International Journal of Statistics in Medical Research

A Double Truncated Binomial Model to Assess Psychiatric Health through Brief Psychiatric Rating Scale: When is Intervention Useful?

Alka Sabharwal, Babita Goyal, Vinit Singh — Thu, 11 Jan 2024 00:00:00 +0000

A double truncated binomial distribution model with ‘u’ classes truncated on left and ‘v’ classes truncated on right is introduced. Its characteristics, namely, generating functions; and the measures of skewness and kurtosis have been obtained. The unknown parameter has been estimated using the method of maximum likelihood and the method of moments. The confidence interval of the estimate has been obtained through Fisher’s information matrix.

The model is applied on cross sectional data obtained through Brief Psychiatric Rating Scale (BPRS) administered on a group of school going adolescent students; and the above-mentioned characteristics have been evaluated. An expert, on the basis of the BPRS score values, suggested an intervention program. The BPRS scores of the students who could be administered the intervention program lied in a range (which was above the lowest and below the highest possible values) suggested by the expert. Whereas the complete data suggested the average number of problem areas is four (which was not in consonance with the observations given by the expert), the double truncated model suggested the number of such areas as five which was consistent with the observations made by the expert. This establishes the usefulness of double truncated models in such scenarios.

Analysis of Wide Modified Rankin Score Dataset using Markov Chain Monte Carlo Simulation

Thu, 18 Jan 2024 00:00:00 +0000

Brain hemorrhage and strokes are serious medical conditions that can have devastating effects on a person's overall well-being and are influenced by several factors. We often encounter such scenarios specially in medical field where a single variable is associated with several other features. Visualizing such datasets with a higher number of features poses a challenge due to their complexity. Additionally, the presence of a strong correlation structure among the features makes it hard to determine the impactful variables with the usual statistical procedure. The present paper deals with analysing real life wide Modified Rankin Score dataset within a Bayesian framework using a logistic regression model by employing Markov chain Monte Carlo simulation. Latterly, multiple covariates in the model are subject to testing against zero in order to simplify the model by utilizing a model comparison tool based on Bayes Information Criterion.

Triglyceridemic Waist Phenotypes as Risk Factors for Type 2 Diabetes Mellitus: A Systematic Review and Meta-Analysis

Fiorella E. Zuzunaga-Montoya, Víctor Juan Vera-Ponce — Mon, 19 Feb 2024 00:00:00 +0000

Introduction: Triglyceride waist phenotypes, which combine high triglyceride levels and central obesity, have recently emerged as an area of interest in metabolic disease research.

Objective: To conduct a systematic review (SR) with meta-analysis to determine if triglyceride waist phenotypes are a risk factor for T2DM.

Materials: SR with meta-analysis of cohort studies. The search was conducted in four databases: PubMed/Medline, Scopus, Web of Science, and EMBASE. Participants were classified into four groups, based on triglyceride level and waist circumference (WC): 1) Normal WC and normalConduct triglyceride level (NWNT); 2) Normal WC and high triglyceride level (NWHT), 3) Altered WC and normal triglyceride level (EWNT) and 4) Altered WC and high triglyceride level (EWHT). For the meta-analysis, only studies whose measure of association were presented as Hazard ratio (HR) along with 95% confidence intervals (CI95%) were used.

Results: Compared to people with NWHT, a statistically significant association was found for those with NWHT (HR: 2.65; CI95% 1.77–3.95), EWNT (HR: 2.54; CI95% 2.05–3.16) and EWHT (HR: 4.41; CI95% 2.82–6.89).

Conclusions: There is a clear association between triglyceride waist phenotypes and diabetes, according to this SR and meta-analysis. Although central obesity and high triglyceride levels are associated with a higher risk of the aforementioned disease, their combination appears to pose an even greater risk. Therefore, in the clinical setting, it is important to consider this when assessing the risk of diabetes.

Adaptive Elastic Net on High-Dimensional Sparse Data with Multicollinearity: Application to Lipomatous Tumor Classification

Narumol Sudjai, Monthira Duangsaphon, Chandhanarat Chandhanayingyong — Fri, 29 Mar 2024 00:00:00 +0000

Predictive models can experience instabilities because of the combination of high-dimensional sparse data and multicollinearity problems. The adaptive Least Absolute Shrinkage and Selection Operator (adaptive Lasso) and adaptive elastic net were developed using the adaptive weight on penalty term. These adaptive weights are related to the power order of the estimators. Therefore, we concentrate on the power of adaptive weight on these penalty functions. This study purposed to compare the performances of the power of the adaptive Lasso and adaptive elastic net methods under high-dimensional sparse data with multicollinearity. Moreover, we compared the performances of the ridge, Lasso, elastic net, adaptive Lasso, and adaptive elastic net in terms of the mean of the predicted mean squared error (MPMSE) for the simulation study and the classification accuracy for a real-data application. The results of the simulation and the real-data application showed that the square root of the adaptive elastic net performed best on high-dimensional sparse data with multicollinearity.