A large proportion of E. coli strains that produce deadly Shiga toxin go under-tested and thus under-diagnosed, according to a study by Washington State Department of Health epidemiologists who surveyed health laboratories in the state to determine how extensively they test for E. coli infections.

The study, published in the March 2012 Emerging Infectious Diseases, found that the increased use of testing for Shiga toxin in Washington between 2005 and 2010 directly correlated with increased diagnoses of non-O157 infection. Still, as of 2010, only 40 percent of stool samples tested in state labs were screened for both O157 and non-O157 strains.

E. coli O157 is the most well-known strain of Shiga toxin-producing E. coli (STEC), but many other strains have gained prominence in recent years after causing illness outbreaks of their own. Infections from many of these other strains, often referred to as “non-O157” STEC, are considered clinically indistinguishable from O157 infections, which can lead to severe diarrhea, hospitalization, kidney failure and even death. Recently, the U.S. Department of Agriculture has targeted six additional strains of non-O157 STEC to join O157 as a banned contaminant in meat.
 
But while indistinguishable from the others in its health impacts, O157 does require its own specific type of culture test to identify it. And up until 2005, it was largely the only type of E. coli strain being tested in clinical labs. A different test for Shiga toxin can detect any of type of STEC, including O157, but it takes longer to perform than the 24-hour O157 test.
 
In 2009, the Centers for Disease Control and Prevention published official STEC testing recommendations for labs, suggesting that all stool samples be tested for O157 with the culture test and all STEC strains with the Shiga toxin test simultaneously. Even though the Shiga toxin test also detects O157, the CDC recommends performing both tests to more rapidly detect O157, which is still the most common strain of STEC.

Testing for more strains of STEC beyond O157 is the only way to more accurately diagnose E. coli infection and foster more effective epidemiology, the study’s authors argued.

“Detecting a larger proportion of cases means we’re able to run PFGE [genetic fingerprinting] at Washington Public Health laboratories and detect clusters of infection so we can remove the common source,” said Kathleen Stigi, lead author of the study and an epidemiologist at the Washington State Department of Health. “With more detection comes a better understanding of the virulence factors of these strains.”

The authors surveyed each of the 57 clinical labs in Washington that test stool samples, all but one of which routinely performed at least one type of STEC test.

As mentioned already, 40 percent of stool samples in Washington were tested according to the CDC’s recommendation of dual O157 and STEC tests at the end of 2010. Labs that tested only for O157 screened 47 percent of samples, while 13 percent were only screened with the STEC test.

From 2005 through 2010, the number of labs testing for non-O157 strains rose from two to 19. During that same time, the number of detected non-O157 infections shot from eight in 2005 to 76 in 2010. The increases in non-O157 detections happened most sharply from 2008 to 2010, corresponding to the years in which the majority of the labs began testing for STEC beyond O157.

The correlation between increased testing and increased diagnoses is self-apparent, the authors said: If more labs test for non-O157 STEC, more people infected will be properly diagnosed and more outbreaks might be detected.

The authors estimated that in 2010, half of all non-O157 STEC infections that could have been detected were not, simply because the labs testing potentially infected stool samples did not perform the Shiga toxin test.

Stigi said the biggest factors holding back labs from performing both types of tests were costs, procedural changes and staffing constraints. Along with that, more STEC cases result in more epidemiology work: Every reported STEC infection requires an epidemiological investigation for local health officials. Testing for more pathogens means requiring more man-hours on the investigation side during a time when agencies are pulling their purse strings tighter.

Regardless, the authors recommended labs adopt the CDC’s dual testing strategy for the sake of public health.

“The potential virulence of non-O157 STEC infections underscores the need for enhanced laboratory testing and epidemiologic research,” the study read. “To encourage adherence to STEC testing recommendations, healthcare providers should request Shiga toxin testing if it is not routinely performed at their laboratory.”

To the authors’ knowledge, their study is the first of its kind to specifically proportion STEC testing rates in terms of stool samples tested.

Washington had 945 confirmed STEC infections between 2005 and 2010, with 83 percent of them O157 and 17 percent non-O157. The percentage of those infected, however, rose each year during the six-year study period, from 6 percent of infections in 2005 to 41 percent in 2010. Again, these results correlated with the increase in non-O157 STEC testing. The authors predicted that non-O157 infections would have accounted for 60 percent of all STEC infections in 2010 if all labs had the infrastructure to test for them.

Four strains of STEC accounted for the large majority of all non-O157 infections in the state: O26 (48 percent), O103 (18 percent), O121 (12 percent) and O111 (5 percent).

Stigi said she believes more labs have implemented Shiga toxin testing procedures since the surveys were conducted in early 2011, and she expects the number to grow, as epidemiologists recognize that accurately tracking disease trends will rely on more complete data.

“These enhanced testing practices have likely contributed to the increased detection of STEC in Washington,” she said. “It’s important for public health professionals to assess laboratory STEC testing practices in order to correctly interpret disease trends.”