The idea is simple: when a patient has a bacterial infection, doctors should find out what kind of bacteria it is so they can give the most effective treatment.
But growing bacteria in a lab to identify it takes hours to days. That means doctors have to make their best guess, and because there are dozens of possible strains, they guess wrong up to a third of the time.
Seattle biotech startup ID Genomics is developing a solution to that conundrum: a system that tests the biological “fingerprint” of a bacteria and connects it to a huge database to determine what bacteria it is, and what the best treatment is.
A new clinical study of the technology found that it was able to bring down errors in diagnosing bacterial infections a huge amount, to as low as 5 percent in some cases. For comparison, some areas of the country routinely see error rates of 35 to 40 percent.
The trial took place at a large emergency care center in Seattle, where it was able to identify the different strains of E. coli bacteria in patients with urinary tract infections in under 30 minutes.
Once identified, the tech developed a resistance profile of the patient’s infection, which told doctors what drugs would work most effectively for each patient.
Other than helping patients get the correct treatment more quickly, the technology can also help the medical field reduce the over-prescription of broad-spectrum antibiotics, the study’s authors wrote. This is important because the widespread use of those antibiotics leads to drug-resistant bacteria that are difficult to treat.
“Precision technology is a critical tool for preventing the overuse of broad-spectrum antibiotics by encouraging the use of first-line antibiotics, as recommended by current guidelines,” Dr. Evgeni Sokurenko said in a press release. Sokurenko is the co-founder of ID Genomics and a professor of microbiology at the University of Washington School of Medicine.
“The results in this article show how ID Genomics’ system may improve patient outcomes. We are looking forward to applying the technique to a broad range of infectious diseases,” he said.
In January, ID Genomics announced a nation-wide consortium of eight hospitals and medical centers would start contributing data on bacteria to its BacNet database, helping build the database’s ability to quickly identify a bacterial infection and the best treatment for it.
The database takes into account the biological “fingerprint” of the bacteria, tested using ID Genomics’ CloNet device. It combines the fingerprint with records of local bacterial outbreaks and other information when it makes a diagnosis.
The study of ID Genomics’ system was conducted by researchers from the University of Washington, Kaiser Permanente Washington, Seattle Children’s Research Institute, the Minneapolis VA Medical Center, the University of Minnesota, and ID Genomics, Inc.
ID Genomics spun out of the UW in 2013, and now employs 10 in its Seattle office and its lab in Bellevue, Wash.