---start---- epi review 2/17 those of you having problems with the past few days of class should look at his handwritten answers, which will be photocopied and placed in the library for your perusal. stuff to know for exam 8 questions will be on it. it's closed book and lasts 50 minutes. question in which you are told things about an endemic or epidemic and you have to a) make a flow chart. note: you should be able to generalize what you already know into a novel situation. this includes a new box you haven't seen, and a new arrow you haven't seen. the rules are the same. you shouldn't have a problem. there is also a given where some parameter values are given, and it should explain how to do part of it. you must label all compartments and arrows. it doesn't matter what letters you use. make a key. give numerical values - so you need to be able to calculate the parameters from the given info. put in units. b) write differential equations. derive formula for basic reproduction ratio, Ro. some of you, after doing last week's problems, realized that you were getting the same Ro over and over. this Ro in this problem is different. you are goign to have to apply the rules and figure it out. figure out formula for Nt critical minimum threshold density. LEARN ALL THE DEFINITIONS there will be one question on the lecture of evolution of virulence, so know basic principles. there is a question on how various control strategies affect different aspects of the basic and effective reproduction ratio.. note service notes on that lecture aren't good. (he says) a general question on eradication strategies - do not rely on note service - he says it is bad. two galligan quesions - one on decision analysis, one on sensitivity and specificity. this should be worth about 5% of the exam. -- the final exam will cover only this material from *after* the chicken stuff (outbreak investigation). note that that lecture was not about chickens, it was about outbreak investigation. he'll conjure up some answers to our other practice problems and have them put in the library by friday. --- he's going to give a quick overview of the entire course now. you have a disease, it's described, the first thing you do is draw flow chart by dividing host population into blocks based on infection status. usually we call these X, Y, Z but you can call them wahtever you want. these are the densities of animals in the box. then you draw arrrows in direction of animal movement. if you want to turn an epidemic into an endemic model, you add a birth rate coming into the first susceptible box, and death rates from each box. assume no disease mortality, and death rate equals birth rate, and population density remains constant. if animals spend 14 days in box Y, then delta the recovery rate is 1/14 per animal per day. the only loss rate that is peculiar is the one from susceptible to infectious box - because the rate of movement depends on rate of contact with Y group, and some probability constant. so BY is your rate of movement into Y from X. the transmission loss rate is this BY thing. differential equations: count the arrows. for endemic model, there are three arrows stuck to the X box - that means you'll have three things to the right of the = sign. arrows coming in are + things, going out are - things. dX/dt = uN - BYX - uX note that the birth rate is uN. N = X+Y+Z to make rate, look at name of rate and multiply times the box doing the losing. always write a key! dY/dt = BYX - dY - uY dZ/dt = dY - uZ why is the birth rate uN and the death loss from X is uX? because we assume that additions to population due to birth are attributable to all boxes, despite infection status. but the deaths only refer to deaths from X category. u = birth rate /animal/day and also death rate/animal/day B = transmission constant things in boxes XYZ = animals/unit of area making Ro basic repro ratio: steps: ask yourself what must be true for Y to increase. answer: dY/dt must be positive then: ask what must be true for dY/dt to be positive? answer: sum of all positive terms must be bigger than sum of all negative terms BXY/(d+u)Y --> BXY >(d+u)Y cancel Ys. substitute N for X since this is time zero. look for sum of positives over sum of negatives, replace X with N. the second that the first individual enters the population and infects someone, N becomes X. now you have the effective repro ratio -this is what you try to alter by vaccination and so forth. critical minimum threshold density - how to calculate: BN/d+u > 1 for epidemic to occur,right? figure out value of N.use cross multiplication. N > (d+u)/B so that's Nt. control things: BX/(d+u) vaccinatoin alters X test and remove alters d culling alters u and X quarantine alters d nutrition alters B -- rules: box is density of animals. arrow represents where animals go when they leave box. rest is common sense. think of where boxes come in order and what happens when animal leaves the box. repro ratio always pertains to Y follow rules for deriving it and you'll always get it right. when he says "derive the formula" do you include all the steps? sure. why not. it improves your chance of getting partial credit. recrudescent animals do affect Ro. dY/dt = BXY + BXR - dY - uY Ro = BXY + BXR all divided by dY + uY now you can't cancel the Ys. -- evolution of virulence: simple. tendency is to produce more virus or bacteria in a single host. that will happen unless it turns into a bad thing to do. it's good because the more microbes, the more shedding, the likelier transmission. when microbial load correlates with sickness, then you could increase morbidity such that transmission is impossible because infectious host is immobile and never meets new host. moribund host won't move around. but for a microbe that's transmitted by a vector like an arthropod or a handler, then causing host to be moribund isn't a disadvantage for the microbe. whenever transmission doesn't rely on movement of host - vectorborne or attendant borne transmission, water borne transmisison, etc - there is no benefit from reduced virulence. evolution will continue to push for increased production of pathogen and hosts get sick. so this kind of infecction tends to be more virulent. -- control strategies - vaccines - two kinds. one: protects host completely. other: doesn't protect host from infection, does protect from consequences of infection. often the second kind also reduces B, increases d, and reduces Ro. so in some circumstances you can use these as eradication tools, although normally you'd want the first kind. test and removal - usually this increases effective recovery rate d. culling - generally mass culling reduces X to below critical minimum threshold density treatment - like simple culling - reduces length of time infectious animal is infectious - increases d. eradication programs almost always use more than one strategy. as time goes on and there is more success, the mix of strategies will change. you start with cheap acceptable option that farmers will do. as you deal with fewer herds, you can be more draconian - in later stages, you go to depopulation of affected herds. originally, too many herds are affected and you can't afford to depopulate and it pisses people off. later, it's more acceptable b/c only a few herds are affected. also the PPV of tests declines with prevalence so you need more sensitive and specific tests. advice re: galligan - learn in generalities. they are multiple choice questions, and there are only two of them. ---end---