A self-censoring model for multivariate nonignorable nonmonotone missing data.

We introduce an itemwise modeling approach called "self-censoring" for multivariate nonignorable nonmonotone missing data, where the missingness process of each outcome can be affected by its own value and associated with missingness indicators of other outcomes, while conditionally independent of the other outcomes. The self-censoring model complements previous graphical approaches for the analysis of multivariate nonignorable missing data. It is identified under a completeness condition stating that any variability in one outcome can be captured by variability in the other outcomes among complete cases. For estimation, we propose a suite of semiparametric estimators including doubly robust estimators that deliver valid inferences under partial misspecification of the full-data distribution. We also provide a novel and flexible global sensitivity analysis procedure anchored at the self-censoring. We evaluate the performance of the proposed methods with simulations and apply them to analyze a study about the effect of highly active antiretroviral therapy on preterm delivery of HIV-positive mothers.

Controlling the chaos: Information management in home-infusion central-line-associated bloodstream infection (CLABSI) surveillance.

Objectives: Access to patient information may affect how home-infusion surveillance staff identify central-line-associated bloodstream infections (CLABSIs). We characterized information hazards in home-infusion CLABSI surveillance and identified possible strategies to mitigate information hazards. Design: Qualitative study using semistructured interviews. Setting and participants: The study included 21 clinical staff members involved in CLABSI surveillance at 5 large home-infusion agencies covering 13 states and the District of Columbia. Methods: Interviews were conducted by 1 researcher. Transcripts were coded by 2 researchers; consensus was reached by discussion. Results: Data revealed the following barriers: information overload, information underload, information scatter, information conflict, and erroneous information. Respondents identified 5 strategies to mitigate information chaos: (1) engage information technology in developing reports; (2) develop streamlined processes for acquiring and sharing data among staff; (3) enable staff access to hospital electronic health records; (4) use a single, validated, home-infusion CLABSI surveillance definition; and (5) develop relationships between home-infusion surveillance staff and inpatient healthcare workers. Conclusions: Information chaos occurs in home-infusion CLABSI surveillance and may affect the development of accurate CLABSI rates in home-infusion therapy. Implementing strategies to minimize information chaos will enhance intra- and interteam collaborations in addition to improving patient-related outcomes.

Implementing and validating a home-infusion central-line-associated bloodstream infection surveillance definition.

OBJECTIVE: Central-line-associated bloodstream infection (CLABSI) surveillance in home infusion therapy is necessary to track efforts to reduce infections, but a standardized, validated, and feasible definition is lacking. We tested the validity of a home-infusion CLABSI surveillance definition and the feasibility and acceptability of its implementation. DESIGN: Mixed-methods study including validation of CLABSI cases and semistructured interviews with staff applying these approaches. SETTING: This study was conducted in 5 large home-infusion agencies in a CLABSI prevention collaborative across 14 states and the District of Columbia. PARTICIPANTS: Staff performing home-infusion CLABSI surveillance. METHODS: From May 2021 to May 2022, agencies implemented a home-infusion CLABSI surveillance definition, using 3 approaches to secondary bloodstream infections (BSIs): National Healthcare Safety Program (NHSN) criteria, modified NHSN criteria (only applying the 4 most common NHSN-defined secondary BSIs), and all home-infusion-onset bacteremia (HiOB). Data on all positive blood cultures were sent to an infection preventionist for validation. Surveillance staff underwent semistructured interviews focused on their perceptions of the definition 1 and 3-4 months after implementation. RESULTS: Interrater reliability scores overall ranged from κ = 0.65 for the modified NHSN criteria to κ = 0.68 for the NHSN criteria to κ = 0.72 for the HiOB criteria. For the NHSN criteria, the agency-determined rate was 0.21 per 1,000 central-line (CL) days, and the validator-determined rate was 0.20 per 1,000 CL days. Overall, implementing a standardized definition was thought to be a positive change that would be generalizable and feasible though time-consuming and labor intensive. CONCLUSIONS: The home-infusion CLABSI surveillance definition was valid and feasible to implement.

Examining the Causal Mediating Role of Cardiovascular Disease on the Effect of Subclinical Cardiovascular Disease on Cognitive Impairment via Separable Effects.

BACKGROUND: An important epidemiological question is understanding how vascular risk factors contribute to cognitive impairment. Using data from the Cardiovascular Health Cognition Study, we investigated how subclinical cardiovascular disease (sCVD) relates to cognitive impairment risk and the extent to which the hypothesized risk is mediated by the incidence of clinically manifested cardiovascular disease (CVD), both overall and within apolipoprotein E-4 (APOE-4) subgroups. METHODS: We adopted a novel "separable effects" causal mediation framework that assumes that sCVD has separably intervenable atherosclerosis-related components. We then ran several mediation models, adjusting for key covariates. RESULTS: We found that sCVD increased overall risk of cognitive impairment (risk ratio [RR] = 1.21, 95% confidence interval [CI]: 1.03, 1.44); however, there was little or no mediation by incident clinically manifested CVD (indirect effect RR = 1.02, 95% CI: 1.00, 1.03). We also found attenuated effects among APOE-4 carriers (total effect RR = 1.09, 95% CI: 0.81, 1.47; indirect effect RR = 0.99, 95% CI: 0.96, 1.01) and stronger findings among noncarriers (total effect RR = 1.29, 95% CI: 1.05, 1.60; indirect effect RR = 1.02, 95% CI: 1.00, 1.05). In secondary analyses restricting cognitive impairment to only incident dementia cases, we found similar effect patterns. CONCLUSIONS: We found that the effect of sCVD on cognitive impairment does not seem to be mediated by CVD, both overall and within APOE-4 subgroups. Our results were critically assessed via sensitivity analyses, and they were found to be robust. Future work is needed to fully understand the relationship between sCVD, CVD, and cognitive impairment.

Nonmyopic and pseudo-nonmyopic approaches to optimal sequential design in the presence of covariates.

In sequential experiments, subjects become available for the study over a period of time, and covariates are often measured at the time of arrival. We consider the setting where the sample size is fixed but covariate values are unknown until subjects enrol. Given a model for the outcome, a sequential optimal design approach can be used to allocate treatments to minimize the variance of the estimator of the treatment effect. We extend existing optimal design methodology so it can be used within a nonmyopic framework, where treatment allocation for the current subject depends not only on the treatments and covariates of the subjects already enrolled in the study, but also the impact of possible future treatment assignments within a specified horizon. The nonmyopic approach requires recursive formulae and suffers from the curse of dimensionality. We propose a pseudo-nonmyopic approach which has a similar aim to the nonmyopic approach, but does not involve recursion and instead relies on simulating trajectories of future possible decisions. Our simulation studies show that, for the simple case of a logistic regression with a single binary covariate and a binary treatment, and a more realistic case with four binary covariates, binary treatment and treatment-covariate interactions, the nonmyopic and pseudo-nonmyopic approaches provide no competitive advantage over the myopic approach, both in terms of the size of the estimated treatment effect and also the efficiency of the designs. Results are robust to the size of the horizon used in the nonmyopic approach, and the number of simulated trajectories used in the pseudo-nonmyopic approach.

Causal determinants of postoperative length of stay in cardiac surgery using causal graphical learning.

OBJECTIVE: We aimed to learn the causal determinants of postoperative length of stay in cardiac surgery patients undergoing isolated coronary artery bypass grafting or aortic valve replacement surgery. METHODS: For patients undergoing isolated coronary artery bypass grafting or isolated aortic valve replacement surgeries between 2011 and 2016, we used causal graphical modeling on electronic health record data. The Fast Causal Inference (FCI) algorithm from the Tetrad software was used on data to estimate a Partial Ancestral Graph (PAG) depicting direct and indirect causes of postoperative length of stay, given background clinical knowledge. Then, we used the latent variable intervention-calculus when the directed acyclic graph is absent (LV-IDA) algorithm to estimate strengths of causal effects of interest. Finally, we ran a linear regression for postoperative length of stay to contrast statistical associations with what was learned by our causal analysis. RESULTS: In our cohort of 2610 patients, the mean postoperative length of stay was 219 hours compared with the Society of Thoracic Surgeons 2016 national mean postoperative length of stay of approximately 168 hours. Most variables that clinicians believe to be predictors of postoperative length of stay were found to be causes, but some were direct (eg, age, diabetes, hematocrit, total operating time, and postoperative complications), and others were indirect (including gender, race, and operating surgeon). The strongest average causal effects on postoperative length of stay were exhibited by preoperative dialysis (209 hours); neuro-, pulmonary-, and infection-related postoperative complications (315 hours, 89 hours, and 131 hours, respectively); reintubation (61 hours); extubation in operating room (-47 hours); and total operating room duration (48 hours). Linear regression coefficients diverged from causal effects in magnitude (eg, dialysis) and direction (eg, crossclamp time). CONCLUSIONS: By using retrospective electronic health record data and background clinical knowledge, causal graphical modeling retrieved direct and indirect causes of postoperative length of stay and their relative strengths. These insights will be useful in designing clinical protocols and targeting improvements in patient management.

Semiparametric Inference for Nonmonotone Missing-Not-at-Random Data: The No Self-Censoring Model.

We study the identification and estimation of statistical functionals of multivariate data missing non-monotonically and not-at-random, taking a semiparametric approach. Specifically, we assume that the missingness mechanism satisfies what has been previously called "no self-censoring" or "itemwise conditionally independent nonresponse," which roughly corresponds to the assumption that no partially-observed variable directly determines its own missingness status. We show that this assumption, combined with an odds ratio parameterization of the joint density, enables identification of functionals of interest, and we establish the semiparametric efficiency bound for the nonparametric model satisfying this assumption. We propose a practical augmented inverse probability weighted estimator, and in the setting with a (possibly high-dimensional) always-observed subset of covariates, our proposed estimator enjoys a certain double-robustness property. We explore the performance of our estimator with simulation experiments and on a previously-studied data set of HIV-positive mothers in Botswana.

Leveraging Machine Learning to Predict 30-Day Hospital Readmission After Cardiac Surgery.

BACKGROUND: Hospital readmission within 30 days of discharge is a well-studied outcome. Predicting readmission after cardiac surgery, however, is notoriously challenging; the best-performing models in the literature have areas under the curve around .65. A reliable predictive model would enable clinicians to identify patients at risk for readmission and to develop prevention strategies. METHODS: We analyzed The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database at our institution, augmented with electronic medical record data. Predictors included demographics, preoperative comorbidities, proxies for intraoperative risk, indicators of postoperative complications, and time series-derived variables. We trained several machine learning models, evaluating each on a held-out test set. RESULTS: Our analysis cohort consisted of 4924 cases from 2011 to 2016. Of those, 723 (14.7%) were readmitted within 30 days of discharge. Our models included 141 STS-derived and 24 electronic medical records-derived variables. A random forest model performed best, with test area under the curve 0.76 (95% confidence interval, 0.73 to 0.79). Using exclusively preoperative variables, as in STS calculated risk scores, degraded the area under the curve, to 0.64 (95% confidence interval, 0.60 to 0.68). Key predictors included length of stay (12.5 times more important than the average variable) and whether the patient was discharged to a rehabilitation facility (11.2 times). CONCLUSIONS: Our approach, augmenting STS variables with electronic medical records data and using flexible machine learning modeling, yielded state-of-the-art performance for predicting 30-day readmission. Separately, the importance of variables not directly related to inpatient care, such as discharge location, amplifies questions about the efficacy of assessing care quality by readmissions.

Discussion on "Causal mediation of semicompeting risks" by Yen-Tsung Huang.

Huang proposes a method for assessing the impact of a point treatment on mortality either directly or mediated by occurrence of a nonterminal health event, based on data from a prospective cohort study in which the occurrence of the nonterminal health event may be preemptied by death but not vice versa. The author uses a causal mediation framework to formally define causal quantities known as natural (in)direct effects. The novelty consists of adapting these concepts to a continuous-time modeling framework based on counting processes. In an effort to posit "scientifically interpretable estimands," statistical and causal assumptions are introduced for identification. In this commentary, we argue that these assumptions are not only difficult to interpret and justify, but are also likely violated in the hepatitis B motivating example and other survival/time to event settings as well.

Auto-G-Computation of Causal Effects on a Network.

Methods for inferring average causal effects have traditionally relied on two key assumptions: (i) the intervention received by one unit cannot causally influence the outcome of another; and (ii) units can be organized into nonoverlapping groups such that outcomes of units in separate groups are independent. In this article, we develop new statistical methods for causal inference based on a single realization of a network of connected units for which neither assumption (i) nor (ii) holds. The proposed approach allows both for arbitrary forms of interference, whereby the outcome of a unit may depend on interventions received by other units with whom a network path through connected units exists; and long range dependence, whereby outcomes for any two units likewise connected by a path in the network may be dependent. Under network versions of consistency and no unobserved confounding, inference is made tractable by an assumption that the networks outcome, treatment and covariate vectors are a single realization of a certain chain graph model. This assumption allows inferences about various network causal effects via the auto-g-computation algorithm, a network generalization of Robins' well-known g-computation algorithm previously described for causal inference under assumptions (i) and (ii). Supplementary materials for this article are available online.

General Identification of Dynamic Treatment Regimes Under Interference.

In many applied fields, researchers are often interested in tailoring treatments to unit-level characteristics in order to optimize an outcome of interest. Methods for identifying and estimating treatment policies are the subject of the dynamic treatment regime literature. Separately, in many settings the assumption that data are independent and identically distributed does not hold due to inter-subject dependence. The phenomenon where a subject's outcome is dependent on his neighbor's exposure is known as interference. These areas intersect in myriad real-world settings. In this paper we consider the problem of identifying optimal treatment policies in the presence of interference. Using a general representation of interference, via Lauritzen-Wermuth-Freydenburg chain graphs (Lauritzen and Richardson, 2002), we formalize a variety of policy interventions under interference and extend existing identification theory (Tian, 2008; Sherman and Shpitser, 2018). Finally, we illustrate the efficacy of policy maximization under interference in a simulation study.

Deriving Bounds and Inequality Constraints Using Logical Relations Among Counterfactuals.

Causal parameters may not be point identified in the presence of unobserved confounding. However, information about non-identified parameters, in the form of bounds, may still be recovered from the observed data in some cases. We develop a new general method for obtaining bounds on causal parameters using rules of probability and restrictions on counterfactuals implied by causal graphical models. We additionally provide inequality constraints on functionals of the observed data law implied by such causal models. Our approach is motivated by the observation that logical relations between identified and non-identified counterfactual events often yield information about non-identified events. We show that this approach is powerful enough to recover known sharp bounds and tight inequality constraints, and to derive novel bounds and constraints.

Full Law Identification in Graphical Models of Missing Data: Completeness Results.

Missing data has the potential to affect analyses conducted in all fields of scientific study including healthcare, economics, and the social sciences. Several approaches to unbiased inference in the presence of non-ignorable missingness rely on the specification of the target distribution and its missingness process as a probability distribution that factorizes with respect to a directed acyclic graph. In this paper, we address the longstanding question of the characterization of models that are identifiable within this class of missing data distributions. We provide the first completeness result in this field of study - necessary and sufficient graphical conditions under which, the full data distribution can be recovered from the observed data distribution. We then simultaneously address issues that may arise due to the presence of both missing data and unmeasured confounding, by extending these graphical conditions and proofs of completeness, to settings where some variables are not just missing, but completely unobserved.

Identification and Estimation of Causal Effects Defined by Shift Interventions.

Causal inference quantifies cause effect relationships by means of counterfactual responses had some variable been artificially set to a constant. A more refined notion of manipulation, where a variable is artificially set to a fixed function of its natural value is also of interest in particular domains. Examples include increases in financial aid, changes in drug dosing, and modifying length of stay in a hospital. We define counterfactual responses to manipulations of this type, which we call shift interventions. We show that in the presence of multiple variables being manipulated, two types of shift interventions are possible. Shift interventions on the treated (SITs) are defined with respect to natural values, and are connected to effects of treatment on the treated. Shift interventions as policies (SIPs) are defined recursively with respect to values of responses to prior shift interventions, and are connected to dynamic treatment regimes. We give sound and complete identification algorithms for both types of shift interventions, and derive efficient semi-parametric estimators for the mean response to a shift intervention in a special case motivated by a healthcare problem. Finally, we demonstrate the utility of our method by using an electronic health record dataset to estimate the effect of extending the length of stay in the intensive care unit (ICU) in a hospital by an extra day on patient ICU readmission probability.

Identification in Causal Models With Hidden Variables.

Targets of inference that establish causality are phrased in terms of counterfactual responses to interventions. These potential outcomes operationalize cause effect relationships by means of comparisons of cases and controls in hypothetical randomized controlled experiments. In many applied settings, data on such experiments is not directly available, necessitating assumptions linking the counterfactual target of inference with the factual observed data distribution. This link is provided by causal models. Originally defined on potential outcomes directly (Rubin, 1976), causal models have been extended to longitudinal settings (Robins, 1986), and reformulated as graphical models (Spirtes et al., 2001; Pearl, 2009). In settings where common causes of all observed variables are themselves observed, many causal inference targets are identified via variations of the expression referred to in the literature as the g-formula (Robins, 1986), the manipulated distribution (Spirtes et al., 2001), or the truncated factorization (Pearl, 2009). In settings where hidden variables are present, identification results become considerably more complicated. In this manuscript, we review identification theory in causal models with hidden variables for common targets that arise in causal inference applications, including causal effects, direct, indirect, and path-specific effects, and outcomes of dynamic treatment regimes. We will describe a simple formulation of this theory (Tian and Pearl, 2002; Shpitser and Pearl, 2006b,a; Tian, 2008; Shpitser, 2013) in terms of causal graphical models, and the fixing operator, a statistical analogue of the intervention operation (Richardson et al., 2017).

Feature engineering with clinical expert knowledge: A case study assessment of machine learning model complexity and performance.

Incorporating expert knowledge at the time machine learning models are trained holds promise for producing models that are easier to interpret. The main objectives of this study were to use a feature engineering approach to incorporate clinical expert knowledge prior to applying machine learning techniques, and to assess the impact of the approach on model complexity and performance. Four machine learning models were trained to predict mortality with a severe asthma case study. Experiments to select fewer input features based on a discriminative score showed low to moderate precision for discovering clinically meaningful triplets, indicating that discriminative score alone cannot replace clinical input. When compared to baseline machine learning models, we found a decrease in model complexity with use of fewer features informed by discriminative score and filtering of laboratory features with clinical input. We also found a small difference in performance for the mortality prediction task when comparing baseline ML models to models that used filtered features. Encoding demographic and triplet information in ML models with filtered features appeared to show performance improvements from the baseline. These findings indicated that the use of filtered features may reduce model complexity, and with little impact on performance.

Spatial Radiation Dose Influence on Xerostomia Recovery and Its Comparison to Acute Incidence in Patients With Head and Neck Cancer.

PURPOSE: Radiation-induced xerostomia is one of the most prevalent symptoms during and after head and neck cancer radiation therapy (RT). We aimed to discover the spatial radiation dose-based (voxel dose) importance pattern in the major salivary glands in relation to the recovery of xerostomia 18 months after RT, and to compare the recovery voxel dose importance pattern to the acute incidence (injury) pattern. METHODS AND MATERIALS: This study included all patients within our database with xerostomia outcomes after completion of curative intensity modulated RT. Common Terminology Criteria for Adverse Events xerostomia grade was used to define recovered versus nonrecovered group at baseline, between end of treatment and 18 months post-RT, and beyond 18 months, respectively. Ridge logistic regression was performed to predict the probability of xerostomia recovery. Voxel doses within geometrically defined parotid glands (PG) and submandibular glands (SMG), demographic characteristics, and clinical factors were included in the algorithm. We plotted the normalized learned weights on the 3-dimensional PG and SMG structures to visualize the voxel dose importance for predicting xerostomia recovery. RESULTS: A total of 146 head and neck cancer patients from 2008 to 2016 were identified. The superior region of the ipsilateral and contralateral PG was the most influencial for xerostomia recovery. The area under the receiver operating characteristic curve evaluated using 10-fold cross-validation for ridge logistic regression was 0.68 ± 0.07. Compared with injury, the recovery voxel dose importance pattern was more symmetrical and was influenced by lower dose voxels. CONCLUSIONS: The superior portion of the 2 PGs (low dose region) are the most influential on xerostomia recovery and seem to be equal in their contribution. The dissimilarity of the influence pattern between injury and recovery suggests different underlying mechanisms. The importance pattern identified by spatial radiation dose and machine learning methods can improve our understanding of normal tissue toxicities in RT. Further external validation is warranted.

Causal inference, social networks and chain graphs.

Traditionally, statistical inference and causal inference on human subjects rely on the assumption that individuals are independently affected by treatments or exposures. However, recently there has been increasing interest in settings, such as social networks, where individuals may interact with one another such that treatments may spill over from the treated individual to their social contacts and outcomes may be contagious. Existing models proposed for causal inference using observational data from networks of interacting individuals have two major shortcomings. First, they often require a level of granularity in the data that is infeasible in practice to collect in most settings and, second, the models are high dimensional and often too big to fit to the available data. We illustrate and justify a parsimonious parameterization for network data with interference and contagion. Our parameterization corresponds to a particular family of graphical models known as chain graphs. We argue that, in some settings, chain graph models approximate the marginal distribution of a snapshot of a longitudinal data-generating process on interacting units. We illustrate the use of chain graphs for causal inference about collective decision making in social networks by using data from US Supreme Court decisions between 1994 and 2004 and in simulations.

Examining the causal mediating role of brain pathology on the relationship between diabetes and cognitive impairment: the Cardiovascular Health Study.

The paper examines whether diabetes mellitus leads to incident mild cognitive impairment and dementia through brain hypoperfusion and white matter disease. We performed inverse odds ratio weighted causal mediation analyses to decompose the effect of diabetes on cognitive impairment into direct and indirect effects, and we found that approximately a third of the total effect of diabetes is mediated through vascular-related brain pathology. Our findings lend support for a common aetiological hypothesis regarding incident cognitive impairment, which is that diabetes increases the risk of clinical cognitive impairment in part by impacting the vasculature of the brain.