Using Nurse Triage on Call™ as an Indicator for the Flu Season

By Dr. Ravi Raheja and Charu Raheja, PhD

As we know, the timing of the flu is unpredictable and can vary from season to season. According to the CDC website, flu activity most commonly peaks in the United States in January or February. The 2012-2013 influenza season, however, began relatively early compared to recent seasons. The worry among health officials is that flu will continue to peak during January and February making 2013 one of the worst flu seasons in history. While it is not possible to predict when the flu season will peak, we can use some data to try to estimate how flu activity is changing over time.

There are several potential indicators for flu activity. The CDC looks at the proportion of people seeing a healthcare provider for influenza-like illness and hospitalizations associated with influenza. The CDC also takes into account deaths attributed to pneumonia and influenza. While this is a good way to get information on people who actually have the flu, the information depends on the health care centers providing data on the disease to the CDC. The difficulty then is in figuring out the cases where people have flu-like symptoms, but the symptoms do not get reported. Further, it can take a few days for the cases to be reported, making it difficult to evaluate and follow the disease over time.

People who have a telephone triage nurse available are likely to seek advice from a nurse about how to take care of their symptoms, even if they don't go see a healthcare provider. This enables a nurse triage system to get additional information about the number of sick cases.

We used triage nurse data to help estimate the flu season and to get an estimate of flu activity over time. Nurse triage phone calls are a great resource to get timed data because we can divide the calls based on when they came (per week, or per month). We can also get updated data as soon as the calls are finished.

Last month we reported flu data comparing the first half of November against the second half of December. We found that almost 35% of the callers in the last two weeks of December had symptoms related to flu such as cold, cough, fever, or pneumonia. In contrast, the first two weeks of November had about a 24% of callers with flu-like symptoms. We now compare the late December and January numbers.

The chart below gives a week-by-week percentage of callers with flu-like symptoms from the last week of December until January 20. The data consists of an average of about 4,000 calls a week. As we can see in the chart, the percentage of callers with flu-like symptoms has been steadily decreasing, suggesting that we might have already peaked in the flu season. The percentage of callers between Jan 14 and 20 was down to about 26%. This could be great news if the trend continues because it would mean that the flu season is behind us.

We will continue to monitor the percentage of callers with flulike symptoms to help us continue to follow flu activity and predict outbreaks.