Misconceptions About Nested Studies, Matched Analysis, Significancy, Power, Normality and Multivariate Regression Analysis

which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Yan et al. have published a useful paper about thyroid dysfunction (TD) in cases with hepatitis C by treatment with interferon (1). There are some comments which may explicit benefits of this paper in more details. When we report a methodological or statistical issue it is necessary to mention important details which are informative and necessary to know. We should report such details as brief as possible. For example we must mention that type of study is nested case-control or (nested) case-cohort. We should also mention each odds ratio (OR) is crude or adjusted. This study has been nested case-cohort and not nested case-control. In a case-cohort study the comparison group, called a sub-cohort, is selected at random from the initial cohort at baseline regardless of the outcome (2). Nested case-control or case-cohort studies are good approach specifically for evaluating markers which are unknown at the beginning of the cohort or expensive tests. However, when we select cases from the baseline cohort it is case-cohort and when we select simultaneous controls which have similar exposure time with cases in cohort (select controls as the cases are occurring) it is called as nested case-control. Explanation of authors shows that it is a case-cohort study. However, authors declared that it is a nested case-control. It is specifically of importance because when we select our controls among patients of base cohort, there is probability of selecting cases as controls. Since, the prevalence of caseness (having TD) is 11%, there is probability of presence of at least six cases among controls which is considerable and may dilute associations and confound the results. I am interested to know how much is hepatitis C prevalence in southwest of China? They treat approximately 2500 in-patients and 75 000 outpatients per year. Data collection has been done between 2004 and 2011. It seems that there are more than 674 cases under treatment for HCV during this period. How they have selected some of them? In addition , I did not understand that 82 cases are excluded because they either did not meet eligibility criteria or they refused to participate. It seems that they did not have eligibility criteria. However, they have mentioned that participation rate was 87.8% which shows some of cases refused from participation. If the latter is correct, authors …