arXiv subjects

Gianluca Baio

Publications and source records attributed to Gianluca Baio.

At least 19 recordsLinked to original sources

arXiv2026

A Bayesian Longitudinal Model for Imputing Item-Level Missing Data in Trial-Based Economic Evaluations

Trial-based economic evaluations are widely used to assess the cost-effectiveness of healthcare interventions and inform decision-making. Cost and effectiveness outcomes are typically collected using multi-item questionnaires administered at multiple time points, and are often subject to item-level missingness. In principle, imputation (i.e., replacing missing value with estimated or substituted values) should be performed at the item level to fully exploit available information. However, this is rarely implemented in practice due to several statistical challenges, including the longitudinal data structure, cross-item dependence, heterogeneous missingness patterns, and the mixture of skewed cost and count data. In this paper, we develop a Bayesian longitudinal model for imputing item-level missing data in trial-based economic evaluations that accommodates these complexities within a unified framework. The approach combines a transition-model formulation for longitudinal dependence with flexible distributional assumptions and explicit modelling of cross-item relationships, allowing item-level responses of different types to be coherently modelled over time. Motivated by a real-world trial, we demonstrate the flexibility and practical applicability of the proposed approach. We further discuss how the model can be extended to settings where data may be missing not at random.

stat.ME↗

arXiv2026

Calculating the Expected Value of Sample Information accounting for missing data

The Expected Value of Sample Information (EVSI) is a powerful instrument to determine the value of additional evidence to inform an economic model. However, EVSI has been applied only to idealized data collection mechanisms, thereby reducing its potential applications in realistic studies. In this paper, we define a methodology to calculate EVSI when the additional evidence we aim to collect exhibits missing data; a very common challenge in real-world studies. First, we define how to simulate individual-level data and how to induce missingness inside the simulated data. We will reproduce Missing Completely At Random (MCAR), Missing At Random (MAR), and Missing Not At Random (MNAR) missing data mechanisms. Then, we will apply the multiple imputation method to adjust for the bias related to the missing data. Finally, we use the imputed data to compute the EVSI using nonparametric regression methods. We apply the novel methodology to two different health economic models and compare the EVSI computed on data without missingness with the EVSI with different types of missing data. We show that the EVSI decreases when the additional evidence suffers from missingness, and therefore, we define a method to efficiently compute the sample size we need to collect to recover the idealized EVSI (without missingness). We find out that the number of additional samples needed to correct for missing data exceeds that coming from standard approaches. With this methodology, we compute EVSI when the additional data are affected by non-trivial forms of missingness, modeling both the MAR and MNAR mechanisms, and extending EVSI calculation to more realistic scenarios. The fact that the necessary sample size to recover the idealized EVSI exceeds that of standard methods suggests that this methodology potentially represents a novel technique for sample size calculation in realistic studies.

stat.ME↗

arXiv2026

A Bayesian Approach for Nonignorable Dropout in Bivariate Longitudinal Models

Longitudinal data collected in clinical trials are almost always incomplete due to some of the participants dropping out from the study during the planned follow-up. A common strategy to handle nonresponse expresses missingness in terms of a dropout process, which is jointly analysed with the outcome process to facilitate the formulation of the missingness assumptions. However, when the outcome is multivariate, the identification of the dropout process becomes problematic, especially when individuals have different dropout times for each type of response, and sensitivity analysis is difficult. The modelling task may be also be complicated by data complexities (e.g. skewness and spikes) which are difficult to capture through standard parametric methods. An example of this analysis framework occurs in trial-based economic evaluations, where a longitudinal bivariate response, formed by suitably-defined measures of effectiveness and costs, is analysed to inform policymakers about the cost-effectiveness of alternative interventions. We present a novel Bayesian nonparametric approach to handle a missing bivariate longitudinal outcome by jointly modelling the dropout process associated with each type of response while also taking into account the complexities of the data. We specify a flexible nonparametric model for the observed data and partially identify the distribution of the missing data with identifying restrictions conditional on the dropout indicators and sensitivity parameters. We explore alternative nonignorable scenarios through different priors for the sensitivity parameters. Our approach is motivated by, and applied to, data from a trial assessing the cost-effectiveness of a new treatment for intellectual disability and challenging behaviour.

stat.ME↗

arXiv2025

Marginal and conditional summary measures: transportability and compatibility across studies

Marginal and conditional summary measures do not generally coincide, have different interpretations and correspond to different decision questions. While these aspects have primarily been recognized for non-collapsible summary measures, they are equally problematic for some collapsible measures in the presence of effect modification. We clarify the interpretation and properties of several marginal and conditional summary measures, considering different types of outcomes and hypothetical outcome-generating mechanisms. We describe implications of the choice of summary measure for transportability, highlighting that covariates not conventionally described as effect modifiers can modify population-level treatment effects. Finally, we illustrate existing summary measure incompatibility issues in the context of evidence synthesis, using the case of covariate adjustment methods for indirect treatment comparisons. Because marginal and conditional summary measures do not generally coincide, their naïve pooling in evidence synthesis can produce bias. Almost invariably, care is needed to ensure that evidence synthesis methods are combining compatible summary measures, and this may be easier to accomplish with full access to individual patient data.

stat.ME↗

arXiv2025

Diffusion piecewise exponential models for survival extrapolation using Piecewise Deterministic Monte Carlo

The piecewise exponential model is a flexible non-parametric approach for time-to-event data, but extrapolation beyond final observation times typically relies on random walk priors and deterministic knot locations, resulting in unrealistic long-term hazards. We introduce the diffusion piecewise exponential model, a prior framework consisting of a discretised diffusion for the hazard, that can encode a wide variety of information about the long-term behaviour of the hazard, time changed by a Poisson process prior for knot locations. This allows the behaviour of the hazard in the observation period to be combined with prior information to inform extrapolations. Efficient posterior sampling is achieved using Piecewise Deterministic Markov Processes, whereby we extend existing approaches using sticky dynamics from sampling spike-and-slab distributions to more general transdimensional posteriors. We focus on applications in Health Technology Assessment, where the need to compute mean survival requires hazard functions to be extrapolated beyond the observation period, showcasing performance on datasets for Colon cancer and Leukaemia patients.

stat.ME↗

arXiv2025

A Bayesian hierarchical mixture cure modelling framework to utilize multiple survival datasets for long-term survivorship estimates: A case study from previously untreated metastatic melanoma

Time to an event of interest over a lifetime is a central measure of the clinical benefit of an intervention used in a health technology assessment (HTA). Within the same trial, multiple end-points may also be considered. For example, overall and progression-free survival time for different drugs in oncology studies. A common challenge is when an intervention is only effective for some proportion of the population who are not clinically identifiable. Therefore, latent group membership as well as separate survival models for identified groups need to be estimated. However, follow-up in trials may be relatively short leading to substantial censoring. We present a general Bayesian hierarchical framework that can handle this complexity by exploiting the similarity of cure fractions between end-points; accounting for the correlation between them and improving the extrapolation beyond the observed data. Assuming exchangeability between cure fractions facilitates the borrowing of information between end-points. We undertake a comprehensive simulation study to evaluate the model performance under different scenarios. We also show the benefits of using our approach with a motivating example, the CheckMate 067 phase 3 trial consisting of patients with metastatic melanoma treated with first line therapy.

stat.AP↗

arXiv2024

Effect modification and non-collapsibility leads to conflicting treatment decisions: a review of marginal and conditional estimands and recommendations for decision-making

Effect modification occurs when a covariate alters the relative effectiveness of treatment compared to control. It is widely understood that, when effect modification is present, treatment recommendations may vary by population and by subgroups within the population. Population-adjustment methods are increasingly used to adjust for differences in effect modifiers between study populations and to produce population-adjusted estimates in a relevant target population for decision-making. It is also widely understood that marginal and conditional estimands for non-collapsible effect measures, such as odds ratios or hazard ratios, do not in general coincide even without effect modification. However, the consequences of both non-collapsibility and effect modification together are little-discussed in the literature. In this paper, we set out the definitions of conditional and marginal estimands, illustrate their properties when effect modification is present, and discuss the implications for decision-making. In particular, we show that effect modification can result in conflicting treatment rankings between conditional and marginal estimates. This is because conditional and marginal estimands correspond to different decision questions that are no longer aligned when effect modification is present. For time-to-event outcomes, the presence of covariates implies that marginal hazard ratios are time-varying, and effect modification can cause marginal hazard curves to cross. We conclude with practical recommendations for decision-making in the presence of effect modification, based on pragmatic comparisons of both conditional and marginal estimates in the decision target population. Currently, multilevel network meta-regression is the only population-adjustment method capable of producing both conditional and marginal estimates, in any decision target population.

stat.ME↗

arXiv2024

Averaging polyhazard models using Piecewise deterministic Monte Carlo with applications to data with long-term survivors

Polyhazard models are a class of flexible parametric models for modelling survival over extended time horizons. Their additive hazard structure allows for flexible, non-proportional hazards whose characteristics can change over time while retaining a parametric form, which allows for survival to be extrapolated beyond the observation period of a study. Significant user input is required, however, in selecting the number of latent hazards to model, their distributions and the choice of which variables to associate with each hazard. The resulting set of models is too large to explore manually, limiting their practical usefulness. Motivated by applications to stroke survivor and kidney transplant patient survival times we extend the standard polyhazard model through a prior structure allowing for joint inference of parameters and structural quantities, and develop a sampling scheme that utilises state-of-the-art Piecewise Deterministic Markov Processes to sample from the resulting transdimensional posterior with minimal user tuning.

stat.ME↗

arXiv2024

Model-based standardization using multiple imputation

When studying the association between treatment and a clinical outcome, a parametric multivariable model of the conditional outcome expectation is often used to adjust for covariates. The treatment coefficient of the outcome model targets a conditional treatment effect. Model-based standardization is typically applied to average the model predictions over the target covariate distribution, and generate a covariate-adjusted estimate of the marginal treatment effect. The standard approach to model-based standardization involves maximum-likelihood estimation and use of the non-parametric bootstrap. We introduce a novel, general-purpose, model-based standardization method based on multiple imputation that is easily applicable when the outcome model is a generalized linear model. We term our proposed approach multiple imputation marginalization (MIM). MIM consists of two main stages: the generation of synthetic datasets and their analysis. MIM accommodates a Bayesian statistical framework, which naturally allows for the principled propagation of uncertainty, integrates the analysis into a probabilistic framework, and allows for the incorporation of prior evidence. We conduct a simulation study to benchmark the finite-sample performance of MIM in conjunction with a parametric outcome model. The simulations provide proof-of-principle in scenarios with binary outcomes, continuous-valued covariates, a logistic outcome model and the marginal log odds ratio as the target effect measure. When parametric modeling assumptions hold, MIM yields unbiased estimation in the target covariate distribution, valid coverage rates, and similar precision and efficiency than the standard approach to model-based standardization.

stat.ME↗

arXiv2023

Bayesian Interrupted Time Series for evaluating policy change on mental well-being: an application to England's welfare reform

Factors contributing to social inequalities are also associated with negative mental health outcomes leading to disparities in mental well-being. We propose a Bayesian hierarchical model which can evaluate the impact of policies on population well-being, accounting for spatial/temporal dependencies. Building on an interrupted time series framework, our approach can evaluate how different profiles of individuals are affected in different ways, whilst accounting for their uncertainty. We apply the framework to assess the impact of the United Kingdoms welfare reform, which took place throughout the 2010s, on mental well-being using data from the UK Household Longitudinal Study. The additional depth of knowledge is essential for effective evaluation of current policy and implementation of future policy.

stat.ME↗

arXiv2023

Methodological considerations for novel approaches to covariate-adjusted indirect treatment comparisons

We examine four important considerations in the development of covariate adjustment methodologies for indirect treatment comparisons. Firstly, we consider potential advantages of weighting versus outcome modeling, placing focus on bias-robustness. Secondly, we outline why model-based extrapolation may be required and useful, in the specific context of indirect treatment comparisons with limited overlap. Thirdly, we describe challenges for covariate adjustment based on data-adaptive outcome modeling. Finally, we offer further perspectives on the promise of doubly-robust covariate adjustment frameworks.

stat.ME↗

arXiv2023

A framework for estimating and visualising excess mortality during the COVID-19 pandemic

COVID-19 related deaths underestimate the pandemic burden on mortality because they suffer from completeness and accuracy issues. Excess mortality is a popular alternative, as it compares observed with expected deaths based on the assumption that the pandemic did not occur. Expected deaths had the pandemic not occurred depend on population trends, temperature, and spatio-temporal patterns. In addition to this, high geographical resolution is required to examine within country trends and the effectiveness of the different public health policies. In this tutorial, we propose a framework using R to estimate and visualise excess mortality at high geographical resolution. We show a case study estimating excess deaths during 2020 in Italy. The proposed framework is fast to implement and allows combining different models and presenting the results in any age, sex, spatial and temporal aggregation desired. This makes it particularly powerful and appealing for online monitoring of the pandemic burden and timely policy making.

stat.AP↗

arXiv2022

Counterfactual Learning with Multioutput Deep Kernels

In this paper, we address the challenge of performing counterfactual inference with observational data via Bayesian nonparametric regression adjustment, with a focus on high-dimensional settings featuring multiple actions and multiple correlated outcomes. We present a general class of counterfactual multi-task deep kernels models that estimate causal effects and learn policies proficiently thanks to their sample efficiency gains, while scaling well with high dimensions. In the first part of the work, we rely on Structural Causal Models (SCM) to formally introduce the setup and the problem of identifying counterfactual quantities under observed confounding. We then discuss the benefits of tackling the task of causal effects estimation via stacked coregionalized Gaussian Processes and Deep Kernels. Finally, we demonstrate the use of the proposed methods on simulated experiments that span individual causal effects estimation, off-policy evaluation and optimization.

cs.LG↗

arXiv2022

Interpretable Deep Causal Learning for Moderation Effects

In this extended abstract paper, we address the problem of interpretability and targeted regularization in causal machine learning models. In particular, we focus on the problem of estimating individual causal/treatment effects under observed confounders, which can be controlled for and moderate the effect of the treatment on the outcome of interest. Black-box ML models adjusted for the causal setting perform generally well in this task, but they lack interpretable output identifying the main drivers of treatment heterogeneity and their functional relationship. We propose a novel deep counterfactual learning architecture for estimating individual treatment effects that can simultaneously: i) convey targeted regularization on, and produce quantify uncertainty around the quantity of interest (i.e., the Conditional Average Treatment Effect); ii) disentangle baseline prognostic and moderating effects of the covariates and output interpretable score functions describing their relationship with the outcome. Finally, we demonstrate the use of the method via a simple simulated experiment.

cs.LG↗

arXiv2022

A Bayesian hierarchical model for improving exercise rehabilitation in mechanically ventilated ICU patients

Patients who are mechanically ventilated in the intensive care unit (ICU) participate in exercise as a component of their rehabilitation to ameliorate the long-term impact of critical illness on their physical function. The effective implementation of these programmes is hindered, however, by the lack of a scientific method for quantifying an individual patient's exercise intensity level in real time, which results in a broad one-size-fits-all approach to rehabilitation and sub-optimal patient outcomes. In this work we have developed a Bayesian hierarchical model with temporally correlated latent Gaussian processes to predict $\dot VO_2$, a physiological measure of exercise intensity, using readily available physiological data. Inference was performed using Integrated Nested Laplace Approximation. For practical use by clinicians $\dot VO_2$ was classified into exercise intensity categories. Internal validation using leave-one-patient-out cross-validation was conducted based on these classifications, and the role of probabilistic statements describing the classification uncertainty was investigated.

stat.AP↗

arXiv2022

Blended Survival Curves: A New Approach to Extrapolation for Time-to-Event Outcomes from Clinical Trial in Health Technology Assessment

Background Survival extrapolation is essential in the cost-effectiveness analysis to quantify the lifetime survival benefit associated with a new intervention, due to the restricted duration of randomized controlled trials (RCTs). Current approaches of extrapolation often assume that the treatment effect observed in the trial can continue indefinitely, which is unrealistic and may have a huge impact on decisions for resource allocation. Objective We introduce a novel methodology as a possible solution to alleviate the problem of performing survival extrapolation with heavily censored data from clinical trials. Method The main idea is to mix a flexible model (e.g., Cox semi-parametric) to fit as well as possible the observed data and a parametric model encoding assumptions on the expected behaviour of underlying long-term survival. The two are "blended" into a single survival curve that is identical with the Cox model over the range of observed times and gradually approaching the parametric model over the extrapolation period based on a weight function. The weight function regulates the way two survival curves are blended, determining how the internal and external sources contribute to the estimated survival over time. Results A 4-year follow-up RCT of rituximab in combination with fludarabine and cyclophosphamide v. fludarabine and cyclophosphamide alone for the first-line treatment of chronic lymphocytic leukemia is used to illustrate the method. Conclusion Long-term extrapolation from immature trial data may lead to significantly different estimates with various modelling assumptions. The blending approach provides sufficient flexibility, allowing a wide range of plausible scenarios to be considered as well as the inclusion of genuine external information, based e.g. on hard data or expert opinion. Both internal and external validity can be carefully examined.

stat.ME↗

arXiv2022

Marginalization of Regression-Adjusted Treatment Effects in Indirect Comparisons with Limited Patient-Level Data

Population adjustment methods such as matching-adjusted indirect comparison (MAIC) are increasingly used to compare marginal treatment effects when there are cross-trial differences in effect modifiers and limited patient-level data. MAIC is sensitive to poor covariate overlap and cannot extrapolate beyond the observed covariate space. Current outcome regression-based alternatives can extrapolate but target a conditional treatment effect that is incompatible in the indirect comparison. When adjusting for covariates, one must integrate or average the conditional estimate over the population of interest to recover a compatible marginal treatment effect. We propose a marginalization method based on parametric G-computation that can be easily applied where the outcome regression is a generalized linear model or a Cox model. In addition, we introduce a novel general-purpose method based on multiple imputation, which we term multiple imputation marginalization (MIM) and is applicable to a wide range of models. Both methods can accommodate a Bayesian statistical framework, which naturally integrates the analysis into a probabilistic framework. A simulation study provides proof-of-principle for the methods and benchmarks their performance against MAIC and the conventional outcome regression. The marginalized outcome regression approaches achieve more precise and more accurate estimates than MAIC, particularly when covariate overlap is poor, and yield unbiased marginal treatment effect estimates under no failures of assumptions. Furthermore, the marginalized regression-adjusted estimates provide greater precision and accuracy than the conditional estimates produced by the conventional outcome regression, which are systematically biased because the measure of effect is non-collapsible.

stat.ME↗

arXiv2022

Parametric G-computation for Compatible Indirect Treatment Comparisons with Limited Individual Patient Data

Population adjustment methods such as matching-adjusted indirect comparison (MAIC) are increasingly used to compare marginal treatment effects when there are cross-trial differences in effect modifiers and limited patient-level data. MAIC is based on propensity score weighting, which is sensitive to poor covariate overlap and cannot extrapolate beyond the observed covariate space. Current outcome regression-based alternatives can extrapolate but target a conditional treatment effect that is incompatible in the indirect comparison. When adjusting for covariates, one must integrate or average the conditional estimate over the relevant population to recover a compatible marginal treatment effect. We propose a marginalization method based on parametric G-computation that can be easily applied where the outcome regression is a generalized linear model or a Cox model. The approach views the covariate adjustment regression as a nuisance model and separates its estimation from the evaluation of the marginal treatment effect of interest. The method can accommodate a Bayesian statistical framework, which naturally integrates the analysis into a probabilistic framework. A simulation study provides proof-of-principle and benchmarks the method's performance against MAIC and the conventional outcome regression. Parametric G-computation achieves more precise and more accurate estimates than MAIC, particularly when covariate overlap is poor, and yields unbiased marginal treatment effect estimates under no failures of assumptions. Furthermore, the marginalized regression-adjusted estimates provide greater precision and accuracy than the conditional estimates produced by the conventional outcome regression, which are systematically biased because the measure of effect is non-collapsible.

stat.ME↗