Biology Reference
In-Depth Information
transplants, Roberts did not mind the heat, having moved from her native
Michigan to escape the winter cold. “I knew it was hot because everyone was
complaining, and you know, the air conditioning isn't working right, but for
me the heat just meant more time in the pool,” Roberts told me when I spoke
with her.
On August 8, 2007, with the temperature in downtown Raleigh reaching
nearly 103°F, officials at the Division Public Health, including the future
N.C. State Health Director, Dr. Jeffrey Engel, the Department's Public
Information Officer at the time, Debbie Crane, and Dr. Jean-Marie Maillard
decided to use the NC DETECT system to track the impacts of the heat. Dr.
Maillard e-mailed Roberts and asked her to query the system regarding
the heat, although at first, neither Maillard nor Roberts were exactly sure
what they were looking for. The NC DETECT system provided a lot of data,
but it had never been used to track a heat wave before, and it was unclear
whether the effects of heat would be seen in emergency department visit
records. At the division's request, Amy Ising and Clifton Barnett of Anna
Waller's University of North Carolina team built a filter or record selection
algorithm to select patient records in the NC DETECT database that had
information that might indicate a heat-related hospital visit. Roberts then
ran that filter on the data and searched through the free text portion of
those records to get a sense of why people were being admitted for heat-
related reasons.
The system's functionality allowed Roberts to analyze records of patients
who visited emergency departments for obvious indicators of heat distress,
such as chief complaints of dehydration or heat exhaustion, as well as for less
obvious indicators of heat-related distress, such as “passed out on the foot-
ball field” or “felt dizzy when gardening.” Roberts was not sure whether she
would find anything as she methodically plotted the chief complaint data, but
once she organized her data into graphs, she knew she was looking at some-
thing important.
“I thought something went wrong,” Roberts recalled, “because they [the
heat admissions and the temperature trends] were so right together. It was
like a perfect epidemiological story.” Her figures indicated that the hottest
days of the summer were associated with the highest peaks in cases. She
shared the graphs with Dr. Maillard who suggested she plot the number of
cases directly against the daily high temperature reported at Raleigh Durham
Airport (Figure 16.1). Again, the results, reported back to Maillard a few days
later, were striking. From August 3 to August 11, North Carolina experienced
a dramatic surge in heat-related emergency department visits, peaking at
between 150 and 160 admissions per day on August 8 and 9. Although heat-
related admissions were lower before and after the worst days of the heat,
looking back over the summer, each temperature increase was accompanied
by a corresponding increase in heat-related admissions. Roberts was excited;
as she recalled, “This was exactly what you go to school for in epidemiology.
You look at that graph and anyone can see that pattern.”
