This is the second post dedicated to on-line teaching. The subject is epidemics. Feel free to comment below!
The image of via Latina (Roma), with its roman tombs, bears no relationship with the post’s subject and has been added as a visual break.
On-line teaching #2ultima modifica: 2020-03-10T09:30:46+01:00da
This is a question I received by e-mail; I copy it here so that the discussion is made available to all students.
I have a couple of questions about the S-C-I model:
1) Why do we consider “Immune” even the people who died from the disease? Should not they be removed from the population poll?
2) Why the S-C-I model is considered irreversible, if some disease can have a reactivation, even if rare, like EBV and HSV?
My answers:
1) The scope of the Reed-Frost model is to predict the time course of epidemics, and especially of new cases. The statistical formula used requires to distinguish the sub-population that may become affected and thus may propagate the disease from the sub-population that cannot become affected and propagate the disease. In this sense, immune and dead belong to the same group, as they are both unable to propagate further the disease. If the letality of the disease is known, it is easy to split the sub-population I into two groups, which we may call D(eceased) and H(ealed), and if you like I can easily modify the model accordingly. Indeed in the study of historical epidemics one usually refers to the death registers, which are held by parishes since the XVI century, and the dynamics of the epidemics is analyzed from this information alone, which parallels the I(mmune) parameter of the Reed-Frost model.
2) Reinfections are usually uncommon and do not warrant the effort, except in some diseases (e.g. malaria). The Reed-Frost model is statistically sound and was meant to demonstrate some effects of epidemics and of public health measures, using the limited computation instruments commonly available at the time (essentially rulers and logarithmic tables). We have nowadays much more complex models that may be computationally demanding, but their scope is beyond that of our course, and beyond my personal expertise. The scope of our elective is to demonstrate that the course of epidemics changes rapidly and significantly the incidence of the disease, and hence the parameters of our statistical reasoning (e.g. the parameters in Bayes’ formula).
I received a question regarding the registration of attendance. This was added as an afterthought to the material already available on the web site, and might have problems. However:
1) at present there are some 50 registrations for porphyrias and 20 for epidemics; thus the program works.
2) The message of confirmation that appears at the end of the page is bomb-proof: if the page says that your number has been registered, you may feel sure that this is the case.
I received a question about the time window to participate to attend the on-line lectures: is option 1 or 2 correct below?
1. We can expect all future lectures to be available the day before they are scheduled on the timetable and we have to complete them by the day/time the lecture was set to finish.
2. We can expect all future lectures to be available at the day/time the lecture is scheduled on the timetable and we have until the end of the day to complete it.
My answer:
the lectures are available on the web site with no time limitations. I am modifying them in order to make them more suitable for on-line teaching, as they were originally meant to accompany the conventional lecturing activity. However I shall activate the attendance registration only from when the lecture is scheduled on our timetable, in order to maintain the logical progression of the course. Once attendance registration is activated there will be no time limits, and you can attend a week later, if you like, and be registered. There would be no point in setting a time limit in the access to the contents, as you need the website also in preparation for the exam. I shall close attendance registration at the end of the semester (may 31st).