Return to Main

FAQ

Why is it important to define infectious disease baseline activity?

The Station provides information regarding when to expect disease activity and its implications for the community. This is important for preventive health activities such as:

  • promoting immunizations;
  • prompting patients and healthcare providers to consider a particular disease diagnosis at the right time; and
  • encouraging preventive, healthy behaviors such as hand washing and disinfecting common work areas.

Many people base important health decisions, such as whether to vaccinate, on this kind of anticipatory information- especially if we are seeing highly unusual conditions. Occasionally, we notify healthcare providers who work in critical areas of our healthcare system such as the intensive care unit of a hospital, to prepare for unexpectedly high patient demand. This important information helps communities respond appropriately to health crises.

What data is used to determine infectious disease baseline activity?

We use public health data provided by local, state, and national public health agencies. Because the data is stripped of any information that identifies individual people, the data we receive is simply the number of patients diagnosed with a particular infectious disease (e.g. influenza) on a weekly basis. The data for our baselines must be reliable, accurate, and (ideally) produced the same way, year after year.

How are activities and baselines of infectious diseases established?

Although there are many ways to estimate future infectious disease activity, we are currently using an approach borrowed from weather forecasters referred to as "unskilled forecasting." An example of an unskilled weather forecast is a graph of the average monthly rainfall. We are able to provide reliable information for several infectious diseases by publishing a range of the most likely number of cases expected per week, up to one year in the future.

We are employing unskilled forecasting to gradually build public trust in this fairly reliable, although occasionally imprecise process. Over time, we expect to be able to introduce a more sophisticated approach called "skilled forecasting." This category of forecasting enables us to predict, on a given date, the precise number of cases of an infectious disease expected. However, at this time, skilled forecasting is more experimental and less reliable than unskilled forecasting.

How applicable is weather forecasting to defining infectious disease baselines?

Weather forecasting is very similar to defining infectious disease baseline and prognosis. Both rely on data that has been reported on a regular basis over a long period of time. Studying past patterns allows scientists to make predictions such as future weather or disease activity. In both cases, the information must then be communicated effectively to the public. Weather forecasters have been doing this since the 1800s and have learned many valuable lessons that are applicable for defining infectious disease activity as well!

For more technical information regarding our operational approach to raising public awareness and education, please see the World Weather Research Programme (WWRP) / World Climate Research Programme (WCRP) Joint Working Group on Forecast Verification Research's website

Sometimes it looks like your baseline doesn't match what is actually happening. Why?

If the baseline fails to predict when a disease might hit our community, or to what magnitude, this typically indicates unusual disease activity. For example, consider a year when high numbers of people come down with flu so severe they have to be hospitalized -- or worse -- they die from it. The baseline "failed" because it hadn't seen that kind of flu pattern before, at least, not often enough to make it part of flu's "normal" pattern in our community. However, this unusual pattern then becomes part of future instances and observations, especially if that pattern of severe flu repeats itself. So, if severe flu became a new normal for a particular region, then in a few years, the baseline would recognize this, and people wouldn't be taken by surprise whenever another season of severe flu hits. Instead, the local healthcare system would have expected the possibility, and would likely be prepared for it.

What are the limitations of the designed system?

All intelligent systems have inherent limitations.

  • We know, for example, that diseases are sometimes reported more often after there is media attention. This may give us an erroneous impression there is more disease activity than there actually is.
  • Another example is if the population of a community is expanding rapidly. More people sometimes means more disease, which then may create the impression that we are seeing more disease. But this isn’t necessarily true because the cases seen per 100,000 individuals (otherwise known as "incidence") hasn't actually changed.
  • Disease reporting may increase over time due to improvements in the healthcare system, and this may result in increased alerting that is not actually due to true increase in disease.

The data is not perfect and neither is the prognosis. In summary, this site and its resources should be used as a tool to help us recognize changes in disease activity that still needs people to properly interpret the information.