Editor's note: CIDRAP's Promising Practices: Pandemic Influenza Preparedness Tools (www.publichealthpractices.org) online database showcases peer-reviewed practices, including useful tools to help others with their planning. This article is one of a series exploring the development of these practices. We hope that describing the process and context of these practices enhances pandemic planning.
Jun 26, 2009 (CIDRAP News) – A couple of years ago, the system for reporting and evaluating influenza-like illness (ILI) outbreaks in Fort Worth, Tex., schools—a key part of Tarrant County's disease surveillance program—didn't exactly yield fine-grained information in real time.
School districts gathered up attendance information from their schools once a week and sent it by fax, or in some cases e-mail, to Tarrant County Public Health (TCPH). At TCPH, officials analyzed the data manually. In case of outbreaks, they issued fact sheets and reports to the school districts on an as-needed basis.
That somewhat creaky system began to change about 2 school years back, when TCPH developed a Web-based system that permits school nurses to submit daily reports on absenteeism and ILI directly to the health department. The same system allows TCPH to analyze the data in a more automated fashion and to provide abundant information to school nurses via the Web.
Now the Tarrant County School Health Surveillance System (SHSS) program is making headway, TCPH officials say. It has been embraced by about a third of the schools in Tarrant County, and it has overcome a number of glitches. Officials say they are encouraged.
Under the old system, TCPH received only weekly, school-district–wide data, making it hard to quickly pinpoint outbreaks, and if a report wasn't submitted, data for a whole district was missing, said Dean Lampman, regional surveillance coordinator for TCPH.
"So we wanted to fix all those things, and the system we're developing is our attempt to do that. And it's been mostly successful," Lampman said.
Quest for surveillance improvements
Within TCPH is the Southwest Center for Advanced Public Health Practice, a program funded by the Centers for Disease Control and Prevention (CDC) to develop exemplary public health practices that can be readily replicated by others. There are seven such centers around the country, Lampman said.
In line with that program, Lampman said he was looking for ways to improve the whole spectrum of TCPH's disease surveillance tools, including hospital data, over-the-counter drug sales, ambulance calls, emergency medical service data, and school absenteeism reports, among other things.
The idea of developing an Internet-based system for school disease surveillance emerged from discussions Lampman had with Tabatha Powell, MPH, who at the time was a doctoral student in public health, and others, he said.
"I thought if I could develop a system that could collect data electronically and have it flow into the ESSENCE system, we might be on to something," he said. ESSENCE is the computer program Electronic Surveillance System for the Early Notification of Community-based Epidemics, which is widely used in public health to look for statistical anomalies in disease data, he explained.
Electronic reporting and automated analysis were two of the key objectives, said Lampman. Two others were to gather reports more than once a week and obtain them from individual schools, not just whole districts.
Speaking of the novel H1N1 virus, he said, "Let's say this was a deadly pandemic like we had in 1918—if we collected weekly, we wouldn't be happy with that. So we asked for daily collection."
More generally, the program was designed to ease information exchange between TCPH and county schools, support early detection of ILI, focus public health resources in response, and give school nurses one-stop access to information to help them promote disease prevention, according to a report Lampman and colleagues wrote after the program's first year.
Dual development tasks
Developing the system involved the dual tasks of laying the groundwork with school nurses and building the electronic tools.
TCPH held half a dozen meetings with nurses from the county's approximately 300 public schools to enlist their support and gather suggestions on how the system would work, Lampman said.
Getting their support "wasn't as easy as we might've hoped," he said. "However, we didn't get into this with any misconceptions that it was going to be easy. Even with the old program we had less than 100% compliance."
None of the nurses thought the SHSS was a bad idea, but some were concerned about how much time it might take, Lampman said. In the end, "We decided there was enough of a green light to go ahead."
On the electronic side, the system was built on an open-source platform called Dot Net Nuke. The heart of the tool is a daily report form that nurses see after they log in. There, nurses are asked to report total absences, absences due to ILI, the number of students seen by the nurse, and how many of those students had an ILI.
In addition, nurses are asked to report whether the number of students with ILI seems to be increasing, decreasing, or static. Additional fields ask for information on faculty absences, including those related to ILI.
Other pages on the Web portal provide maps of changing disease patterns, access to flu prevention resources, news items, analysis, and action items suggested by TCPH, according to Lampman's report.
With the use of open-source software, the SHSS wasn't very expensive to launch and was supported by TCPH's Advanced Practice Center funds, he said. Later, TCPH applied and received flu surveillance money from the Texas Department of State Health Services.
TCPH launched the system in the fall of 2007. It quickly became clear that broad participation wasn't going to be achieved overnight.
"We rolled it out to everybody all at once, sent all the instructions—and then found that almost nobody started reporting immediately," Lampman said.
Eventually, TCPH officials met with the nurses to promote the program further and get their feedback, Lampman said. "It took the epidemiologist to engage with the school nurse and the lead nurse and have a face-to-face meeting, and then it came around," he said.
Part of the problem was that at the start of the school year, nurses are focusing on dealing with children's vaccination records, said Diana Cervantes, a TCPH epidemiologist. "I think at the beginning of the year the reporting is lower because that's their emphasis," she said.
Lampman's official report says more than 200 school nurses in 7 of the county's 16 school districts were trained to use the system and the online report form, which they could complete in less than 5 minutes. However, the initial data harvest varied in quality and was hard to assess because only about 60% of the nurses in the seven districts actually reported.
A chart of system statistics shows that participation grew over the past 2 school years and peaked late in the flu season each year. The peak for the 2008-09 year came the first week in March, with 104 schools from nine districts reporting. (The chart covered records only through April.) The previous year, the peak came in mid-February, with 84 schools from four districts reporting.
When the novel H1N1 virus emerged in April and May, officials saw increased reporting, Lampman reported. He said the epidemic seemed to improve the quality of the reporting, as several nurses took time to call or e-mail corrections to some of their data.
When H1N1 cases broke out in Fort Worth children, the school district there closed at the recommendation of TCPH. The health department advised another district to shut down, but instead the district closed just the one school where a case was found, Lampman said.
Those closure recommendations were not driven specifically by the SHSS, said Cervantes. They were shaped more by traditional "shoe-leather epidemiology," the gathering and testing of specimens from local clinics and hospitals, she said. "The School Health Surveillance System is still kind of in its infancy; it wasn't something that we really used when H1N1 hit to really detect any increase in absenteeism," she added.
One of the key remaining challenges for making the SHSS fully useful is to determine baseline levels of ILI in the schools, Cervantes said. "We still need to work to establish a baseline so we know when there's an increase in ILIs or other illnesses. At this point it's difficult to say if a certain level of ILI or illness is unusual."
Another challenge is to persuade school nurses to report daily. "They were doing it as they had time, which was basically once or twice a week," Cervantes said.
She hopes that when the system becomes more mature, it will be of greater help in making school closure recommendations and other actions.
"That's our hope, that when the reporting becomes a little more consistent and we're able to extract the data and use it more in a real-time sense, and we can establish a baseline, that in the future it'll help us make those decisions."
But already the system has strengthened the health department's links with school nurses and their ties with one another, she reported.
"They feel it's a good way to get information from the health department, and keep two-way communication open with us. They feel it's more personalized; they do enjoy that. There's a field [on the report form] where they can add comments and give us information," which was helpful when some Streptococcus infections cropped up.
"It also kind of helps build community with other nurses," she added. "It had them speaking to other school nurses in their area. Whenever I meet with them, that's the feedback."
Lampman said the system needs some technical enhancements. One is to make the data transfer and analysis more automated. Right now a technician has to transfer data from the system into ESSENCE, a step that Lampman hopes to eliminate with a change in the database structure.
Other future plans call for developing a "full-blown guidance document" and conducting an exercise simulating a hypothetical disease outbreak. "If H1N1 didn't sufficiently impress or engage some people, presumably a tabletop exercise will," he said.
SHSS materials on CIDRAP Promising Practices site