ijsmr
|
|
Abstract: Cancer researchers are often interested in identifying biomarkers that are indicative of poor outcomes (prognostic biomarkers) or response to specific therapies (predictive biomarkers). In designing a biomarker study, the first statistical issue encountered is the sample size requirement for adequate detection of a biomarker effect. In biomarker studies, the desired effect size is typically larger than those targeted in therapeutic trials and the biomarker prevalence is rarely near the optimal 50%. In this article, we review sample size formulas that are routinely used in designing therapeutic trials. We then conduct simulation studies to evaluate the performances of these methods when applied to biomarker studies. In particular, we examine the impact that deviations from certain statistical assumptions (i.e., biomarker positive prevalence and effect size) have on statistical power and type I error. Our simulation results indicate that when the true biomarker prevalence is close to 50%, all methods perform well in terms of power regardless of the magnitude of the targeted biomarker effect. However, when the biomarker positive prevalence rate deviates from 50%, the empirical power based on some existing methods may be substantially different from the nominal power, and this discrepancy becomes more profound for large biomarker effects. The type I error is maintained close to the 5% nominal level in all scenarios we investigate, although there is a slight inflation as the targeted effect size increases. Based on these results, we delineate the range of parameters within which the use of some sample size methods may be sufficiently robust. Keywords: Sample size methods, biomarker study, prognostic biomarker, predictive biomarker, survival data. |
|
|
Abstract: Purpose: To determine the risk factors for the development of serious adverse events (AEs) in black adult patients on combination antiretroviral therapy (cART). Keywords: Adverse events, cohort event monitoring, combination antiretroviral therapy, pharmacovigilance, risk factors, South Africa. |
|
|
Abstract: In this study, we adapt a Cox-based model for recurrent events; the Prentice, Williams and Peterson Total -Time (PWP-TT) that has largely, been used under the assumption of non-informative censoring and evaluate it under an informative censoring setting. Empirical evaluation was undertaken with the aid of the semi-parametric framework for recurrent events suggested by Huang [1] and implemented in R Studio software. For validation we used data from a typical HIV care setting in Kenya. Of the three models under consideration; the standard Cox Model had gender hazard ratio (HR) of 0.66 (p-value=0.165), Andersen-Gill had HR 0.46 (with borderline p-value=0.054) and extended PWP TT had HR 0.22 (p-value=0.006). The PWP-TT model performed better as compared to other models under informative setting. In terms of risk factors under informative setting, LTFU due to stigma; gender [base=Male] had HR 0.544 (p-value =0.002), age [base is < 37] had HR 0.772 (p-value=0.008), ART regimen [base= First line] had HR 0.518 (p-value= 0.233) and differentiated care model (Base=not on DCM) had HR 0.77(p-value=0.036). In conclusion, in spite of the multiple interventions designed to address incidences of LTFU among HIV patients, within-person cases of LTFU are usually common and recurrent in nature, with the present likelihood of a person getting LTFU influenced by previous occurrences and therefore informative censoring should be checked. Keywords: Recurrent events, Loss to follow-up, HIV, Prentice, Williams and Peterson Gap-Time, Informative censoring. |
|
|
Abstract: Often in epidemiological research, introducing a stratified Cox model can account for the existence of interactions of some inherent factors with some major/noticeable factors. This paper aims at modelling correlated variables in infant mortality with the existence of some inherent factors affecting the infant survival function. A Stratified Cox model is proposed with a view to taking care of multi-factor-level that has interactions with others. This, however, is used as a tool to model infant mortality data from Nigeria Demographic and Health Survey (NDHS) with g-level-factor (Tetanus, Polio and Breastfeeding) having correlations with main factors (Sex, infant Size and Mode of Delivery). Asymptotic properties of partial likelihood estimators of regression parameters are also studied via simulation. The proposed models are tested via data and it shows good fit and performs differently depending on the levels of the interaction of the strata variable Z*. An evidence that the baseline hazard functions and regression coefficients are not the same from stratum to stratum provides a gain in information as against the usage of the Cox model. Simulation result shows that the present method produces better estimates in terms of bias, lower standard errors, and or mean square errors. Keywords: Stratified Cox, Semiparametric model, infant mortality, multifactor-level, confounding variables. |
|
|
Abstract: In analyzing longitudinal data the correlations between responses obtained from same individual need to be taken into account. Various models can be used to handle such correlations. This article focuses on the application of transition modeling using Bayesian approach for analyzing longitudinal binary data. For Bayesian estimation asymmetric loss functions, such as, linear exponential (LINEX) and modified linear exponential (MLINEX) loss function and Tierney and Kadnae (T.K.) approximation has been used. Comparison is made using Bayes factor and Bayesian approach under LINEX loss function can be suggested to estimate the parameters of transition model. Keywords: Bayesian approach, Bayes Factor (BF), Linear exponential (LINEX), Longitudinal data, Markov model, Modified linear exponential (MLINEX). |