The role of inference in AI: Start S.M.A.L.L. with mindful modeling

This chapter explores trustworthiness in AI and penetrates the black-box opacity through explainable, fair, and ethical AI solutions. AI remains a spirited topic within academic, government, and industrial literature. Much has occurred since the last AI winter in the early 1990's;yet, numerous sources indicate the initial successes solving problems like computer vision, speech recognition, and natural sciences may wane — plunging AI into another winter. Many factors contributed to advances in AI: more data science courses in universities producing data-science capable graduates, high venture capital funding levels encouraging startups, and a decade of broadening awareness among corporate executives about AI promises, real or perceived. Nonetheless, could sources like Gartner be right? Are we approaching another AI winter? As the world learned during the COVID-19 pandemic, when we find ourselves in a crisis, focusing on the fundamentals can have a powerful effect to easing the troubles. As AI makes history, it relies on progress from other domains such as data availability, computing power, and algorithmic advances. Balance among elements maintains a healthy system. AI is no different. Too much or too little of any elemental capability can slow down overall progress. This chapter integrates fundamental ideas from psychology (heuristics and bias), mindfulness in modeling (conceptual models in group settings), and inference (both classical and contemporary). Practitioners may find the techniques proposed in this chapter useful next steps in AI evolution aimed at understanding human behavior. The techniques we discuss can protect against negative impacts resulting from a future AI winter through proper preparation: appreciating the fundamentals, understanding AI assumptions and limitations, and approaching AI assurance in a mindful manner as it evolves. This chapter will address the fundamentals in a unifying example focused on healthcare, with opportunities for trustworthy AI that is impartial, fair, and unbiased. © 2023 Elsevier Inc. All rights reserved.

What drives the performance and causality of green bond indices?

This paper empirically assesses the performance of green bond indices and the causality of that performance using a range of financial and commodity data. We present new insights from the novel application of datasets, neural networks and performance measurements. We find that green bond indices do not outperform the market when factors beyond market return are considered. We find that Brent crude oil has the most significant effect on certain indices, a finding that contrasts with other studies on green bonds. A greater sensitivity to oil prices and global green equities also evinces a negative impact on a green bond index's ability to outperform the market. For the first time, a linear causal relationship is established between Title Transfer Facility (TTF) returns and green bond index returns. Additionally, a fundamental shift in causal relationships is observed over the COVID-19 period. In this way, we contribute to the literature on sustainable green bonds and the impact of COVID-19. These insights provide more clarity to market participants for navigating the uncertainties of both the global energy transition and the postpandemic period. © 2023 Informa UK Limited, trading as Taylor & Francis Group.

Defining estimands in clinical trials: A unified procedure.

The ICH E9 (R1) addendum proposes five strategies to define estimands by addressing intercurrent events. However, mathematical forms of these targeted quantities are lacking, which might lead to discordance between statisticians who estimate these quantities and clinicians, drug sponsors, and regulators who interpret them. To improve the concordance, we provide a unified four-step procedure for constructing the mathematical estimands. We apply the procedure for each strategy to derive the mathematical estimands and compare the five strategies in practical interpretations, data collection, and analytical methods. Finally, we show that the procedure can help ease tasks of defining estimands in settings with multiple types of intercurrent events using two real clinical trials.

Using Randomized Controlled Trials in the Sports Medicine and Performance Environment: Is It Time to Reconsider and Think Outside the Methodological Box?

SYNOPSIS: Randomized controlled trials (RCTs) are ubiquitous in medicine and have facilitated great strides in clinical care. However, when applied in sport, RCTs have limitations that hinder implementing effective interventions in the real-world clinical setting. Pragmatic clinical trials offer some solutions. Yet due to the competitive, high-pressure nature of sport at the individual, team, and governing body level, RCTs are likely infeasible in certain sport settings. The small number of athletes at the elite team level, along with the potential financial consequences of randomizing at the individual athlete and team level, also restricts study power and feasibility, limiting conclusions. Consequently, researchers may need to "think outside the box" and consider other research methodology, to help improve athlete care. In this Viewpoint, we detail alternative study designs that can help solve real-world problems in sports medicine and performance, while maintaining robust research standards and accounting for the challenges that RCTs pose. We also provide practical examples of alternative designs. J Orthop Sports Phys Ther 2023;53(6):1-4. Epub: 18 April 2023. doi:10.2519/jospt.2023.11824.

The effects of international research collaboration on the policy impact of research: A causal inference drawing on the journal Lancet

Research findings have been widely used as evidence for policy-making. The internationalisation of research activities has been increasing in recent decades, particularly during the COVID-19 pandemic. Previous studies have revealed that international research collaboration can enhance the academic impact of research. However, the effects that international research collaboration exerts on the policy impact of research are still unknown. This study aims to examine the effects of international research collaboration on the policy impact of research (as measured by the number of citations in policy documents) using a causal inference approach. Research articles published by the journal Lancet between 2000 and 2019 were selected as the study sample (n = 6098). The number of policy citations of each article was obtained from Overton, the largest database of policy citations. Propensity score matching analysis, which takes a causal inference approach, was used to examine the dataset. Four other matching methods and alternative datasets of different sizes were used to test the robustness of the results. The results of this study reveal that international research collaboration has significant and positive effects on the policy impact of research (coefficient = 4.323, p < 0.001). This study can provide insight to researchers, research institutions and grant funders for improving the policy impact of research. © The Author(s) 2023.

Personalized exercise recommendation method based on causal deep learning: Experiments and implications

The COVID-19 pandemic has accelerated innovations for supporting learning and teaching online. However, online learning also means a reduction of opportunities in direct communication between teachers and students. Given the inevitable diversity in learning progress and achievements for individual online learners, it is difficult for teachers to give personalized guidance to a large number of students. The personalized guidance may cover many aspects, including recommending tailored exercises to a specific student according to the student's knowledge gaps on a subject. In this paper, we propose a personalized exercise recommendation method named causal deep learning (CDL) based on the combination of causal inference and deep learning. Deep learning is used to train and generate initial feature representations for the students and the exercises, and intervention algorithms based on causal inference are then applied to further tune these feature representations. Afterwards, deep learning is again used to predict individual students' score ratings on exercises, from which the Top-N ranked exercises are recommended to similar students who likely need enhancing of skills and understanding of the subject areas indicated by the chosen exercises. Experiments of CDL and four baseline methods on two real-world datasets demonstrate that CDL is superior to the existing methods in terms of capturing students' knowledge gaps in learning and more accurately recommending appropriate exercises to individual students to help bridge their knowledge gaps. © 2022 The Author(s).

Evaluating the effect of SARS-CoV-2 spike mutations with a linear doubly robust learner.

Driven by various mutations on the viral Spike protein, diverse variants of SARS-CoV-2 have emerged and prevailed repeatedly, significantly prolonging the pandemic. This phenomenon necessitates the identification of key Spike mutations for fitness enhancement. To address the need, this manuscript formulates a well-defined framework of causal inference methods for evaluating and identifying key Spike mutations to the viral fitness of SARS-CoV-2. In the context of large-scale genomes of SARS-CoV-2, it estimates the statistical contribution of mutations to viral fitness across lineages and therefore identifies important mutations. Further, identified key mutations are validated by computational methods to possess functional effects, including Spike stability, receptor-binding affinity, and potential for immune escape. Based on the effect score of each mutation, individual key fitness-enhancing mutations such as D614G and T478K are identified and studied. From individual mutations to protein domains, this paper recognizes key protein regions on the Spike protein, including the receptor-binding domain and the N-terminal domain. This research even makes further efforts to investigate viral fitness via mutational effect scores, allowing us to compute the fitness score of different SARS-CoV-2 strains and predict their transmission capacity based solely on their viral sequence. This prediction of viral fitness has been validated using BA.2.12.1, which is not used for regression training but well fits the prediction. To the best of our knowledge, this is the first research to apply causal inference models to mutational analysis on large-scale genomes of SARS-CoV-2. Our findings produce innovative and systematic insights into SARS-CoV-2 and promotes functional studies of its key mutations, serving as reliable guidance about mutations of interest.

Impact of the COVID-19 Pandemic on Clinical Findings in Medical Imaging Exams in a Nationwide Israeli Health Organization: Observational Study.

BACKGROUND: The outbreak of the COVID-19 pandemic had a major effect on the consumption of health care services. Changes in the use of routine diagnostic exams, increased incidences of postacute COVID-19 syndrome (PCS), and other pandemic-related factors may have influenced detected clinical conditions. OBJECTIVE: This study aimed to analyze the impact of COVID-19 on the use of outpatient medical imaging services and clinical findings therein, specifically focusing on the time period after the launch of the Israeli COVID-19 vaccination campaign. In addition, the study tested whether the observed gains in abnormal findings may be linked to PCS or COVID-19 vaccination. METHODS: Our data set included 572,480 ambulatory medical imaging patients in a national health organization from January 1, 2019, to August 31, 2021. We compared different measures of medical imaging utilization and clinical findings therein before and after the surge of the pandemic to identify significant changes. We also inspected the changes in the rate of abnormal findings during the pandemic after adjusting for changes in medical imaging utilization. Finally, for imaging classes that showed increased rates of abnormal findings, we measured the causal associations between SARS-CoV-2 infection, COVID-19-related hospitalization (indicative of COVID-19 complications), and COVID-19 vaccination and future risk for abnormal findings. To adjust for a multitude of confounding factors, we used causal inference methodologies. RESULTS: After the initial drop in the utilization of routine medical imaging due to the first COVID-19 wave, the number of these exams has increased but with lower proportions of older patients, patients with comorbidities, women, and vaccine-hesitant patients. Furthermore, we observed significant gains in the rate of abnormal findings, specifically in musculoskeletal magnetic resonance (MR-MSK) and brain computed tomography (CT-brain) exams. These results also persisted after adjusting for the changes in medical imaging utilization. Demonstrated causal associations included the following: SARS-CoV-2 infection increasing the risk for an abnormal finding in a CT-brain exam (odds ratio [OR] 1.4, 95% CI 1.1-1.7) and COVID-19-related hospitalization increasing the risk for abnormal findings in an MR-MSK exam (OR 3.1, 95% CI 1.9-5.3). CONCLUSIONS: COVID-19 impacted the use of ambulatory imaging exams, with greater avoidance among patients at higher risk for COVID-19 complications: older patients, patients with comorbidities, and nonvaccinated patients. Causal analysis results imply that PCS may have contributed to the observed gains in abnormal findings in MR-MSK and CT-brain exams.

Learning linear non-Gaussian directed acyclic graph with diverging number of nodes

An acyclic model, often depicted as a directed acyclic graph (DAG), has been widely employed to represent directional causal relations among collected nodes. In this article, we propose an efficient method to learn linear non-Gaussian DAG in high dimensional cases, where the noises can be of any continuous non-Gaussian distribution. The proposed method leverages the concept of topological layer to facilitate the DAG learning, and its theoretical justification in terms of exact DAG recovery is also established under mild conditions. Particularly, we show that the topological layers can be exactly reconstructed in a bottom-up fashion, and the parent-child relations among nodes can also be consistently established. The established asymptotic DAG recovery is in sharp contrast to that of many existing learning methods assuming parental faithfulness or ordered noise variances. The advantage of the proposed method is also supported by the numerical comparison against some popular competitors in various simulated examples as well as a real application on the global spread of COVID-19. ©2022 Ruixuan Zhao, Xin He, and Junhui Wang.

Using Mobile Device Data to Understand the Effect of Stay-at-Home Orders on Residents' Mobility

Problem definition: This study addresses three important questions concerning the effectiveness of stay-at-home orders and sociodemographic disparities. (1) What is the average effect of the orders on the percentage of residents staying at home? (2) Is the effect heterogeneous across counties with different percentages of vulnerable populations (defined as those without health insurance or who did not attend high school)? (3) If so, why are the orders less effective for some counties than for others? Academic/practical relevance: To combat the spread of coronavirus disease 2019 (COVID-19), a number of states in the United States implemented stay-at-home orders that prevent residents from leaving their homes except for essential trips. These orders have drawn heavy criticism from the public because whether they are necessary and effective in increasing the number of residents staying at home is unclear. Methodology: We estimate the average effect of the orders using a difference-in-differences model, where the control group is the counties that did not implement the orders and the treatment group is the counties that did implement the orders during our study period. We estimate the heterogeneous effects of the orders by interacting county features with treatment dummies in a triple-difference model. Results: Using a unique set of mobile device data that track residents' mobility, we find that, although some residents already voluntarily stayed at home before the implementation of any order, the stay-at-home orders increased the number of residents staying at home by 2.832 percentage points (or 11.25%). We also find that these orders are less effective for counties with higher percentages of uninsured or less educated (i.e., did not attend high school) residents. To explore the mechanisms behind these results, we analyze the effect of the orders on the average number of work and nonwork trips per person. We find that the orders reduce the number of work trips by 0.053 (or 7.87%) and nonwork trips by 0.183 (or 6.50%). The percentage of uninsured or less educated residents in a county negatively correlates with the reduction in the number of work trips but does not correlate with the reduction in the number of nonwork trips. Managerial implications: Our results suggest that uninsured and less educated residents are less likely to follow the orders because their jobs prevent them from working from home. Policy makers must take into account the differences in residents' socioeconomic status when developing new policies or allocating limited healthcare resources. © 2021 INFORMS.

Estimands and their Estimators for Clinical Trials Impacted by the COVID-19 Pandemic: A Report from the NISS Ingram Olkin Forum Series on Unplanned Clinical Trial Disruptions

The COVID-19 pandemic continues to affect the conduct of clinical trials globally. Complications may arise from pandemic-related operational challenges such as site closures, travel limitations and interruptions to the supply chain for the investigational product, or from health-related challenges such as COVID-19 infections. Some of these complications lead to unforeseen intercurrent events in the sense that they affect either the interpretation or the existence of the measurements associated with the clinical question of interest. In this article, we demonstrate how the ICH E9(R1) Addendum on estimands and sensitivity analyses provides a rigorous basis to discuss potential pandemic-related trial disruptions and to embed these disruptions in the context of study objectives and design elements. We introduce several hypothetical estimand strategies and review various causal inference and missing data methods, as well as a statistical method that combines unbiased and possibly biased estimators for estimation. To illustrate, we describe the features of a stylized trial, and how it may have been impacted by the pandemic. This stylized trial will then be revisited by discussing the changes to the estimand and the estimator to account for pandemic disruptions. Finally, we outline considerations for designing future trials in the context of unforeseen disruptions.Copyright © 2022 American Statistical Association.

CauRuler: Causal irredundant association rule miner for complex patient trajectory modelling.

BACKGROUND AND OBJECTIVES: Discovering causal associations between variables is one of the main goals of clinical trials, with the ultimate aim of identifying the causes of specific health status. Prior knowledge of causal paths could help ensure patients do not develop the resultant conditions. In recent years, thanks to the enormous amount of health data stored with the support of digital tools, attempts have been made to employ Machine Learning to infer causality. Those methodologies suffer from some deficiencies in controlling cofounders when analysing causality, as well as providing causal rules general enough to be useful in healthcare practice. Conversely, this work presents and evaluates CauRuler, a new approach to deal with causality from association rules. The proposed approach uses a pruning strategy to reduce the association rule set, which does not compromise the causality learning capability of the algorithm. This behaviour makes the algorithm suitable for exploiting large health databases with thousands of patients and medical instances. CauRuler can control a larger number of confounders than other proposals, bringing robustness to causal analysis and avoiding the identification of spurious associations. Additionally, the method generalizes causality using anti-monotone properties to obtain complex and general causal paths. The method can target correct causal associations in complex medical databases with retrospective data. METHOD: CauRuler extends association rule mining with an irredundancy property so that the set of rules learnt is reduced in size and generalized. General association rules, conformed by fewer items, enable controlling more confounding variables to verify, with more statistical evidence on available data, if they represent causal paths in patient disease trajectories. RESULTS: CauRuler has been tested on a complex real medical database (3,5 M visits to the primary care services between 2019 and 2020, and controlling over 15.000 different variables including diagnoses and demographic and other clinical patient data). The reduction of the rule set achieved by the pruning strategy goes from 7.732 to 2.240 rules, from which 46 have been found to have causality relationships in the patient trajectories, and generalized to 14 rules tested as true causal relationships thanks to the confounding analysis. These rules have been validated by clinicians with the support of a graphical map. The obtained causal paths control in average of 906 confounder variables, retrieving robust results. CONCLUSIONS: Causal relationships enable predicting causal paths between health conditions according to patient trajectories. Knowing these causal paths is crucial for understanding and preventing the appearance or worsening of diseases in patients. CauRuler, with high demanding thresholds, has proven its efficiency and effectiveness in targeting previously known causal associations between diagnoses, reaching consensus in the medical community. Softening these thresholds should help target interesting general causal paths.

Identifying and estimating effects of sustained interventions under parallel trends assumptions.

Many research questions in public health and medicine concern sustained interventions in populations defined by substantive priorities. Existing methods to answer such questions typically require a measured covariate set sufficient to control confounding, which can be questionable in observational studies. Differences-in-differences rely instead on the parallel trends assumption, allowing for some types of time-invariant unmeasured confounding. However, most existing difference-in-differences implementations are limited to point treatments in restricted subpopulations. We derive identification results for population effects of sustained treatments under parallel trends assumptions. In particular, in settings where all individuals begin follow-up with exposure status consistent with the treatment plan of interest but may deviate at later times, a version of Robins' g-formula identifies the intervention-specific mean under stable unit treatment value assumption, positivity, and parallel trends. We develop consistent asymptotically normal estimators based on inverse-probability weighting, outcome regression, and a double robust estimator based on targeted maximum likelihood. Simulation studies confirm theoretical results and support the use of the proposed estimators at realistic sample sizes. As an example, the methods are used to estimate the effect of a hypothetical federal stay-at-home order on all-cause mortality during the COVID-19 pandemic in spring 2020 in the United States.

Effects of Antidepressants on COVID-19 Outcomes: Retrospective Study on Large-Scale Electronic Health Record Data.

BACKGROUND: Antidepressants exert an anticholinergic effect in varying degrees, and various classes of antidepressants can produce a different effect on immune function. While the early use of antidepressants has a notional effect on COVID-19 outcomes, the relationship between the risk of COVID-19 severity and the use of antidepressants has not been properly investigated previously owing to the high costs involved with clinical trials. Large-scale observational data and recent advancements in statistical analysis provide ample opportunity to virtualize a clinical trial to discover the detrimental effects of the early use of antidepressants. OBJECTIVE: We primarily aimed to investigate electronic health records for causal effect estimation and use the data for discovering the causal effects of early antidepressant use on COVID-19 outcomes. As a secondary aim, we developed methods for validating our causal effect estimation pipeline. METHODS: We used the National COVID Cohort Collaborative (N3C), a database aggregating health history for over 12 million people in the United States, including over 5 million with a positive COVID-19 test. We selected 241,952 COVID-19-positive patients (age >13 years) with at least 1 year of medical history. The study included a 18,584-dimensional covariate vector for each person and 16 different antidepressants. We used propensity score weighting based on the logistic regression method to estimate causal effects on the entire data. Then, we used the Node2Vec embedding method to encode SNOMED-CT (Systematized Nomenclature of Medicine-Clinical Terms) medical codes and applied random forest regression to estimate causal effects. We used both methods to estimate causal effects of antidepressants on COVID-19 outcomes. We also selected few negatively effective conditions for COVID-19 outcomes and estimated their effects using our proposed methods to validate their efficacy. RESULTS: The average treatment effect (ATE) of using any one of the antidepressants was -0.076 (95% CI -0.082 to -0.069; P<.001) with the propensity score weighting method. For the method using SNOMED-CT medical embedding, the ATE of using any one of the antidepressants was -0.423 (95% CI -0.382 to -0.463; P<.001). CONCLUSIONS: We applied multiple causal inference methods with novel application of health embeddings to investigate the effects of antidepressants on COVID-19 outcomes. Additionally, we proposed a novel drug effect analysis-based evaluation technique to justify the efficacy of the proposed method. This study offers causal inference methods on large-scale electronic health record data to discover the effects of common antidepressants on COVID-19 hospitalization or a worse outcome. We found that common antidepressants may increase the risk of COVID-19 complications and uncovered a pattern where certain antidepressants were associated with a lower risk of hospitalization. While discovering the detrimental effects of these drugs on outcomes could guide preventive care, identification of beneficial effects would allow us to propose drug repurposing for COVID-19 treatment.

"Does a Respiratory Virus Have an Ecological Niche, and If So, Can It Be Mapped?" Yes and Yes.

Although the utility of Ecological Niche Models (ENM) and Species Distribution Models (SDM) has been demonstrated in many ecological applications, their suitability for modelling epidemics or pandemics, such as SARS-Cov-2, has been questioned. In this paper, contrary to this viewpoint, we show that ENMs and SDMs can be created that can describe the evolution of pandemics, both in space and time. As an illustrative use case, we create models for predicting confirmed cases of COVID-19, viewed as our target "species", in Mexico through 2020 and 2021, showing that the models are predictive in both space and time. In order to achieve this, we extend a recently developed Bayesian framework for niche modelling, to include: (i) dynamic, non-equilibrium "species" distributions; (ii) a wider set of habitat variables, including behavioural, socio-economic and socio-demographic variables, as well as standard climatic variables; (iii) distinct models and associated niches for different species characteristics, showing how the niche, as deduced through presence-absence data, can differ from that deduced from abundance data. We show that the niche associated with those places with the highest abundance of cases has been highly conserved throughout the pandemic, while the inferred niche associated with presence of cases has been changing. Finally, we show how causal chains can be inferred and confounding identified by showing that behavioural and social factors are much more predictive than climate and that, further, the latter is confounded by the former.

Design and analysis of outcomes following SARS-CoV-2 infection in veterans.

BACKGROUND: Understanding how SARS-CoV-2 infection impacts long-term patient outcomes requires identification of comparable persons with and without infection. We report the design and implementation of a matching strategy employed by the Department of Veterans Affairs' (VA) COVID-19 Observational Research Collaboratory (CORC) to develop comparable cohorts of SARS-CoV-2 infected and uninfected persons for the purpose of inferring potential causative long-term adverse effects of SARS-CoV-2 infection in the Veteran population. METHODS: In a retrospective cohort study, we identified VA health care system patients who were and were not infected with SARS-CoV-2 on a rolling monthly basis. We generated matched cohorts within each month utilizing a combination of exact and time-varying propensity score matching based on electronic health record (EHR)-derived covariates that can be confounders or risk factors across a range of outcomes. RESULTS: From an initial pool of 126,689,864 person-months of observation, we generated final matched cohorts of 208,536 Veterans infected between March 2020-April 2021 and 3,014,091 uninfected Veterans. Matched cohorts were well-balanced on all 39 covariates used in matching after excluding patients for: no VA health care utilization; implausible age, weight, or height; living outside of the 50 states or Washington, D.C.; prior SARS-CoV-2 diagnosis per Medicare claims; or lack of a suitable match. Most Veterans in the matched cohort were male (88.3%), non-Hispanic (87.1%), white (67.2%), and living in urban areas (71.5%), with a mean age of 60.6, BMI of 31.3, Gagne comorbidity score of 1.4 and a mean of 2.3 CDC high-risk conditions. The most common diagnoses were hypertension (61.4%), diabetes (34.3%), major depression (32.2%), coronary heart disease (28.5%), PTSD (25.5%), anxiety (22.5%), and chronic kidney disease (22.5%). CONCLUSION: This successful creation of matched SARS-CoV-2 infected and uninfected patient cohorts from the largest integrated health system in the United States will support cohort studies of outcomes derived from EHRs and sample selection for qualitative interviews and patient surveys. These studies will increase our understanding of the long-term outcomes of Veterans who were infected with SARS-CoV-2.

Effectiveness of Localized Lockdowns in the COVID-19 Pandemic.

Nonpharmaceutical interventions, such as social distancing and lockdowns, have been essential to control of the coronavirus disease 2019 (COVID-19) pandemic. In particular, localized lockdowns in small geographic areas have become an important policy intervention for preventing viral spread in cases of resurgence. These localized lockdowns can result in lower social and economic costs compared with larger-scale suppression strategies. Using an integrated data set from Chile (March 3-June 15, 2020) and a novel synthetic control approach, we estimated the effect of localized lockdowns, disentangling its direct and indirect causal effects on transmission of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Our results showed that the effects of localized lockdowns are strongly modulated by their duration and are influenced by indirect effects from neighboring geographic areas. Our estimates suggest that extending localized lockdowns can slow down SARS-CoV-2 transmission; however, localized lockdowns on their own are insufficient to control pandemic growth in the presence of indirect effects from contiguous neighboring areas that do not have lockdowns. These results provide critical empirical evidence about the effectiveness of localized lockdowns in interconnected geographic areas.

Hunting for protective drugs at the break of a pandemic: Causal inference from hospital data.

At the break of a pandemic, the protective efficacy of therapeutic interventions needs rapid evaluation. An experimental approach to the problem will not always be appropriate. An alternative route are observational studies, whether based on regional health service data or hospital records. In this paper, we discuss the use of methods of causal inference for the analysis of such data, with special reference to causal questions that may arise in a pandemic. We apply the methods by using the aid of a directed acyclic graph (DAG) representation of the problem, to encode our causal assumptions and to logically connect the scientific questions. We illustrate the usefulness of DAGs in the context of a controversy over the effects of renin aldosterone system inhibitors (RASIs) in hypertensive individuals at risk of (or affected by) severe acute respiratory syndrome coronavirus 2 disease. We consider questions concerning the existence and the directions of those effects, their underlying mechanisms, and the possible dependence of the effects on context variables. This paper describes the cognitive steps that led to a DAG representation of the problem, based on background knowledge and evidence from past studies, and the use of the DAG to analyze our hospital data and assess the interpretive limits of the results. Our study contributed to subverting early opinions about RASIs, by suggesting that these drugs may indeed protect the older hypertensive Covid-19 patients from the consequences of the disease. Mechanistic interaction methods revealed that the benefit may be greater (in a sense to be made clear) in the older stratum of the population.