I went to the first Veterinary epidemiology class yesterday. It was really a hectic day.
We bought a Dohoo book; Veterinary Epidemiologic Research. It is a really nice, thick book of all Modern methods of data analysis techniques inside.
There’s another 3 students; Anouk, Hiestke and Pathija from Thai. We get to know about them that morning. Hietske is Phd student in public health at Julius Centre, starting off with case control study of MRSA in veals. Anouk and Pathija does not have basic in Veterinary epidemiology (looks like we are cleverer huh??) Pathija comes from Erasmus Rotterdam.
Mirjam taught us back about the study design that we have learned earlier in our Msc. Its basically the 4 study design, observational+experimental; Cohort, Cross Sectional, Case Control and RCT and how we apply it in Veterinary field. She used Andro and Hietske research example. Mirjam also taught about confounding and asked me about me. Huhuhu, I forgot the best example, but I make a smoking example and the need to make the causality inferences. Mirjam added, like having lighter in your pocket and if you do the data analysis, it is significant.
Hietske then gave the example smoking and asbestos, which is a nice example. Because confounding is related to the outcome and risk factors. So if you want to know about asbestos in people you must also check for smoking because both can cause lung problems. It is a confounder.
Later we learn about Experimental study design. Hietske is doing MRSA, which cause problems in human but not in animal. MRSA~Methicillin Resistant Staphylococcal Aureus is found in the piglets. Last time it was not known but now, it has been found in almost all animal. So Hietske said, the research questions are; if giving antibiotic reduce the problem and if hygiene protocol helps reduce the problem. Later, Mirjam asked Andro about his experiment. It was about if vaccination efficacy in AI birds. But his research is based on negotiation, so it is not a randomize trial. Mirjam said, people who wanted to enter the study might do it because their old protocol is not working. So there is danger of under or over estimation. While RCT need to be randomize. She also stressed about the introduction of diseases. If there is no introduction of AI, there will not be measure of vaccine efficacy at all. So people tend to be happy with what they have and do not want to change anything.
She also stressed about the outcome of the study. She said, before starting the project we need to have in our mind the outcome we want, easiest in AI is mortality. Probably if you have titre, its different and its difficult to do. So we need to think hard about it.
She gave us quiz and expect us to do some reading next week.
In the afternoon, the practical is exhaustive! I need to answer back the questions. Mirjam wanted to discuss on Friday.