Transportability of Trial Results Using Inverse Odds of Sampling Weights

Westreich, Daniel; Edwards, Jessie K.; Lesko, Catherine R.; Stuart, Elizabeth A.; Cole, Stephen R.

doi:10.1093/aje/kwx164

Cited by 240 publications

(280 citation statements)

References 16 publications

Supporting

Mentioning

279

Contrasting

Unclassified

Order By: Relevance

“…Variables meeting this criterion were further assessed using interaction analyses with additional adjustment for potential confounders of their association with adenoma outcomes as well as using standardization methods (see below) to compare results across the two studies. To standardize the participant population in the CPPS or the VCPPS trial to the distribution of observed effect measure modifier in the other trial, we used inverse odds of sampling weights . Briefly, inverse odds of sampling weights use a form of model‐based standardization to transport the effects of an intervention in a given study population (in this case, the effects of calcium supplementation on adenoma risk in one of the two trials) to another target population (here, the other trial).…”

Section: Methodsmentioning

confidence: 99%

“…To standardize the participant population in the CPPS or the VCPPS trial to the distribution of observed effect measure modifier in the other trial, we used inverse odds of sampling weights. 11 Briefly, inverse odds of sampling weights use a form of model-based standardization 12 to transport the effects of an intervention in a given study population (in this case, the effects of calcium supplementation on adenoma risk in one of the two trials) to another target population (here, the other trial). These methods have been applied to a variety of randomized clinical trials when seeking to generalize or transport treatment effects to other target populations of interest.…”

Section: Statistical Analysesmentioning

confidence: 99%

See 1 more Smart Citation

Body mass index, calcium supplementation and risk of colorectal adenomas

Barry

Lund

Westreich

et al. 2018

Intl Journal of Cancer

Self Cite

View full text Add to dashboard Cite

Calcium supplementation (1,200 mg/day) did not significantly reduce colorectal adenomas in our recent randomized, controlled trial (Vitamin D/Calcium Polyp Prevention Study, VCPPS, 2004-2013) in contrast to our previous trial (Calcium Polyp Prevention Study, CPPS, 1988-1996). To reconcile these findings, we identified participant characteristics that differed between the study populations and modified the effect of calcium supplementation on adenomas or high-risk findings (advanced or multiple adenomas). Compared to the CPPS, more participants in the VCPPS were obese (body mass index (BMI) ≥30 kg/m ; 37.5% vs. 24.4%) and fewer had normal BMI (BMI <25 kg/m ; 18.5% vs. 31%). BMI appeared to modify the effect of calcium supplementation on adenomas and especially on high risk-findings: in the VCPPS, there was a 44% reduction in high-risk findings among individuals whose BMI was normal (RR = 0.56, 95% CI = 0.26-1.23), but not among overweight (RR = 1.09, 95% CI = 0.62-1.91) or obese (RR = 1.54, 95% CI = 0.92-2.57) individuals (p = 0.03). Similarly, in the CPPS, there was a 56% reduction in high-risk findings among individuals whose BMI was normal (RR = 0.44, 95% CI = 0.26-0.74), but not among overweight (RR = 0.87, 95% CI = 0.55-1.39) or obese (RR = 1.02, 95% CI = 0.57-1.82) individuals (p = 0.02). Standardization of each trial's findings to the BMI distribution in the other attenuated calcium's protective effect on adenomas in the CPPS but enhanced it in the VCPPS. In conclusion, 1,200 mg/day calcium supplementation may reduce risk of colorectal adenomas among those with normal BMI but not in overweight or obese individuals; and differences in BMI distribution partially account for the apparent difference in calcium efficacy between the two trials.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Statistical Analysesmentioning

confidence: 99%

Body mass index, calcium supplementation and risk of colorectal adenomas

Barry

Lund

Westreich

et al. 2018

Intl Journal of Cancer

Self Cite

View full text Add to dashboard Cite

show abstract

“…We use the term generalizability when the target population coincides or is a subset of the trial‐eligible population and transportability when the target population includes at least some individuals who are not trial‐eligible (and who, by definition, cannot be trial participants; others have proposed different definitions). In generalizability analyses, the target population often has a different distribution of effect modifiers compared with the participant population, even though both populations meet the trial eligibility criteria.…”

Section: Extending Trial Findings To a Target Populationmentioning

confidence: 99%

“…This tutorial builds on a growing literature to show that, instead of relying exclusively on informal assessments, we can use formal statistical methods to extend causal inferences from trial participants to a new target population. We focus on the common setting in which individual‐level data on time‐fixed treatments, outcomes, and baseline covariates are available from a randomized trial, but only individual‐level data on baseline covariates are available from the target population of nonparticipants.…”

Section: Introductionmentioning

confidence: 99%

Extending inferences from a randomized trial to a new target population

Dahabreh

Robertson

Steingrimsson

et al. 2020

Statistics in Medicine

146

196

View full text Add to dashboard Cite

When variables that are treatment effect modifiers also influence the decision to participate in a clinical trial, the average effect among trial participants will differ from the effect in other populations of trial-eligible individuals. In this tutorial, we consider methods for transporting inferences about a time-fixed treatment from trial participants to a new target population of trial-eligible individuals, using data from a completed randomized trial along with baseline covariate data from a sample of nonparticipants. We examine methods based on modeling the expectation of the outcome, the probability of participation, or both (doubly robust). We compare the finite-sample performance of different methods in a simulation study and provide example code to implement the methods in software. We illustrate the application of the methods to the Coronary Artery Surgery Study, a randomized trial nested within a cohort of trial-eligible patients to compare coronary artery surgery plus medical therapy versus medical therapy alone for patients with chronic coronary artery disease. Lastly, we discuss issues that arise when using the methods in applied transportability analyses.

show abstract

“…The proof of approach 5 is closely related to that for approach 4. Westreich et al [27] provide a full proof in the appendix.…”

Section: A Appendixmentioning

confidence: 99%

Effect heterogeneity and variable selection for standardizing causal effects to a target population

et al. 2019

View full text Add to dashboard Cite

Randomized trials, and even observational studies used for causal inference, are rarely representative of the populations in which the results will be used to inform clinical decision making. In order to account for differences between populations, we may consider standardizing results to a target population. In this paper, we discuss different approaches for reasoning about which covariates must be standardized over, and different approaches for computing the standardized causal effect in the target population. We compare and contrast these approaches, with an emphasis on evaluating whether the necessary assumptions underlying each method can be expected to hold in standard clinical and pharmacoepidemiological settings. We conclude that the choice must be informed by expert beliefs about which assumptions are most likely to be approximately true in the specific scientific context.

show abstract

Transportability of Trial Results Using Inverse Odds of Sampling Weights

Cited by 240 publications

References 16 publications

Body mass index, calcium supplementation and risk of colorectal adenomas

Body mass index, calcium supplementation and risk of colorectal adenomas

Extending inferences from a randomized trial to a new target population

Effect heterogeneity and variable selection for standardizing causal effects to a target population

Contact Info

Product

Resources

About