by the joint set of potential outcomes under all actions. For example, the Hippocratic principle to "do no harm" implies that the cost of causing death to a patient who would otherwise survive without treatment is greater than the cost of forgoing life-saving treatment. We consider optimal policy learning with asymmetric utility functions of this form. We show that asymmetric utilities lead to an unidentifiable social welfare function, and so we first partially identify it. Drawing on statistical decision theory, we then derive minimax decision rules by minimizing the maximum regret relative to alternative policies. We show that one can learn minimax decision rules from observed data by solving intermediate classification problems. We also establish that the finite sample regret of this procedure is bounded by the mis-classification rate of these intermediate classifiers.We apply this conceptual framework and methodology to the decision about whether or not to use right heart catheterization for patients with possible pulmonary hypertension.

show abstract

Statistical Reinforcement Learning and Dynamic Treatment Regimes

Shen,

Cui

2024

ICSA Book Series in Statistics

View full text Add to dashboard Cite

Individualized Decision Making Under Partial Identification: ThreePerspectives, Two Optimality Results, and One Paradox

Cited by 5 publications

References 40 publications

Identification of Vaccine Effects When Exposure Status Is Unknown

Identification of Vaccine Effects When Exposure Status Is Unknown

Policy learning with asymmetric utilities

Statistical Reinforcement Learning and Dynamic Treatment Regimes

Contact Info

Product

Resources

About