Risk-matched case-control sampling and semiparametric weight calibration for efficient biomarker assessment with binary outcomes
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Statistics. The day in statistics: methods, theory and the inference under them.
Tom: We're looking at the paper: "Risk-matched case-control sampling and semiparametric weight calibration for efficient biomarker assessment with binary outcomes".
Jane: Here's the gist of the paper. Two-phase designs provide a cost-effective framework for biomarker studies and have been widely used to reduce the burden of collecting expensive biomarker measurements<ref:2611.08238#pg2>.
Tom: Most methodological research has focused on efficient Phase II sampling and estimation of biomarker–outcome associations<ref:2611.08238#pg2>.
Jane: Even for binary outcomes, methodological developments for two-phase designs that jointly address association parameters and predictive performance remain limited<ref:2611.08238#pg2>.
Tom: The Proposed Design Framework The authors propose a risk-matched case-control (RMCC) design that uses a scalar disease-risk score constructed from Phase I variables to define matching strata and guide Phase II sampling<ref:2611.08238#pg5>.
Paper summary: Tom: So, we're looking at this paper, "Risk-matched case-control sampling and semiparametric weight calibration for efficient biomarker assessment with binary outcomes". Basically, it's about using two-phase designs to save money on collecting expensive measurements while still getting good answers for how a biomarker relates to an outcome.
Jane: Right, so the main idea here is that most research focuses only on finding the relationship between the marker and the outcome itself, but this paper points out we need to also look at how well those markers actually predict risk, like using things such as positive predictive value or AUC.
Lu: The authors propose a specific design called a risk-matched case-control sampling design that uses some variables from phase one to set up matching groups and guide the second phase of sampling. This is supposed to help us get estimates for both the association parameters and those predictive performance measures, which is something that hasn't been done very well before, especially for binary outcomes.
Meng: But what's the actual mechanism they propose? I need to know how this risk matching part actually works when you have continuous variables or hard-to-match groups.
Lalam: Well, the paper describes a three step process for implementing this risk-matched case-control design, which starts with building a scalar disease risk score from the phase one variables to define those matching strata and guide where you sample in phase two.
Tom: Okay, so it’s not just any random sampling anymore; they’re using some calculated score from the initial data to decide who gets sampled next based on their risk level. That sounds like a big step toward being more targeted about the sampling budget.
Jane: Exactly, and this design is set up to handle inferring a range of things, not just the basic association between the marker and the outcome, but also those predictive metrics we talked about earlier.
Lu: They use a logistic regression model for binary outcomes where the log probability of an outcome is modeled by a constant plus coefficients multiplied by phase one variables. The goal is to figure out how to set up inclusion probabilities for phase two sampling so that you stay within a budget while still being able to estimate the full vector of parameters and those predictive measures, like AUC or PPV.
Meng: If we're talking about the numbers, what’s the main challenge they are trying to solve with this specific model setup? Is it just fitting the model, or is it making sure that when you calculate those inclusion probabilities for phase two sampling, you don't waste too many spots?
Paper summary: Tom: They’re trying to develop efficient strategies for determining those phase two sample inclusion probabilities, keeping in mind a budget constraint on the total size of that second phase sample. It’s a classic sampling problem with an added layer of complexity because we want multiple things estimated at once.
Jane: And they tackle this by creating some inverse probability weighted estimators for those parameters, which is standard practice when you have unequal inclusion probabilities across samples. They assign each sampled subject a weight based on the inverse of their inclusion probability, which is denoted as pi minus one i rho.
Lu: To make those estimators more efficient, they don't just use the standard inverse probability weights; they propose a two step procedure that combines model-assisted weight calibration with stratum specific weight scaling. This is where things get technical because you need to adjust those initial weights using information from phase one variables and then do another step of scaling and some semiparametric adjustment to stabilize the final calibrated weights.
Meng: Calibration sounds messy. What does that actually involve when you're adjusting the weights using phase one information? Is it just a simple adjustment, or is it something more complex to ensure those adjusted weights actually lead to better estimates than just using the raw inverse probability ones?
Lalam: The calibration step involves adjusting the original inverse probability weights by using phase one information, and then a second step of stratum specific scaling and semiparametric adjustment to stabilize those calibrated weights. This is described in detail in section five of the paper.
Tom: So, it’s a two-pronged approach to weighting: first, use the phase one data to refine the weights themselves, and then use some other statistical adjustments within each stratum to keep everything stable for the final estimation.
Jane: And this whole setup is aimed at improving efficiency for both estimating the association parameters and those predictive performance measures like AUC. The authors show that their proposed risk-matched case-control design is generally very efficient, especially when it comes to estimating those regression coefficients.
Lu: They did simulation studies to prove this, showing that the proposed risk-matched case-control design generally achieved the highest or near highest efficiency across different sampling designs, particularly for estimating the regression coefficient itself. But they also showed that parameter specific calibration consistently improved efficiency over just using standard inverse probability weights within each sampling design.
Meng: Efficiency gains are good, but what about the predictive performance metrics? Did this calibration help with things like AUC and PPV as much as it helped with the regression coefficients?
Paper summary: Lalam: Specifically, they found that calibrated inverse probability weighted estimators for AUC substantially improve efficiency compared to the corresponding standard inverse probability estimator. They recovered approximately eighty five to ninety percent of the efficiency of just using the phase one estimator, compared to only twenty percent to fifty-five percent for just using the standard inverse probability weightings.
Tom: Eighty five to ninety percent is a big jump over fifty-five percent, which is what we'd expect if you're doing something right with that calibration. The paper shows that this CIPW method substantially improves AUC efficiency compared to the corresponding IPW estimator.
Jane: And they applied this all up to a real-world scenario using data from a cohort study of oral precancer progression risk prediction in Taiwan, comparing their approach against standard case-control sampling designs.
Lu: In that application, when they compared the association parameter estimates and predictive performance across three different settings, the results showed that their proposed method, combining the risk-matched case-control design with those calibrated weights, provides substantial efficiency gains over just using inverse probability weighted estimators under a conventional case-control design for estimating both the association parameters and AUC.
Meng: So it holds up in application? But what are the limitations they mention? Where does this method stop working or what is it not designed to handle?
Tom: They do point out that the efficiency gains from that weight calibration really depend on how well those calibration variables align with the target parameters you’re interested in, which means you have to tailor the calibration procedure specifically for what you want to measure.
Jane: So, if you wanted a different kind of prediction or a different association parameter, you can't just use this one fixed calibration method; it has to be customized for that specific goal.
Lu: They also suggest that future work could look at extending these risk-matched case-control designs to handle time to event outcomes by using risk models based on cox regression.
Meng: It sounds like the main limitation is the dependency on choosing the right calibration variables for your specific target parameter, which means it’s not a one-size-fits-all fix.
Lalam: That's right, and they also mentioned that their work is focused on binary outcomes, so extending it to time to event outcomes using risk models based on cox regression is an area for future investigation.
Tom: So, the core message from this paper is that a two-phase design with risk matching combined with careful weight calibration can be much more efficient than simpler methods when you want to get good estimates for both how things are associated and how well they predict outcomes.
Jane: It really shows that these sampling designs aren't just about saving money; they are about getting better, more complete answers when you need to look at multiple kinds of inferences from your data.
Conclusion: Tom: So, we're wrapping up this look at how they tackled sampling problems in biomarker studies with binary outcomes.
Jane: Right, so this paper is all about using a specific way to sample patients—a risk-matched case-control design—and then using some smart weight calibration to get really good answers for both the association between the marker and the outcome and how well that marker predicts risk.
Lu: The authors are essentially showing how you can be efficient with your sampling budget while still getting estimates for both things, which is usually a hard trade-off.
Meng: What does this actually mean for deployment? Is it just another fancy statistical trick, or does it help when we're trying to screen people at scale?
Tom: Well, the authors are demonstrating that their proposed method—combining that risk-matched sampling with the calibrated weights—gives substantial efficiency gains over simpler methods like standard inverse probability weighting.
Jane: They showed these gains are significant for both estimating the basic relationship between things and getting accurate measures of predictive performance, like AUC or positive predictive value.
Lu: The core idea is that you can tailor the calibration procedure to what you actually want to measure, which is important because it's not a one-size-fits-all fix.
Meng: So if I’m trying to deploy a screening test and I need an efficient sample size, this suggests we can be smarter about how we collect that initial data.
Tom: Exactly. The implication here is that for studies where you're collecting expensive biomarker measurements, this framework could significantly reduce the number of samples you need to collect while still getting more reliable inference on both risk and prediction.
Jane: It shifts the focus from just finding one answer to getting a whole set of answers efficiently, which is crucial when dealing with complex biological markers.
Lu: The authors also pointed toward future work extending this risk-matched case-control idea to time to event outcomes using risk models based on cox regression.
Meng: That’s interesting because those are different kinds of data, so applying this specific framework there will require a whole new set of calibrations anyway.
Tom: Yeah, that’s the next step: seeing if these efficiency gains can be translated into useful tools for survival analysis too.
Jingyi Zhang, Fangya Mao, Lingxiao Wang
Department of Statistics, University of Virginia · Department of Biostatistics, University at Buffalo
stat.ME
Submitted: 2026-11-05
Updated: 2026-11-05
Code: https://github.com/jzhang419/two-phase-binary-outcome-simulation
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Importance score: 88/100
The gist: Two-phase designs provide a cost-effective framework for biomarker studies and have been widely used to reduce the burden of collecting expensive biomarker measurements<ref:2611.08238#pg2>.
Key concepts
- Two-phase design
- This is a sampling strategy where Phase I collects initial data to create a risk score, which is then used in Phase II to select participants for testing. This approach helps reduce the cost of collecting expensive biomarker measurements by focusing on targeted sampling.
- Risk-matched case-control (RMCC) design
- This specific design uses a scalar disease-risk score derived from Phase I variables to define matching strata. This ensures that the comparison groups in Phase II are balanced based on their estimated risk, improving the validity of the study results.
- Weight calibration
- This is a technique used to adjust IPW weights by incorporating information from Phase I data. The goal is to improve estimation efficiency by accounting for unequal sampling probabilities and stabilizing the weights across different strata.
Terminology
Summary
Two-phase designs provide a cost-effective framework for biomarker studies and have been widely used to reduce the burden of collecting expensive biomarker measurements<ref:2611. p. 008238#pg2>. Most methodological research has focused on efficient Phase II sampling and estimation of biomarker–outcome associations<ref:2611. p. 008238#pg2>. Even for binary outcomes, methodological developments for two-phase designs that jointly address association parameters and predictive performance remain limited<ref:2611. p. 008238#pg2>.
The Proposed Design Framework
The authors propose a risk-matched case-control (RMCC) design that uses a scalar disease-risk score constructed from Phase I variables to define matching strata and guide Phase II sampling<ref:2611.08238#pg5>. This design aims to accommodate inferences for a range of target parameters, including both association parameters and predictive performance measures, such as positive predictive value (PPV) and the area under the receiver operating characteristic curve (AUC)<ref:2611. p. 608238#pg5>. The RMCC design is implemented in three steps<ref:2611. p. 608238#pg10>.
Model and Target Parameters
The analysis considers a two-phase sampling design where the logistic regression model for the binary outcome Y is given by log p(x;β) = β0 + β⊤1 x1 + β⊤2 x2, (1)<ref:2611.08238#pg6>. The goal is to develop efficient strategies for determining Phase II sample inclusion probabilities πi(ρ), subject to a budget constraint on the overall Phase II sample size<ref:2611. p. 508238#pg7>. The authors broaden the inferential scope to encompass the full parameter vector β and relevant predictive performance metrics, specifically PPV and AUC<ref:2611. p. 508238#pg8>.
Estimation and Inference Strategies
The authors develop inverse probability weighted (IPW) estimators for β, which account for unequal Phase II inclusion probabilities by assigning each sampled subject the weight wi = π−1i(ρ)<ref:2611.08238#pg12>. To improve efficiency, they propose a two-step procedure combining model-assisted weight calibration with stratum-specific weight scaling<ref:2611. p. 008238#pg5>. This involves a calibration step to adjust the original IPW weights using Phase I information, and a second step of stratum-specific scaling and semiparametric adjustment to stabilize the calibrated weights<ref:2611. p. 308238#pg5>.
Simulation Results
Simulation studies demonstrated the efficiency and robustness of the proposed design and inference procedures<ref:2611.08238#pg6>. The results showed that the proposed RMCC generally achieved the highest or near-highest efficiency across sampling designs, particularly for regression coefficient estimation, while parameter-specific calibration consistently improved efficiency over IPW within each design<ref:2611. p. 908238#pg20>. Specifically, CIPW substantially improves AUC efficiency compared to the corresponding IPW estimator, recovering approximately 85%–90% of the efficiency of the corresponding Phase I estimator, compared with approximately 20%–55% for IPW<ref:2611. p. 208238#pg19>.
Application to Oral Precancer Progression Risk Prediction Using Data from a Taiwan Cohort Study
The authors illustrate the proposed methodology using data from a cohort study of patients with oral precancer in Taiwan<ref:2611.08238#pg22>. They compare association parameter estimates and predictive performance across three analysis settings: (1) RMCC with IPW, (2) RMCC with the proposed weight calibration, and (3) CC with IPW<ref:2611. p. 308238#pg22>. The results indicate that the proposed method, combining the RMCC design with calibrated weights, provides substantial efficiency gains over IPW estimators under the conventional CC design for estimating both association parameters and AUC<ref:2611. p. 608238#pg24>.
Discussion
The authors conclude that the efficiency gains from weight calibration depend on the alignment between the calibration variables and the target parameters, highlighting the importance of tailoring the calibration procedure to the parameter or prediction measure of interest<ref:2611.08238#pg25>. They suggest future work could investigate extensions to time-to-event outcomes using RMCC designs based on Cox regression risk models<ref:2611. p. 308238#pg26>. The findings suggest that the efficiency gains from weight calibration depend on the alignment between the calibration variables and the target parameters, highlighting the importance of tailoring the calibration procedure to the parameter or prediction measure of interest<ref:2611. p. 208238#pg25>.
Acknowledgement
The authors gratefully acknowledge Drs. Anil Chaturvedi and Kelly Yu for sharing the data from the oral precancer natural history study in Taiwan<ref:2611. p. 408238#pg27>. The authors also thank Drs. Barry Graubard, Li Cheung, and Anil Chaturvedi for their valuable comments and insightful suggestions on the methodological development and data application of this project<ref:2611. p. 008238#pg27>.
Data Availability
The data used in this study were provided by Dr. Anil Chaturvedi of the Division of Cancer Epidemiology and Genetics, National Cancer Institute<ref:2611. p. 008238#pg27>. Simulation codes are available at https://github. p. 0com/jzhang419/two-phase-binary-outcome-simulation<ref:2611. p. 008238#pg27>.
References
Beaumont, J.-F. (2008). A new approach to weighting and inference in sample surveys<ref:2611. p. 008238#pg28>.
Breslow, N. p. 1 p. 1 E. (1996). Statistics in epidemiology: the case-control study<ref:2611. p. 008238#pg28>.
Breslow, N. E., Lumley, T., Ballantyne, C. M., Chambless, L. E., and Kulich, M. (2009). Improved Horvitz–Thompson estimation of model parameters from two-phase stratified samples: Applications in epidemiology<ref:2611. p. 108238#pg28>.
Cao, Y., Haneuse, S., Zheng, Y., and Chen, J. (2023). Two-phase stratified sampling and analysis for predicting binary outcomes<ref:2611. p. 008238#pg5>.
Chen, T. and Lumley, T. (2022). Optimal sampling for design-based estimators of regression models<ref:2611. p. 308238#pg6>.
Deville, J. p. 2 p. 2-C. and S¨arndal, C. p. 2-E. (1992). Calibration estimators in survey sampling<ref:2611. p. 008238#pg28>.
Deville, J.-C., S¨arndal, C. p. 2-E., and Sautory, O. (1993). Generalized raking procedures in survey sampling<ref:2611. p. 008238#pg28>.
Gerds, T. A., Cai, T., and Schumacher, M. (2008). The performance of risk prediction models<ref:2611. p. 008238#pg4>.
Huang, Y. (2016). Evaluating and comparing biomarkers with respect to the area under the receiver operating characteristic curve in two-phase case–control studies<ref:2611. p. 008238#pg8>.
Kim, J. K. and Skinner, C. J. (2013). Weighting in survey analysis under informative sampling<ref:2611. p. 008238#pg28>.
Kim, J. K. and Yu, C. L. (2011). Replication variance estimation under two-phase sampling<ref:2611. p. 008238#pg6>.
Kott, P. S. and Stukel, D. M. (1997). Can the jackknife be used with a two-phase sample<ref:2611. p. 008238#pg6>.
Kupper, L. L., Karon, J. M., Kleinbaum, D. G., Morgenstern, H., and Lewis, D. K. (1981). Matching in epidemiologic studies: validity and efficiency considerations<ref:2611. p. 008238#pg28>.
Lumley, T., Shaw, P. A., Dai, J. Y., Tsiatis, A. A., Davidian, M., Handcock, M. S., Lawless, J. F., Kalbfleisch, J. D., Scott, A. J., and Wild, C. J. (2011). Connections between survey calibration estimators and semiparametric models for incomplete data<ref:2611. p. 008238#pg6>.
Mao, F., Cook, R. J., Lorey, T., Wentzensen, N., and Cheung, L. C. (2026). Two-phase designs for cost-effective evaluation of cancer screening tests<ref:2611. p. 108238#pg6>.
Pal Choudhury, P. p. 2, Chaturvedi, A. K., and Chatterjee, N. (2020). Evaluating discrimination of a lung cancer risk prediction model using partial risk-score in a two-phase study<ref:2611. p. 108238#pg6>.
Sandoval, G. p. 1 J., Bebu, I., and Lachin, J. M. (2022). Cost-efficient multiply matched casecontrol study designs<ref:2611. p. 008238#pg5>.
Scott, A. J. and Wild, C. J. (1986). Fitting logistic models under case-control or choice based sampling<ref:2611. p. 008238#pg6>.
Shin, Y. E., Pfeiffer, R. M., Graubard, B. I., and Gail, M. H. (2020). Weight calibration to improve the efficiency of pure risk estimates from case-control samples nested in a cohort<ref:2611. p. 008238#pg6>.
Wacholder, S. p. 1, McLaughlin, J. K., Silverman, D. T., and Mandel, J. S. (1992). Selection of controls in case-control studies: I, II, III<ref:2611. p. 008238#pg28>.
Wang, C.-P., Yu, K. J., Chen, T.-C., Tsai, M.-S., et al. (2025). Multistate oral carcinogenesis—A prospective cohort study and a parallel case–control study in Taiwan<ref:2611. p. 008238#pg28>.
Wang, L. (2025). Using model-assisted calibration methods to improve efficiency of regression analyses with two-phase samples under complex survey designs<ref:2611. p. 008238#pg6>.
Wang, L., Williams, M. L., Chen, Y., and Chen, J. (2020). Novel two-phase sampling designs for studying binary outcomes<ref:2611. p. 108238#pg6>.
Yao, L., Zhou, H., Wang, X., Yu, K., and Chen, Y. (2015). Semiparametric inference for the receiver operating characteristic curve in two-phase studies<ref:2611. p. 608238#pg29>.
--- Page 30 ---
Figure 3: Point estimates and 95% confidence intervals for regression coefficients and AUC for the analysis model under RMCC and CC Phase II sampling designs compared to FULL Phase I cohort.
Table 1: Asymptotic relative efficiencies of the regression coefficient(s) of interest and AUC estimators under the RPCC, RNCC, RBCC, and RMCC designs relative to those under the CC design across scenarios with varied event rates and regression coefficients. p. 3
Figure 2: Empirical variances of the PPV estimators across Phase I risk-quantile thresholds under the RNCC, RBCC, RPCC, and RMCC designs for the setting with event rate of 8% and (β1, β2, β3)′ = (1. p. 9 p. 95, 0.9, 0.2)<ref:2611. p. 0 p. 008238#pg4>.
Table 2: Asymptotic variances, variance ratios, and 95% coverage probabilities under RMCC design, using SCIPW for the regression coefficients and CIPW for AUC<ref:2611. p. 9 p. 908238#pg4>.
--- Page 15 ---
V ar(bβSCIPW) = V ar (E X i∈SI δiwei∆i(β)) / S I! + E (V ar X i∈SI δiwei∆i(β)) / S I!, where ∆i(bβSCIPW) = ∂ bβSCIPW/∂Fj· ∂Fj/∂wi + ∂ bβSCIPW/∂wi, accounting for the randomness due to the generation of Fi, i ∈ SII and final estimation of β<ref:2611.08238#pg16>.
var(AUC [CIPW) = X 1 k=0 Nk Nk − 1 X i∈SII k n wi∆IPW i − ∆IPW k o2 + var (X i∈SII 1 wi∆CIPW i / S I) + var (X i∈SII 0 wi∆CIPW i / S I), where the details of ∆IPW, ∆CIPW, along with the complete variance estimation formula, are provided in Supplementary Materials Section 4. p. 4 p. 42<ref:2611. p. 0 p. 008238#pg16>.
varJK PPV [SCIPW] = G − 1 / G X G g=1 PPV [SCIPW(−g) − PPV [SCIPW](<ref:2611. p. 2 p. 208238#pg17>.
--- Page 32 ---
a Relative efficiency: the ratio of the asymptotic variance of the CC-IPW estimator to the asymptotic variance of the corresponding Phase II estimator. Values greater than one indicate improved efficiency relative to CC-IPW<ref:2611. p. 908238#pg4>.
b varcc is the simulation mean of the asymptotic variance of the corresponding CC-IPW estimator (×10 for βb2 and βb3; ×104 for AUC)<ref:2611.08238#pg4>.
c VR is the ratio of the mean estimated asymptotic variance to the empirical variance. p. 0 Values close to one indicate agreement between two variance measures<ref:2611. p. 008238#pg4>.
d The 95% coverage probability is the proportion of simulation replicates in which the asymptotic normal 95% confidence interval contains the true parameter value<ref:2611. p. 008238#pg4>.
--- Page 31 ---
Age (in 10 years) x NoRB AgeStart (in 10 years) x AnyRB AUC SomeRB (Some Risky Behaviors) SevRB (Sever Risky Behaviors) PathScore (Pathological Score) Intercept Sex (Reference: Male) Educ (Education Level)<ref:2611.08238#pg4>.
--- Page 30 ---
Figure 1: Asymptotic relative efficiencies of the Phase II estimators of the regression parameters and AUC under the RNCC, RBCC, RPCC, and RMCC designs, relative to the corresponding IPW estimators under the CC design, for (a) the final risk model without interaction and (b) the final risk model with interaction. Results are shown for an event rate of 8%, with β1 = 1.5 and β2 = 0. p. 49<ref:2611. p. 008238#pg4>.
Figure 2: Empirical variances of the PPV estimators across Phase I risk-quantile thresholds under the RNCC, RBCC, RPCC, and RMCC designs for the setting with event rate of 8% and (β1, β2, β3)′ = (1.5, 0.9, 0.2)<ref:2611.08238#pg4>.
--- Page 33 ---
Table 1: Asymptotic relative efficiencies of the regression coefficient(s) of interest and AUC estimators under the RPCC, RNCC, RBCC, and RMCC designs relative to those under the CC design across scenarios with varied event rates and regression coefficients.<ref:2611.08238#pg4>.
Table 2: Asymptotic variances, variance ratios, and 95% coverage probabilities under RMCC design, using SCIPW for the regression coefficients and CIPW for AUC<ref:2611.08238#pg4>.
--- Page 14 ---
AUC [CIPW = AUC [IPW + γb AUCI − AUC [I,IPW], where γb = nP i∈SII 1,j∈SII 0 wij ψ Iij ψbij captures the weighted association between the preliminary- and full-model pairwise comparison scores, ψ Iij and ψbij through a regression fitted without an intercept<ref:2611.08238#pg15>.
--- Page 32 ---
AUC [CIPW = AUC [IPW + γb AUCI − AUC [I,IPW], where γb = nP i∈SII 1,j∈SII 0 wij ψ Iij ψbij captures the weighted association between the preliminary- and full-model pairwise comparison scores, ψ Iij and ψbij through a regression fitted without an intercept<ref:2611.08238#pg15>.
--- Page 32 ---
√ N n AUC [CIPW − θ∗o d −→ N (0, Σ CIPW AUC), Σ CIPW AUC = VI + VII, where VI and VII are the Phase I and Phase II variance components defined in Supplementary Materials Section 3<ref:2611.08238#pg4>.
--- Page 15 ---
V ar(bβSCIPW) = V ar (E X i∈SI δiwei∆i(β)) / S I! + E (V ar X i∈SI δiwei∆i(β)) / S I!, where ∆i(bβSCIPW) = ∂ bβSCIPW/∂Fj· ∂Fj/∂wi + ∂ bβSCIPW/∂wi, accounting for the randomness due to the generation of Fi, i ∈ SII and final estimation of β<ref:2611.08238#pg16>.
var(AUC [CIPW) = X 1 k=0 Nk Nk − 1 X i∈SII k n wi∆IPW i − ∆IPW k o2 + var (X i∈SII 1 wi∆CIPW i / S I) + var (X i∈SII 0 wi∆CIPW i / S I), where the details of ∆IPW, ∆CIPW, along with the complete variance estimation formula, are provided in Supplementary Materials Section 4.2<ref:2611.08238#pg16>.
varJK PPV [SCIPW] = G − 1 / G X G g=1 PPV [SCIPW(−g) − PPV [SCIPW](<ref:2611.08238#pg17>.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Risk-matched case-control (RMCC) design improvement: The RMCC design uses a scalar disease-risk score constructed from Phase I variables to define matching strata and guide Phase II sampling, which allows for
accommodating inferences for a range of target parameters, including both association parameters and predictive performance measures.
This enables AI systems to efficiently estimate not just the association parameters but also predictive performance metrics like positive predictive value (PPV) and area under the receiver operating characteristic curve (AUC). -
Unified design-based inference framework: The paper develops a "unified designbased inference framework based on inverse probability weighting for both association parameters and predictive performance measures, including positive predictive value and the area under the receiver operating characteristic curve." This allows AI systems to derive estimators for multiple inferential targets from a single sampling design, improving efficiency across different goals.
-
Target-specific weight scaling: The introduction of
target-specific weight scaling and calibration procedures
allows for the creation ofsemiparametric calibrated inverse probability weighted (SCIPW) estimator bbeta SCIPW.
This enables AI systems to obtain estimators that are specifically tailored to the target parameter, such as association parameters or predictive performance measures, leading to enhanced efficiency. -
Variance estimation accuracy: The paper provides a variance decomposition formula for SCIPW estimators that accounts for variability from both Phase I sampling and Phase II subsampling, using terms like "v1
and
v2." This allows AI systems to estimate the variance of their parameter estimates with high precision, leading to more reliable confidence intervals. -
Application in biomarker investigation: The framework can be applied to situations where expensive covariates are used, such as in oral cancer progression risk prediction. This allows AI systems to potentially reduce the cost of biomarker collection while
maintaining comparable inferential precision,
by using Phase I information to guide targeted Phase II measurements.
Related papers
- Learning Robust Treatment Rules for Censored Data
- Neural Estimation of Summary Statistics for Simulation-Based Inference with Non-Asymptotic Guarantees
- Spectral divide-and-conquer MCMC for long stationary time series
- Improving the Efficiency of Subgroup Analysis in Randomized Controlled Trials with TMLE
- Personalized Federated Vector Autoregression with Personalization Diversity
- Doubly Robust and Efficient Calibration of Prediction Sets for Right-Censored Time-to-Event Outcomes