Search

Multiple imputation (MI) models can be improved by including auxiliary covariates (AC), but their performance in high-dimensional data is not well understood. We aimed to develop and compare high-dimensional MI (HDMI) approaches using structured and natural language processing (NLP)-derived AC in studies with partially observed confounders. We conducted a plasmode simulation study using data from opioid vs. non-steroidal anti-inflammatory drug (NSAID) initiators (X) with observed serum creatinine labs (Z2) and time-to-acute kidney injury as outcome. We simulated 100 cohorts with a null treatment effect, including X, Z2, atrial fibrillation (U), and 13 other investigator-derived confounders (Z1) in the outcome generation. We then imposed missingness (MZ2) on 50% of Z2 measurements as a function of Z2 and U and created different HDMI candidate AC using structured and NLP-derived features. We mimicked scenarios where U was unobserved by omitting it from all AC candidate sets. Using LASSO, we data-adaptively selected HDMI covariates associated with Z2 and MZ2 for MI, and with U to include in propensity score models. The treatment effect was estimated following propensity score matching in MI datasets and we benchmarked HDMI approaches against a baseline imputation and complete case analysis with Z1 only. HDMI using claims data showed the lowest bias (0.072). Combining claims and sentence embeddings led to an improvement in the efficiency displaying the lowest root-mean-squared-error (0.173) and coverage (94%). NLP-derived AC alone did not perform better than baseline MI. HDMI approaches may decrease bias in studies with partially observed confounders where missingness depends on unobserved factors.

Research over the past few decades has assumed the richness (number of species) to be the primary driver of the biodiversity and ecosystem function (BEF) relationship. However, biodiversity is multi-dimensional, and richness alone does not capture all its attributes. Diversity-Interactions modelling is a regression-based framework that models the biodiversity and ecosystem function relationship by incorporating species-specific effects along with the interactions between species. The species interactions in a Diversity-Interactions model can take several different forms ranging in complexity from a single interaction term (assuming all pairs of species interact in the same way) to many interaction terms (e.g. assuming a separate interaction for all pairs of species). The specification of the interactions may also include a non-linear parameter ({\theta}) as an exponent to the product of the species proportions to allow for non-linear relationship between the response and species interactions, giving rise to Generalized Diversity-Interactions modelling. The structure of the interaction terms describes the underlying biological processes and thus the selection of a correct structure is important. The inclusion of {\theta} introduces complexity to the model selection process: we could (a) select the interaction structure first by assuming {\theta}=1 and then estimate {\theta}, or (b) estimate {\theta} first and then select the appropriate interaction structure by fixing {\theta}, or (c) test for {\theta} and its inclusion for each interaction structure. It is also unknown whether {\theta} is robust to changes in the underlying interaction structure. Using a simulation study, we test the robustness of {\theta} and compare multiple model selection approaches to identify an optimal and computationally efficient model selection procedure for Generalized Diversity-Interactions models., Comment: For code to reproduce the analysis see https://github.com/rishvish/Theta-Simulation-Study

BEF studies aim to understand how ecosystems respond to a gradient of species diversity. Diversity-Interactions (DI) models are suitable for analysing the BEF relationship. These models relate an ecosystem function response of a community to the identity of the species in the community, their evenness (proportions) and interactions. The number of species in the community (richness) is also implicitly modelled through this approach. It is common in BEF studies to model an ecosystem function as a function of richness; while this can uncover trends in the BEF relationship, by definition, species diversity is much broader than richness alone, and important patterns in the BEF relationship may remain hidden. In this paper, we introduce the DImodels R package for implementing DI models. We also compare DI models to traditional modelling approaches to highlight the advantages of using a multi-dimensional definition of species diversity. We show that using DI models can lead to considerably improved model fit over other methods; it does this by incorporating variation due to the multiple facets of species diversity. Predicting from a DI model is not limited to the study design points, the model can extrapolate to predict for any species composition and proportions (assuming there is sufficient coverage of this space in the study design). Expressing the BEF relationship as a function of richness alone can be useful to capture overall trends. However, collapsing the multiple dimensions of species diversity to a single dimension (such as richness) can result in valuable ecological information being lost. DI modelling provides a framework to test the multiple components of species diversity in the BEF relationship. It facilitates uncovering a deeper ecological understanding of the BEF relationship and can lead to enhanced inference.