With the development of hygiene systems and rising awareness of health in the world wide, the influence of infectious diseases which are stubborn in the past is diminishing. However, the quick spreading of COVID-19 tells everyone, there is still necessity to learn epidemiology. Even if COVID-19 disappears, it is unsure whether there will be new terrible pathogens. Therefore, we need to develop epidemiology further to prevent for future threats.
In week 1, we learnt a generalization of epidemiology and public health nowadays. Although we are living in a well sanitary environment, it’s necessary to explore more about epidemiology, especially in the age of COVID-19. In order to solve public health problems, we can analyze problems from individual perspective and population perspective. In the assignment this week, we compare and analyze these methods dialectically. This assignment helps us know and master analysis methods, so that we can apply them to solve problems. In the TA session, we learnt basic operations of R related and means of searching for suitable databases.
In week 2, we learnt relevant generalizations about incidence and prevalence as the first part of session. Professor compared a few groups of generalizations to help us distinguish them. In the second part, we learnt exploratory analysis and summary statistics. Professor introduces different diagrams for us, explaining the application situations. As a important type of database, professor first introduced the survival data and its relevant generalizations. Then he summarized points for estimating survival functions. In the TA session, we learnt categorical and numerical variable, including their center and spread generalizations. In practice, teaching assistant introduce different types of tables for presenting data. Additionally, he showed us method of analyzing data simply.
In week 3, we learnt two sample comparison and regression analysis. Professor derived formula for analyzing variances and taught us the method of calculating regression equation. Among all parameters of the two-analysis method, p-value is one of the most important parameters. It can exam whether data is statistically significant. In the TA session, we learnt analysis methods for survival analysis and two sample comparisons corresponding to professor session. Paired t-test is used to compare two related groups of samples. Classical t-test is applied for equal variance, while Welch t-statistics is applied for unequal variance. We need preliminary test before t-test. Finally, we need p value to decide whether groups are significantly different. For the second assignment, we were acquired to select a topic for our report and find out questions hidden in our topic database. Our group finally chose the risk factors of cervical cancer as our topic, and list questions we interested in.
In week 4, we learnt diseases and prevention. The natural history of diseases is composed of induction, incubation and latency. Population screening is used to classify whether people are effected with diseases. However, screening program requires suitable disease, suitable test, suitable program and good use of resources. Lead time means the duration between detectable, preclinical stage and clinical stage. Survival time is often longer than lead time. Screening test is reliable and validity. Reliability means getting same result at the same conditions. Validity means sensitivity and specificity. After interpreting test results, we can get predictive value for prediction. In the TA session, we learnt linear regression analysis. Compared with other analysis method, it can predict further value on the response variables. Model assessment includes R-Squared, T and F statistic with p value, AIC and BIC, and we need different suitable model to make diagnostic decisions.
In week 5, we learnt intervention and Cohort studies. Analytic study designs are experimental or observational. Intervention trails are experimental, while Cohort, case-control and cross-sectional studies are observational. Intervention studies include therapeutic trails and preventive trails. Cohort studies often start with a population at risk and measure characteristics at baseline. Follow-up the population over time needs surveillance or re-examination. Finally, Cohort studies needs to compare event rates in people with and without characteristics of interest. In the TA session, we learnt advanced topics in linear regression and generalized linear regression.
In week 6, we present our findings in class. Teaching assistant gave us useful advices and we finish our final report successfully.
In the project, I learnt lot about statistics analysis for epidemiology. At the same time, I realize that the methods I learnt is not only for epidemiology, but also for other medicine fields. After project, I will apply knowledge to my future work, continue studying new techniques for further exploration.
