What Scary Disease Lady (aka ‘Maryn McKenna’) wrote a couple of weeks ago about the clinical shift from culturing microorganisms to rapid diagnosis that removes the need to grow the disease-causing critter is very important (boldface mine):

To get diagnoses made more quickly — and thus get patients medical help faster — public health labs have begun using rapid tests that recognize a molecular signature, instead of the traditional culture-based tests that grow the bacterium on a dish and test it. This has a significant downside: No organism means no ability to do the “DNA fingerprinting” that the public health system relies on for PulseNet, a crucial tool that can link infections that occur many miles away from each other. Reacting to that, Caroline Smith DeWaal of the Center for Science in the Public Interest said in a statement:

It is critically important that CDC develop a plan to address the increasing use of diagnostic tests that don’t use lab cultures. Otherwise the trend of declining reporting of outbreaks may continue—not because fewer people are getting sick, but because state health departments and CDC cannot track the outbreaks.

I’ll add one thing to that. Not having a cultured organism also means losing the ability to detect when the foodborne illness is antibiotic-resistant, because resistance assays at the moment rely on having living bacteria. Antibiotic resistance is an increasingly important issue for food production; the now year-long outbreak in chicken from Foster Farms, which has racked up 524 infections in 25 states, involves a Salmonella that is multi-drug resistant. No longer being able to track resistance could mean completely losing track of foodborne epidemics.

We know this can happen, because it already did happen, for gonorrhea, which currently is veering toward untreatability all over the world. As I reported two years ago at Scientific American and here, multi-drug resistant gonorrhea got out ahead of us because public health’s switch to inexpensive rapid tests for STDs sacrificed the ability to examine whether the bacterium’s resistance profile was getting worse.

And I’ll add one thing to that as well. In the foreseeable future, to apply genomic sequencing to clinical microbiology, we’re going to need to culture organisms, either through traditional methods or single-cell methods (e.g., getting cells from blood)–don’t believe all the microbiome-y hype on this, as separating the wheat from the chaff in most clinical situations will either be prohibitively expensive or impossible. If we want to understand how organisms are entering and moving around in the clinic, the evolution and transmission of antibiotic resistance, and track spread on a regional scale, we’re going to need some isolates.

This could be a problem.