Vol 8 No 3 (2022)

Original Article(s)

  • XML | PDF | downloads: 160 | views: 406 | pages: 234-247

    Background: Birth weight and gestational age are important determinants of an infant’s survival and future development. The main objective of this research was to identify the determinants of birth weight and gestational age simultaneously based on Ethiopia's demographic health survey in 2016 which implemented in statistical package R.
    Methods: Cross-sectional study design was used from Ethiopia's demographic health survey in 2016. The bi-variate linear regression model was used to identify factors of birth weight and gestational age simultaneously which had small standard errors as compared to a separate model.
    Results: Bi-variate models of birth weight and gestational age determined the effect of predictors. Therefore, the model shows that the number of tetanus injections before pregnancy, educational level of husband, desire for more children, drink alcohol, and region are statistically significant at 5% level of significance for gestational age in Ethiopia. Similarly, the size of a child, preferred waiting time for birth, the number of tetanus injections before pregnancy was statistically associated with birth weight at 5% level of significance.
    Conclusion: From our finding, we concluded that the number of tetanus injections before pregnancy, educational level of husband, desire for more children, drink alcohol, size of child, preferred waiting time for birth and region were significant predictors of birth weight and gestational age simultaneously at 5% level of significance. Hence, special care should be given to the pregnant during antenatal care for minimizing the risk of low birth weight and small gestational age.

  • XML | PDF | downloads: 110 | views: 125 | pages: 248-257

    Objectives: Cumulative incidence function (CIF) measures the survival time of a particular hazard in the presence of others, while cause-specific (CS) one ignores the competing risks. The present study aimed to fit CIF and CS function for brain stroke (BS) patients and compare the results by the cause of death.

    Materials and method: In the study, 332 patients with the definitive diagnosis of BS were followed up for 10 years, and their mortality status due to BS or other causes was evaluated. In addition, significance tests and parameters were estimated by using STATA 14 software by considering the CS and CIF-based regression model.

    Results: Based on the results of CIF and CS analyses concerning the variables with similar significance, the hazard ratio of age at diagnosis (68-59 years (91%,2.61), ≥76 years (2.14,3.03) during diagnosis enhanced in the death for other causes, while an increase was observed in this ratio for sex (38%,2.35%), as well as the history of heart disease (44%,47%) and blood pressure (57%,64%) regarding BS-caused death, respectively. Regarding the significant variables, the correlation strength of CIF model was more in the BS-caused death by considering p-value, while CS one had stronger correlation in the death due to other causes.

    Conclusion: The estimation of CIF analysis, along with CS one for the competing risks, is suggested to provide more precise information about patients’ status in order to support adopted clinical decisions when aiming at assessing health related to a specific cause economically and determining the probability of occurring an intended event among other causes.

  • XML | PDF | downloads: 68 | PDF | views: 107 | pages: 258-268


    In numerous practical applications, data from neighbouring small areas present spatial correlation. More recently, an extension of the Fay–Herriot model through the spatial (exponential) has been considered. This spatial area-level model like the fundamental area-level model (was first suggested by Fay III and Herriot ) has a powerful assumption of known sampling variance . Several methods have been suggested for smoothing of sampling variance and there is no unique method for sampling variance estimation, more studies need.


    This research examines four techniques for sampling variance estimates including of Direct , Probability Distribution, Bayes and Bootstrap methods. We used households'  food expenditures (HFE) data 2013 and other socio-economic ancillary data to fit the read model and at last conduct a simulation study based on this data to compare the effects of four variance estimation methods on precision of small area estimates.


    The best model on real data showed that the lowest and the highest HFE belonged to Pishva district (in Tehran province) with 26,707 thousand rials (TRs) and Omidiyeh (in Khouzestan province) with 101,961 TRs, respectively. Accordingly on simulation study, the probability distribution and direct methods, respectively and approximately had the smallest and the highest Root Average Mean Square Errors  (RAMSE)  for all conditions.


    The results showed the best fitting with Direct method in real data and best precision with Probability Distribution method in simulation study.

  • XML | PDF | downloads: 70 | views: 99 | pages: 269-280

    BACKGROUND: Colorectal cancer is the most common cause of cancer mortality in Iran. There are differences in the etiology, clinical behavior and pathological features in cancer of the colon versus the rectum. The aim of this study was to evaluate the factors related to survival and cure probability of patients with colon and rectal cancer using a semi-parametric non-mixture cure rate model.

    METHODS: This retrospective cohort study was conducted on 311 patients, with colorectal cancer. Data of all patients with colon and rectum malignances who underwent the first treatment in Omid Hospital, Mashhad, between 2006 and 2011 were gathered through medical records. Patients were followed-up for 9 years until September 2020. Semi-parametric non-mixture cure model was implemented using miCoPTCM package in the R software.

    RESULTS: The mean survival time was 2973.94 days (95% confidence interval [CI], 2694.96 to 3252.93). The 5-year survival rates for colon and rectal cancer patients were 0.54 (%95 CI:(0.45, 0.61)) and 0.57 (%95 CI:(0.48,0.65)), respectively. The proportion of cured patients for colon cancer was 44.0%, and for rectal cancer was 40.0%.Age and stage of the disease were determined as factors related to survival and cure fraction of both colon and rectal cancers. In addition, ethnicity and type of treatment were distinguished as factors related to survival and cure fraction of rectal cancer. While the history of drug abuse only increased the hazard of death in colon cancer patients; overweight as a protective variable increased the survival and cure fraction of rectal cancer patients.

    CONCLUSION: Because the factors associated with colorectal cancer are not necessarily equal to the risk factors for colon and rectal cancer, it is recommended to obtain more accurate and valid results in the survival analysis of colorectal cancer patients, the colon and rectum should be considered separately. It is also appropriate to use cure rate models when there is a cure fraction in the data.

  • XML | PDF | downloads: 53 | PDF | views: 90 | pages: 281-294


    Introduction: Haplotype analysis allows higher resolution analysis in genetic association studies and is used as a reference panel for genotype imputation in genome-wide association studies. Haplotypes estimates from genotypes among unrelated individuals, but misclassification of the haplotype reconstruction will directly affect the accuracy of the results.


    Methods: This study proposes a novel statistical method Gibbs sampler algorithm to estimate haplotype frequency and quantify the influence of misclassification bias of the estimate haplotype. The performance of the algorithm is evaluated on simulated datasets assuming that linkage phase unknown. The simulation used different minor allele frequencies at each single nucleotide polymorphisms (SNPs) and different linkage-disequilibrium between the SNPs.


    Results: The Gibbs sampler algorithm presents higher accuracy among over seven SNPs or less, validated, and deals with missing genotype compared to previous related statistical approaches. Misclassification of estimated haplotypes leads to non-difference bias in exposure and affects haplotype estimates in haplotype analysis. The observed odds ratio underestimates the association between haplotype and phenotype by 36% to 99%.


    Conclusion: The Gibbs sampler algorithm provides higher accuracy and robust effectiveness performance, handles missing genotypes, and provides uncertain probabilities of haplotype frequencies. The misclassification bias of the estimate haplotype underestimates the genetic association by more than forty percent.

  • XML | PDF | downloads: 37 | views: 115 | pages: 295-303

    Introduction: Coordinate-based meta-analysis (CBMA) is a standard method for integrating brain functional patterns in neuroimaging studies. CBMA aims to identify convergency in activated brain regions across studies using coordinates of the peak activation (foci). Here, we aimed to introduce a new application of the Gibbs models for the meta-regression of the neuroimaging studies.

    Methods: We used a dataset acquired from 31 studies by previous work. For each study as well as foci, study features such as SD duration and the average age were extracted. Two widely Gibbs models, Area-interaction and Geyer saturation were fitted on the foci. These models can quantify and test evidence for clusters in foci using an interaction parameter. We included study features in the models to identify their contribution to foci distribution and hence determine sources of the heterogeneity.

    Results: Our results revealed that latent study-specific features have a moderate contribution to the heterogeneity of foci distribution. However, the effect of age and SD duration was not significant (p<0.001). Additionally, the estimated interaction parameter was 1.34 (p<0.001) which denotes strong evidence of clusters in foci.

    Conclusions: Overall, this study highlighted the role of the interaction parameter in CBMA. The results of this work suggest that Gibbs models can be considered as a promising tool for neuroimaging meta-analysis.

  • XML | PDF | downloads: 59 | PDF | views: 110 | pages: 304-313

    Introduction: Recurrent event data are common in many longitudinal studies. Often, a terminating event such as death can be correlated with the recurrent event process. A shared frailty model applied to account for the association between recurrent and terminal events. In some situations, a fraction of subjects experience neither recurrent events nor death; these subjects are cured.

    Methods: In this paper, we discussed the Bayesian approach of a joint frailty model for recurrent and terminal events in the presence of cure fraction. We compared estimates of parameters in the Frequentist and Bayesian approaches via simulation studies in various sample sizes; we applied the joint frailty model in the presence of cure fraction with Frequentist and Bayesian approaches for breast cancer.

    Results: In small sample size Bayesian approach compared to Frequentist approach had a smaller standard error and mean square error, and the coverage probabilities close to nominal level of 95%. Also, in Bayesian approach, the sampling means of the estimated standard errors were close to the empirical standard error.

    Conclusion: The simulation results suggested that when sample size was small, the use of Bayesian joint frailty model in the presence of cure fraction led to more efficiency in parameter estimation and statistical inference

  • XML | PDF | downloads: 74 | views: 191 | pages: 314-327

    Introduction: Disorders can often lead to physical illness and suffering along with associated functional disability which hampers the overall well-being of a person. Consequently, it can lead to loss of productivity at the workplace, absenteeism, and social isolation which eventually affects the individual and the society. Researchers have found a crucial association between childhood traumatic experiences with developing anxiety or panic disorder.

    Methods: The purpose of this study is to do a logistic regression on Add health data to examine whether a history of childhood abuse tends to lead to a diagnosis of anxiety or panic disorder in later life. Add health dataset was used for the analysis. Additionally, medical conditions such as ADHD, PTSD or socio-economic conditions, and addiction were also investigated for their possible contribution to developing anxiety or panic disorder.

    Results: 49.4 % of respondents reported having faced either physical, emotional, or sexual abuse before the age of 18. Among the total respondents, 12.5 % reported having been diagnosed with anxiety disorder and among these individuals, 25.9 % reported having experienced physical abuse, 64.6 % faced emotional abuse, and 10.3 % said they faced sexual abuse earlier in their life. Results from logistic regression showed all types of history of abuse have a significant effect on anxiety disorder. Additionally, the number of abuses experienced also increased the odds of developing an anxiety disorder. Women had higher odds of having such a disorder if they faced maltreatment in their childhood. Moreover, having PTSD and Depression also increased the odds of anxiety substantially.

    Conclusion: Childhood emotional abuse was found to be a more significant contributor to anxiety or panic disorder than other types of abuse. Any kind of childhood abuse experience seemed to have a greater effect on the female portion of the respondents in comparison to the males.

  • XML | PDF | downloads: 59 | views: 109 | pages: 328-339

    Introduction: Anemia is a major public health concern that affects more than fifty-six million women globally. In pregnancy, it is hemoglobin concentration of less than eleven grams per deciliter in venous blood and has significant adverse health consequences, on pregnant women. The major objective of this study is to look into anemia and the factors that contribute to it in the participant.

       Methods: The data in this research came from the 2016 Ethiopian Demographic Health survey. This analysis included 1053 women who are pregnant in total. Partial proportional odd model was used in the analysis of risk factors of anemia status.    

        Results: Among the total, 1053 pregnant women involved in this study: 32, 214, 395 were severe, moderate, and mild anemic respectively. The highest proportions of severe anemic were observed in Somali whereas the smallest was in Tigray. The chance of non-anemic increased by 9.3% for those taking iron pills. Somali pregnant women were 3.66% more likely to get severe anemic. However, pregnant women from a rural area and richest household were 27.1 %, 6.9 percent less likely to be non-anemic; consequently. Being non-anemic was increase with education levels [primary: (AMPE4 = 0.032, P ≤ 5%), Secondary: (AMPE4 = 0.069, P ≤ 5%) and higher: (AMPE4 = 0.176, P ≤ 5%)]

       Conclusion: Finally, the author concludes that education status, iron intake, wealth index, residence, parity, and region have been identified as prognostic factors of anemia status in pregnant women aged 15 to 49. As a result, action targeting these predictor variables is needed to enhance the anemia condition of pregnancies in Ethiopia.