-
outbreak
- a number of cases significantly higher than background expected number of cases in a PARTICULAR AREA over a given PERIOD OF TIME
- if you've never see the disease, expected is 0
-
steps in investigating an outbreak
- 1. verify the diagnosis
- 2. define a case (definition) (line-listing a table of cases. Can have ORs and ANDs)
- 3. determine magnitude of problem (establish existence of outbreak)
- 4. Describe temporal pattern of new cases (epidemic curve). Helps understand time trend, outliers, magnitude, mode of spread, period of exposure
- 5. describe the spatial pattern of new cases (Map!) - increasing frequency or trend in relation to location. Look for direction of spread or clusters
- 6. describe the animal pattern of new cases (what do they have in common? what is attack rate for each characteristics?) Can be feed-specific AR, barn-specific AR, species, age etc
- 7. analysis of outbreak data (compare AR. Calculate AR difference = AR exposed - AR not exposed)
- 8. Formulate a working hypothesis (remove what you think is infecting them)
- 9. clinical and epidemiology follow-up (were you right?
- 10. Report the findings
-
case definition
- standard set of criteria used in outbreak investigation to decide if animal is infected with disease of interest or not (NOT the same as clinical diagnosis)
- includes clinical criteria and describes a case in terms of ANIMAL, PLACE, TIME
- could be suspect, presumptive positive, confirmed positive cases
-
how to determine magnitude of outbreak
- count cases - search for missing or hidden, rule out other reasons for increase (like new lab procedure or reporting, change in population, mass hysteria)
- determine population at risk (susceptibles)
- determine if observed cases actually exceeds expected number.
- CALCULATE ATTACK RATE (AR)
-
Attack Rate
- AR = (number of new cases since onset) / (total number of animals at risk at onset of outbreak) x 100
- At risk can be everyone for a new disease, only not vaccinate, or even everyone who ate the specific food
- sick/total x 100 = %
-
epidemic curve
- visual display of onset of illness among cases over time in outbreak
- helps understand time trend, outliers, magnitude, mode of spread, period of exposure.
-
point source outbreak (temporal pattern)
- group exposed to single noxious influence. Rapid rise and fall in number of cases. Abrupt beginning within range of one incubation period, and does not persist over time.
- No animal-to-animal transmission, just a single source at a single time.
-
continuing source outbreak
- repeatedly or continuously exposed to noxious influence.
- Rapid rise, continuted high level of disease until source is removed (often water or food) or most are affected.
-
propagated outbreak
- host-to-host transmission
- slow rise in cases, multiple cycles.
- Often successively taller peaks, merge into waves (exponential growth).
-
shape of epidemic curve and time scale depend on
- incubation period
- infectiousness
- proportion of susceptible
- animal density/density of population
-
attack rate difference
- AR = (AR of exposed) - (AR non-exposed) = % difference in disease in exposed. Helps pinpoint culprit.
- Negative value seems to be protective
-
period of likely exposure
- count back minimum, average and maximum incubation times = likely period of exposure. Look for something everyone was exposed to.
- If TOO wide, may not be point-source.
-
Power vs P value
- power is likelyhood of being correct
- P-value is likelyhood of being incorrect.
|
|