Identification and estimation of mediational effects of longitudinal modified treatment policies (preprint)

We demonstrate a comprehensive semiparametric approach to causal mediation analysis, addressing the complexities inherent in settings with longitudinal and continuous treatments, confounders, and mediators. Our methodology utilizes a nonparametric structural equation model and a cross-fitted sequential regression technique based on doubly robust pseudo-outcomes, yielding an efficient, asymptotically normal estimator without relying on restrictive parametric modeling assumptions. We are motivated by a recent scientific controversy regarding the effects of invasive mechanical ventilation (IMV) on the survival of COVID-19 patients, considering acute kidney injury (AKI) as a mediating factor. We highlight the possibility of "inconsistent mediation," in which the direct and indirect effects of the exposure operate in opposite directions. We discuss the significance of mediation analysis for scientific understanding and its potential utility in treatment decisions.

The ‘Cost of Living Crisis’ and its effects on health: A qualitative study from the UK (preprint)

OBJECTIVE The study explored the impacts of an economic crisis – the UK’s ‘cost of living crisis’ – on mental and physical health from the perspectives of people themselves. It also explored how people coping with this crisis, and as a secondary objective, explored if, and if so how, the COVID-19 pandemic was shaping, their experiences or views of the cost of living crisis.DESIGN This qualitative study consisted of online focus groups carried out between September 14th- 29th 2022. Data were analysed using a thematic approach.SETTING Focus groups took place via (Zoom) videoconferencing.PARTICIPANTS Participants (n= 28) were all UK adults, recruited as part of the Public Views during the Covid Pandemic (PVCOVID) study.RESULTS Most participants, and particularly those on low or insecure income or living in deprived communities, felt that the cost of living crisis was having negative impact on their mental and emotional health and wellbeing. Analysis generated five main themes, three related to why the cost of living crisis was affecting their health: (1) Anxiety over an uncertain future; (2) Worry about others; (3) A loss of control and two related to the ways in which people were coping, or trying to cope, with the crisis: (1) Resilience and Family and community support. CONCLUSIONS. This study found that the cost of living crisis is having significant impacts on health, particularly on the mental health of those on low- or insecure-incomes or from deprived communities. Further research and policy investment is needed to explore ways to provide emotional as well as practical/financial support for those most vulnerable to economic crises.

Navigating the ‘new normal’: Public attitudes and behaviours two years into the COVID-19 pandemic in the UK (preprint)

Objectives The study explored public attitudes and behaviours in relation to COVID-19 two years into the pandemic. Design This Qualitative study consisted of online focus groups carried out between 15th-30th June 2022. Data were analysed using a framework approachSetting Focus groups took place via online videoconferencingParticipants Participants (n= 28) were all UK residents aged 18 years or older, representing a range of gender and ethnic backgrounds, invited from the Public Views during the Covid Pandemic (PVCOVID) study.Results Most participants reported feeling ‘back to normal’, not having thought much about COVID-19 recently, and were not wearing masks or socially distancing. Lack of media coverage was a big factor cited, as was the perception that new variants were ‘milder.’ A minority of participants were still wearing masks or socially distancing and some argued they felt ‘reconditioned’ to be more cautious or less socially active. Identifying COVID-19 symptoms, and distinguishing them from flu was challenging, with some suggesting they would test on ‘instinct’ or if they felt very unwell. Intention to take a COVID-19 test and to socially distance if unwell was generally high. There was a modest appetite for future boosters, with those already triple-jabbed suggesting they would receive another dose in future, particularly if it was officially recommended. Some participants argued they would only change their behaviour if the situation was ‘serious’ (e.g. if the death rate increased significantly). Most participants said they would adopt more caution if there was a future wave, although many argued that the lack of trust caused by UK political figures breaking rules (‘Partygate’) would harm future compliance. Conclusions. The study suggests that the ‘Living with COVID’ strategy and the relative lack of media coverage have contributed to a sense of things being ‘back to normal’. If future developments in the pandemic require behavioural change, public health policy and communication will need to overcome this perception and barriers such as a lack of trust in government and the perception of the virus as milder.

Corticosteroids in COVID-19: Optimizing Observational Research through Target Trial Emulations (preprint)

Background: Observational research provides a unique opportunity to learn causal effects when randomized trials are not available, but obtaining the correct estimates hinges on a multitude of design and analysis choices. We illustrate the advantages of modern causal inference methods and compare to standard research practice to estimate the effect of corticosteroids on mortality in hospitalized COVID-19 patients in an observational dataset. We use several large RCTs to benchmark our results. Methods: Our retrospective data source consists of 3,293 COVID-19 patients hospitalized at New York Presbyterian March 1-May 15, 2020. We design our study using the Target Trial Emulation framework. We estimate the effect of an intervention consisting of 6 days of corticosteroids administered at the time of severe hypoxia and contrast with an intervention consisting of no corticosteroids administration. The dataset includes dozens of time-varying confounders. We estimate the causal effects using a doubly robust estimator where the probabilities of treatment, outcome, and censoring are estimated using flexible regressions via super learning. We compare these analyses to standard practice in clinical research, consisting of two main methods: (i) Cox models for an exposure of corticosteroids receipt within various time windows of hypoxia, and (ii) a Cox time-varying model where the exposure is daily administration of corticosteroids starting at the time of hospitalization. Results: The effect in our target trial emulation is qualitatively identical to an RCT benchmark, estimated to reduce 28-day mortality from 32% (95% confidence interval: 31-34) to 23% (21-24). The estimated effect from meta-analyses of RCTs for corticosteroids is an odds ratio of 0.66 (0.53-0.82)(1). Hazard ratios from the Cox models range in size and direction from 0.50 (0.41-0.62) to 1.08 (0.80-1.47) and all study designs suffer from various forms of bias. Conclusion: We demonstrate in a case study that clinical research based on observational data can unveil true causal relations. However, the correctness of these effect estimates requires designing and analyzing the data based on principles which are different from the current standard in clinical research. The widespread communication and adoption of these design and analytical techniques is of high importance for the improvement of clinical research based on observational data.

Public attitudes to social care in Wales following the COVID-19 pandemic (preprint)

Background. The COVID-19 pandemic has shone further light on some of the challenges facing social care in Wales, as in many countries, and looks to have exacerbated a crisis that was already extant. This has led to the intensification of longer-standing arguments that social reform is necessary and that the pandemic presents an added impetus and opportunity for reform. Methods. An online survey was completed by 2569 respondents between February 11th and March 11th, 2022. Additionally, online focus groups were conducted with a sample of 14 participants. The inclusion criteria were adults aged 18 years and over living in Wales. Results. Four-in-ten of those who felt that they or someone in their household or close family needed social care during the past two years did not receive or make use of it. The pandemic was cited as a major reason why many of those who may have needed social care didn’t access it Satisfaction with social care was variable, with approximately one-third either very or quite dissatisfied, and a little over half either very or quite satisfied with social care services for themselves or a household or close family member. Discussion. Social care policymakers and providers should seek to understand and address what people feel are the main barriers to accessing or using social care, including increasing provision for those who need it, encouraging and enabling those who feel they need social care to apply, consider broadening the eligibility criteria where appropriate, and simplifying the application process.

In These Uncertain Times: Fake News Amplifies the Desires to Save and Spend in Response to COVID-19

COVID-19 is changing human history. However, it arrived on the heels of another crisis: fake news. We examine how fake news influences consumers’ spending intentions in response to COVID-19. Across three studies we find that concerns about COVID-19 engender uncertainty, and that exposure to fake news amplifies this effect. This uncertainty increases consumers’ tendency toward two competing goals: compensatory consumption and resource conservation. We present three studies in which we measure consumers’ general preferences (study 1), their specific preferences with respect to buying food at a restaurant (study 2), and their choices when selecting from a meal delivery service (study 3). Our findings have important implications for both marketing practitioners and policy makers, which we discuss throughout.

‘Variant fatigue’? Public attitudes to COVID-19 18 months into the pandemic: A qualitative study (preprint)

Objectives: Qualitative study exploring public attitudes to COVID-19 18 months into the pandemic, specifically focused on adherence to infection-reducing behaviours and policy measures during a period of the emergence of a new variant (Omicron)Study design: Qualitative online focus group study Methods: Focus groups were conducted with a diverse sample of 22 adults in the United Kingdom to explore their views. Data were analysed using a framework approach. Findings: Analysis revealed two main groups based on participants’ perceived concern over Omicron: variant fatigue (n=16 (73%)) and deja vu (n=6 (27%)). Those exhibiting variant fatigue reported not adopting any additional caution or infection-reducing behaviours as a result of the new variant. They tended to describe Omicron as ‘just another variant’ and expressed a need to ‘get on’ and ‘live with’ the virus. Those exhibiting deja vu suggested that Omicron was of additional concern to them and for some posed a threat not seen since ‘last year’ (second wave). No demographic patterns emerged, although there was a high amount of variant fatigue (absence of additional caution) (n=5 (83%)) amongst the unvaccinated participants. Those who were concerned about Omicron tended to report reducing social contacts. Few participants reported taking lateral flow tests regularly, except for those required to by their employers. Stated compliance with facemask rules was high. Nearly all participants stated an intention to comply with any future potential policy measures to reduce transmission of COVID-19, including more stringent measures such as lockdowns. Implications: 18 months into the pandemic, there may be habituation to the risk posed by COVID-19, despite the increased risk posed by the new variant Omicron. Due to this risk habituation (‘variant fatigue’) and due to a general decline in engagement with news related to COVID-19, many people may not be, or might be reluctant to, voluntarily adopt additional caution and infection-reducing behaviours. This poses a challenge for public health communication, since a sense of being ‘relaxed’ about, or ‘living with’ COVID-19 may undermine efforts to encourage voluntary adherence to infection-reducing behaviours. However, findings suggest that most people intend to comply (albeit reluctantly) with policy measures (as opposed to ‘advice’)- including stringent measures such as lockdown - if they were required in future.

Public attitudes to COVID-19 ‘booster vaccinations and influenza vaccines: A qualitative focus group study (preprint)

Objectives This study explored public attitudes to COVID-19 ‘booster’ vaccines and influenza vaccines, including participants intentions to have them or not, and their preferences on how and when they should be administered. Study design. Qualitative study using online focus groups and interviews.Methods Group and individual online interviews were conducted with a diverse sample of 21 adults in the United Kingdom to explore their views on the issue of COVID-19 booster and influenza vaccination. Data were analysed using a framework approach.Results Three themes emerged to unpack participants views on COVID-19 booster and flu vaccinations: (1) Booster intentions correspond with decisions around initial vaccination (2) Mixed views on booster and flu vaccines (3) Mixed views on whether to get COVID-19 booster and flu vaccines together or separately. Those who had been more willing to accept their initial COVID-19 vaccine were more willing to accept a booster and a flu vaccine, whereas those who had been more hesitant about their initial COVID-19 vaccine were more hesitant about getting a booster and a flu vaccine – including the idea of getting them together.Conclusions Public health messaging related to COVID-19 boosters might emphasise that they can be seen as: an extension of their initial decision to get vaccination; a way to maintain population immunity over the longer term; a collective act (to protect others); a way to maintain freedoms or ‘normality’. Messaging around boosters and influenza vaccines needs to emphasise they are safe and convenient to be taken together and that both are important, especially for those clinically vulnerable

Optimizing Precision and Power by Machine Learning in Randomized Trials, with an Application to COVID-19 (preprint)

The rapid finding of effective therapeutics requires the efficient use of available resources in clinical trials. The use of covariate adjustment can yield statistical estimates with improved precision, resulting in a reduction in the number of participants required to draw futility or efficacy conclusions. We focus on time-to-event and ordinal outcomes. A key question for covariate adjustment in randomized studies is how to fit a model relating the outcome and the baseline covariates to maximize precision. We present a novel theoretical result establishing conditions for asymptotic normality of a variety of covariate-adjusted estimators that rely on machine learning (e.g., l1-regularization, Random Forests, XGBoost, and Multivariate Adaptive Regression Splines), under the assumption that outcome data is missing completely at random. We further present a consistent estimator of the asymptotic variance. Importantly, the conditions do not require the machine learning methods to converge to the true outcome distribution conditional on baseline variables, as long as they converge to some (possibly incorrect) limit. We conducted a simulation study to evaluate the performance of the aforementioned prediction methods in COVID-19 trials using longitudinal data from over 1,500 patients hospitalized with COVID-19 at Weill Cornell Medicine New York Presbyterian Hospital. We found that using l1-regularization led to estimators and corresponding hypothesis tests that control type 1 error and are more precise than an unadjusted estimator across all sample sizes tested. We also show that when covariates are not prognostic of the outcome, l1-regularization remains as precise as the unadjusted estimator, even at small sample sizes (n = 100). We give an R package adjrct that performs model-robust covariate adjustment for ordinal and time-to-event outcomes.

Public attitudes to COVID-19 vaccines: A qualitative study (preprint)

OBJECTIVE: To explore public attitudes to COVID-19 vaccines in the UK, focused on intentions and decisions around taking vaccines, views on ‘vaccine passports’, and experiences and perspectives on post-vaccination behaviour.DESIGN: Qualitative study consisting of 6 online focus groups conducted between 15th March – 22nd April 2021. SETTING: Online video conferencingPARTICIPANTS: 29 adult UK-based participantsRESULTS: Three main groups regarding participants’ decision or intention to receive a COVID-19 vaccine were identified: (1) Accepters, (2) Delayers and (3) Refusers. Two reasons for vaccine delay were identified: delay due to a perceived need more information and delay until vaccine was “required” in the future. Three main facilitators (Vaccination as a social norm; Vaccination as a necessity; Trust in science) and six barriers (Preference for “natural immunity”; Concerns over possible side effects; Distrust in government; Perceived lack of information; Conspiracy theories; “Covid echo chambers”) to vaccine uptake were identified. For some delayers, 'vaccine passports' were perceived to be a reason why they would get vaccinated in the future. However, vaccine passports were controversial, and were framed in four main ways: as “a necessary evil”; as “Orwellian”; as a “human rights problem”; and as a source of confidence”. Participants generally felt that receiving a vaccine was not changing the extent to which people were adhering to COVID-19 measures. CONCLUSIONS: Overall positive sentiment around vaccines is high. However, there remains a number of potential barriers which might be leading to vaccine delay in some. ‘Vaccine delay’ might be a more useful and precise construct than vaccine hesitancy in explaining why some may initially ignore or be uncertain about vaccination invitations. Vaccine passports may increase or ‘nudge’ uptake in some delayers but may be unpopular in others. Earlier concerns that vaccination might reduce adherence to social distancing measures are not borne out in our data, with most people reporting adherence and caution.