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 pandemic, COVID-19 disease and perinatal health (preprint)

Adverse effects of COVID-19 on perinatal health have been documented, however there is a lack of research that separates individual disease from other changing risks during the pandemic period. We linked California statewide birth and hospital discharge data for 2019-2020, and compared health indicators among 3 groups of pregnancies: [a] 2020 delivery with COVID-19, [b] 2020 delivery with no documented COVID-19, and [c] 2019 pre-pandemic delivery. We aimed to quantify the links between COVID-19 and perinatal health, separating individual COVID-19 disease (a vs b) from the pandemic period (b vs c). We examined the following health indicators: preterm birth, hypertensive disorders of pregnancy, gestational diabetes mellitus and severe maternal morbidity. We applied model based standardization to estimate "average effect of treatment on the treated" risk differences (RD), and adjusted for individual and community-level confounders. Among pregnancies in 2020, those with COVID-19 disease had higher burdens of preterm birth (RD[95% confidence interval (CI)]=2.8%[2.1,3.5]), hypertension (RD[95% CI]=3.3%[2.4,4.1]), and severe maternal morbidity (RD[95% CI]=2.3%[1.9,2.7]) compared with pregnancies without COVID-19 (a vs b) adjusted for confounders. Pregnancies in 2020 without COVID-19 had a lower burden of preterm birth (RD[95% CI]=-0.4%[-0.6,-0.3]), particularly spontaneous preterm, and a higher burden of hypertension (RD[95% CI]=1.0%[0.9,1.2]) and diabetes RD[95%CI]=0.9%[0.8,1.1] compared with pregnancies in 2019 (b vs c) adjusted for confounders. Protective associations of the pandemic period for spontaneous preterm birth may be explained by socioenvironmental and behavioral modifications, while increased maternal conditions may be due to stress and other behavioral changes. To our knowledge, our study is the first to distinguish between individual COVID-19 disease and the pandemic period in connection with perinatal outcomes.

Introducing longitudinal modified treatment policies: a unified framework for studying complex exposures (preprint)

This tutorial discusses a recently developed methodology for causal inference based on longitudinal modified treatment policies (LMTPs). LMTPs generalize many commonly used parameters for causal inference including average treatment effects, and facilitate the mathematical formalization, identification, and estimation of many novel parameters. LMTPs apply to a wide variety of exposures, including binary, multivariate, and continuous, as well as interventions that result in violations of the positivity assumption. LMTPs can accommodate time-varying treatments and confounders, competing risks, loss-to-follow-up, as well as survival, binary, or continuous outcomes. This tutorial aims to illustrate several practical uses of the LMTP framework, including describing different estimation strategies and their corresponding advantages and disadvantages. We provide numerous examples of types of research questions which can be answered within the proposed framework. We go into more depth with one of these examples -- specifically, estimating the effect of delaying intubation on critically ill COVID-19 patients' mortality. We demonstrate the use of the open source R package lmtp to estimate the effects, and we provide code on https://github.com/kathoffman/lmtp-tutorial.