Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userModelling disease progression with multi-level electronic health records data and informative observation times: an application to treating iron deficiency anaemia in primary care of the UK
151
0
0.0
(
0
)
Added by
Li Su
Publication date
2021
fields
Mathematical Statistics
and research's language is
English
Ask ChatGPT about the research
No Arabic abstract
Modelling disease progression of iron deficiency anaemia (IDA) following oral iron supplement prescriptions is a prerequisite for evaluating the cost-effectiveness of oral iron supplements. Electronic health records (EHRs) from the Clinical Practice Research Datalink (CPRD) provide rich longitudinal data on IDA disease progression in patients registered with 663 General Practitioner (GP) practices in the UK, but they also create challenges in statistical analyses. First, the CPRD data are clustered at multi-levels (i.e., GP practices and patients), but their large volume makes it computationally difficult to implement estimation of standard random effects models for multi-level data. Second, observation times in the CPRD data are irregular and could be informative about the disease progression. For example, shorter/longer gap times between GP visits could be associated with deteriorating/improving IDA. Existing methods to address informative observation times are mostly based on complex joint models, which adds more computational burden. To tackle these challenges, we develop a computationally efficient approach to modelling disease progression with EHRs data while accounting for variability at multi-level clusters and informative observation times. We apply the proposed method to the CPRD data to investigate IDA improvement and treatment intolerance following oral iron prescriptions in primary care of the UK.
rate research
Read More
Analyzing electronic health records (EHR) poses significant challenges because often few samples are available describing a patients health and, when available, their information content is highly diverse. The problem we consider is how to integrate sparsely sampled longitudinal data, missing measurements informative of the underlying health status and fixed demographic information to produce estimated survival distributions updated through a patients follow up. We propose a nonparametric probabilistic model that generates survival trajectories from an ensemble of Bayesian trees that learns variable interactions over time without specifying beforehand the longitudinal process. We show performance improvements on Primary Biliary Cirrhosis patient data.
We develop a new methodology for spatial regression of aggregated outputs on multi-resolution covariates. Such problems often occur with spatial data, for example in crop yield prediction, where the output is spatially-aggregated over an area and the covariates may be observed at multiple resolutions. Building upon previous work on aggregated output regression, we propose a regression framework to synthesise the effects of the covariates at different resolutions on the output and provide uncertainty estimation. We show that, for a crop yield prediction problem, our approach is more scalable, via variational inference, than existing multi-resolution regression models. We also show that our framework yields good predictive performance, compared to existing multi-resolution crop yield models, whilst being able to provide estimation of the underlying spatial effects.
Although increasingly used as a data resource for assembling cohorts, electronic health records (EHRs) pose many analytic challenges. In particular, a patients health status influences when and what data are recorded, generating sampling bias in the collected data. In this paper, we consider recurrent event analysis using EHR data. Conventional regression methods for event risk analysis usually require the values of covariates to be observed throughout the follow-up period. In EHR databases, time-dependent covariates are intermittently measured during clinical visits, and the timing of these visits is informative in the sense that it depends on the disease course. Simple methods, such as the last-observation-carried-forward approach, can lead to biased estimation. On the other hand, complex joint models require additional assumptions on the covariate process and cannot be easily extended to handle multiple longitudinal predictors. By incorporating sampling weights derived from estimating the observation time process, we develop a novel estimation procedure based on inverse-rate-weighting and kernel-smoothing for the semiparametric proportional rate model of recurrent events. The proposed methods do not require model specifications for the covariate processes and can easily handle multiple time-dependent covariates. Our methods are applied to a kidney transplant study for illustration.
This paper aims to enhance our understanding of substantive questions regarding self-reported happiness and well-being through the specification and use of multi-level models. To date, there have been numerous quantitative research studies of the happiness of individuals, based on single-level regression models, where typically a happiness index is related to a set of explanatory variables. There are also several single-level studies comparing aggregate happiness levels between countries. Nevertheless, there have been very few studies that attempt to simultaneously take into account variations in happiness and well-being at several different levels, such as individual, household, and area. Here, multilevel models are used with data from the British Household Panel Survey to assess the nature and extent of variations in happiness and well-being to determine the relative importance of the area (district, region), household and individual characteristics on these outcomes. Moreover, having taken into account the characteristics at these different levels in the multilevel models, the paper shows how it is possible to identify any areas that are associated with especially positive or negative feelings of happiness and well-being.
Voluntary medical male circumcision (VMMC) reduces the risk of male HIV acquisition by 60%. Programmes to provide male circumcision (MC) to prevent HIV infection have been introduced in sub-Saharan African countries with high HIV burden. While large-scale provision of MMC is recent, traditional MC has long been conducted as part of male coming-of-age practices. How and at what age traditional MC occurs varies by ethnic groups within countries. Accurate estimates of MC coverage by age and type of circumcision (traditional or medical) over time at sub-national levels are essential for planning and delivering VMMCs to meet targets and evaluating their impacts on HIV incidence. In this paper, we developed a Bayesian competing risks time-to-event model to produce region-age-time-type specific probabilities and coverage of MC with probabilistic uncertainty. The model jointly synthesises data from household surveys and health system data on the number of VMMCs conducted. We demonstrated the model using data from five household surveys and VMMC programme data to produce estimates of MC coverage for 52 districts in South Africa between 2008 and 2019. Nationally in 2008, 24.1% (CI: 23.4-24.8%) of men aged 15-49 were traditionally circumcised and 19.4% (CI: 18.9-20.0%) were medically circumcised. Between 2008 and 2019, five million VMMCs were conducted, and MC coverage among men aged 15-49 increased to 64.0% (CI: 63.2-64.9%) and medical MC coverage to 42% (CI: 41.3-43.0%). MC coverage varied widely across districts, ranging from 13.4-86.3%. The average age of traditional MC ranged between 13 to 19 years, depending on local cultural practices.
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
