Anatomy of an outbreak: ICPs, public health team solve the tale of the tub

Legionella outbreak requires vital partnerships across continuum of care

Hospital infection control professionals in southwestern Virginia recently played key roles in alerting public health officials and assisting in the investigation of a mysterious outbreak of community pneumonia that infected 23 people and caused two deaths.

Investigators from state and local public health departments and the Centers for Disease Control and Prevention ultimately tracked the outbreak of Legionella pneumophila to a whirlpool spa display at a home improvement store. The health care community’s response to the case in the Blacksburg, VA, area provides an apt illustration of how hospital clinicians and public health officials can work in concert to rapidly identify the etiology of an outbreak, track it to its source, and quell community panic.

While the outbreak involved a known pathogen, public health epidemiologists have underscored frequently in recent years how the threat of emerging infectious diseases will require enhanced communication and partnership across a continuum of care. For their part, ICPs in several community hospitals played critical roles in identifying incoming cases of unexplained pneumonia and ensuring they were tested under a broad-spectrum protocol as investigators tried to determine what manner of pathogen they were dealing with. Throughout the mid-October 1996 outbreak, ICPs served as "front-line epidemiologists," says Denise Benkel, MD, a CDC Epidemic Intelligence Service officer who assisted in the investigation.

"We needed them to pull records on people that had pneumonia — before we even knew it was Legionella," she says. "Once we actually figured out we had a Legionnaires’ disease outbreak on our hands, the ICPs helped us by identifying those currently in the hospital that needed to be tested and making sure the test protocol was followed."

The first call to public health officials was made by Jennifer Brumfield, RN, infection control practitioner and employee health nurse at Columbia Montgomery Regional Hospital in Blacksburg. Alerted by an attending physician on call that seven community pneumonia cases had been admitted the previous day, Brumfield began looking for other cases in previous admissions. She found that some 15 patients had been admitted with unexplained pneumonia during Oct. 8-14.

"I started going all over the hospital and found that this doctor had one or two pneumonias and another doctor had one or two," she says. "We ended up having a bunch of pneumonia patients. That’s when I started calling other facilities to see if they were seeing an increase. Then I called the health department to see if anything was going on the community."

Among the other ICPs Brumfield contacted was Betsy Allbee, RN, CIC, infection control practitioner at Radford (VA) Community Hospital, which is about 15 miles away. When Allbee looked at new admissions and tracked back through patient records, she found a similar upsurge in unexplained pneumonias.

"They weren’t all definite pneumonias when they came in," she says. "I looked at things like shortness of breath and coughing. I looked at our ED admissions — those are based on chief complaints of the patient. They are not really reliable, so we really had to look a little deeper, but we did see quite a few with pneumonia."

With both Brumfield and Allbee reporting an increase in unexplained pneumonia, local health officials conferred with state epidemiologists and the CDC came in to assist in the investigation of the unfolding outbreak.

"State health epidemiology and CDC came up with a protocol of the test they wanted us to do," Brumfield says. "I got a list of all new pneumonia cases everyday and made sure all of the orders were followed and all of the tests were being completed."

Though there were many more "probable" cases, review of records at the aforementioned hospitals and two other area facilities ultimately turned up 23 cases that were strongly linked to the Legionella outbreak, Benkel says.

A shotgun approach

Critical to getting the final case count and identifying the etiology of the outbreak in its early stages was implementation of the aforementioned testing protocol, notes Jody Hershey, MD, MPH, one of the principal investigators of the outbreak and director of the New River Health District in Christiansburg, VA.

"The lead physicians and lead infection control nurses were willing to inform their peers and ‘sell’ this protocol," he says. "It was put into place within hours after we developed it."

The protocol took a shotgun approach, including chest X-rays, complete blood count analysis, urine antigen tests for Legionella, sputum cultures, and acute serum specimens.

"We were also ordering nasopharyngeal washes and swabs to try to make sure that it wasn’t a viral etiology — we were trying to diagnose what the pneumonia was," Hershey says.

"Blood cultures were done to make sure it wasn’t strep pneumo or some other septic, bacterial pneumonia. With the acute serum [tests], we were looking for things like parainfluenza, influenza, mycoplasma, Legionella."

Thousands of cases go undetected

The situation was somewhat unusual in that patients with unexplained pneumonia are not typically tested for Legionella, but are simply treated until they recover. Indeed, most of the estimated 10,000 to 15,000 cases of Legionella infection that occur annually in the United States are thought to occur sporadically and go undiagnosed.

"But for purposes of this outbreak we were trying to confirm and diagnose as many cases as possible," he says.

Even as the diagnostics rapidly began pointing to Legionella as the culprit, an educational effort became necessary as word spread within the community that public health officials were investigating an outbreak of unexplained pneumonia.

"People were panicky," he says.

Hershey became the liaison to the local press. He told them he would give them as much information as possible if they in turn would agree to go through him rather them attempting to directly question investigators and clinicians. Brumfield worked with public health officials to develop a staff handout for hospital workers that outlined the basics of community-acquired Legionella, which does not spread from person to person and requires no special isolation of incoming patients. (See handout, p. 52.)

"As soon as we had a handle on what we might be dealing with, I met with all the department managers and handed those out to educate the staff so we wouldn’t have a panic in the hospital," she says.

Still, another important questions remained unanswered. What was the source of the Legionella, and was transmission still occurring in the community?

Public health investigators began individually questioning confirmed cases, using a calendar of all recent public events in an attempt to find a common site of exposure that would create the typical mode of Legionella transmission — aerosolized water that is inhaled. Of particular interest were recent visits to any sites near water sources, particularly fountains, pools, garden centers, and misters in grocery store vegetable sections.

"Within 36 hours after we started the interviews — we had several teams working in groups of twos — one area began surfacing," Hershey says.

The site in question was a nearby home improvement center frequented by many of the patients, many of whom remembered looking at a display of two working whirlpool spas in a high-traffic area of the store. People were not getting into the display spas, but many remembered walking by them and lingering to look at the tubs. When the health department went in the store they found that one of the spas was still on display, but it been recently drained and cleaned. The other spa had been sold and removed from the store on Oct. 11. The spa on display at the store and other environmental cultures taken there did not culture positive for Legionella. Investigators turned their attention to the other tub, which had been sold to a store employee.

"The small hot tub that had been sold to an employee was still in the employee’s garage and had never been put back up," Hershey says. "The bacteria grew out in an exact DNA match from two of the patients."

Store unknowingly broke the chain

By removing the hot tub that was the likely source of the outbreak, the store had inadvertently broken the chain of transmission even before the outbreak investigation began, he notes. The infected patients, however, all fell within the normal incubation period for Legionnaires’ disease of two to 10 days.

"Early on in the investigation, I was wondering why the numbers weren’t staying up," Hershey says. "Something had broken the transmission."

To add an epidemiologic component to the lab evidence, a case-control study was undertaken using three controls for each confirmed case and matching them by age, sex, and underlying medical conditions. Of the 23 cases, 15 were included in the case-control study because one had died and seven had not yet been identified when the study began, according to the CDC report of the investigation.1 A history of having visited the home-improvement center during the two weeks before onset of illness was reported by 14 (93%) of the 15 cases, compared with 12 (27%) of the 45 controls. Of the 13 case-patients and 12 controls who had visited the store and for whom there was a detailed in-store exposure history, cumulative duration of total store visits averaged 79 minutes for cases and 29 minutes for controls. In addition, 10 (77%) case-patients reported spending time in the area surrounding the spas during their visits to the store, compared with three (25%) of the 12 controls.

No other activity, including drinking from the store’s water fountains or visiting other locations in the community, was associated with illness, the CDC reported. The mean age of case-patients was 65 years, and 17 of the patients were men. In general, risk factors for acquiring Legionnaire’s disease include age greater than 50, male gender, a history of heavy smoking or alcohol drinking, or having underlying disease or being otherwise immunocompromised.

"Every one of the individuals who was confirmed in this outbreak had one or more of those risk factors," Hershey says.

Though cultures were not available for all case patients to do DNA comparisons, the matching strain in two patients and the tub filter, combined with the epidemiologic evidence, all point to the store spa as the source of the outbreak, he says.

"Three of our four community hospitals had higher-than-normal admissions for community-acquired pneumonia in the week or two prior [to recognition]," he says. "It happened that Jennifer Brumfield was the one who called — she was sharp and astute. There is no question that the teamwork effort made this a success. In some areas, that may not have occurred. I don’t want to say the outbreak was ‘exciting’ because people died and it was a tragedy, but it was exciting to be part of the effort."

Following the investigation, health officials recommended that whirlpool spas being used as displays be regularly inspected and maintained with biocides and that their filters be regularly changed or decontaminated. In response to an outbreak of Legionnaires’ disease on a cruise ship due to a hot tub, the CDC issued guidelines last year for the maintenance of whirlpool spas on cruise ships.2,3


1. Centers for Disease Control. Legionnaires’ Disease Associated with a whirlpool spa display — Virginia, September-October, 1996. MMWR 1997; 46:83-86.

2. Jernigan DB, Hofmann J, Cetron MS, et al. Outbreak of Legionnaires’ disease among cruise ship passengers exposed to a contaminated whirlpool spa. Lancet 1996; 347:494–499.

3. National Center for Environmental Health/National Center for Infectious Diseases. Final recommendations to minimize transmission of Legionnaires’ disease from whirlpool spas on cruise ships. Atlanta: US Department of Health and Human Services, Public Health Service, CDC; 1996.