When the Medicine Gets Weaker: E. coli, Antibiotic Resistance, and Why Water Treatment Is Your Family's Firewall
Most people meet E. coli as a headline — a beach closing, a food recall, a boil-water notice. What the headline usually leaves out is the part that should concern every family on a private well: some E. coli infections no longer respond to the antibiotics that used to stop them reliably. When the medicine gets weaker, the value of never getting infected in the first place goes up. That is what water treatment is. It is not a luxury appliance. It is a firewall — a barrier that stands between your family and the germ, working before anyone is sick, instead of after.
This article goes into the depth the subject deserves: what E. coli actually is, how it learned to beat our drugs, what researchers at the University of Michigan and Michigan State University have documented about that process, and why testing and treatment at your own tap is the most practical defense a rural family can own.
First, an honest word about E. coli
Escherichia coli is not one germ. It is a large family of bacteria, most of which live harmlessly — helpfully, even — in the intestines of people and animals. The trouble comes from specific pathogenic groups. Shiga toxin–producing E. coli (such as O157:H7) cause severe intestinal illness. Uropathogenic E. coli cause most urinary tract infections — infections so common that half of all women will have one at some point in their lives. Left to climb, those infections can damage kidneys or seed the bloodstream, and bloodstream infection — bacteremia — is where a 'routine' germ turns life-threatening. The University of Michigan Medical School's Mobley laboratory, which studies exactly these Gram-negative pathogens in bacteremia, puts it bluntly: 'We are losing the battle against antibiotic-resistant Gram-negative bacterial pathogens in our hospitals and clinical care facilities.'
Because E. coli is shed in feces, it is also the world's standard indicator organism. You cannot test water for every possible pathogen — there are too many, and each test is slow and expensive. So laboratories test for E. coli as the alarm bell: if it is present, fecal contamination has found a path into the water, and whatever else travels with feces may be present too. Michigan State University runs its own campus water system — 17 groundwater wells serving some 60,000 people — on exactly this logic. Every month the wells are tested for total coliforms; any positive result triggers an automatic test for E. coli; and a positive E. coli result requires public notification. All told, the plant performs about 13,000 tests a year. If a university with a full-time water staff tests that relentlessly, a private well — where the homeowner is the entire water department — deserves no less.
How E. coli learned to beat the drugs
Bacteria do not need to invent resistance from scratch. They have three well-worn strategies, and E. coli is good at all of them.
1. Destroy the drug. Many resistant E. coli produce enzymes called extended-spectrum beta-lactamases (ESBLs) that chemically break apart whole families of antibiotics — penicillins and most cephalosporins — before the drugs can work.
2. Share the instructions. Resistance genes ride on plasmids, small circles of DNA that bacteria pass to one another like a copied page. One bacterium that survives an antibiotic can hand the trick to its neighbors — including bacteria of other species. Resistance is not just inherited; it is published.
3. Clone a winner. Once a resistant strain also happens to be good at colonizing people, it can spread worldwide as a single successful lineage. That is the story of E. coli ST131, and a 2026 study in Clinical Infectious Diseases — a fourteen-site U.S. prospective study whose authors include University of Michigan Medical School infectious-disease researcher Owen R. Albin — measured what it is doing to American medicine. Among 282 patients with E. coli bloodstream infections, 43 percent carried ST131. Ceftriaxone — a workhorse antibiotic — failed against 79 percent of ST131 isolates and 86 percent of its most resistant subclade, compared with 27 percent of other E. coli. The consequence: doctors reached for carbapenems, among the last reliable drugs we have, in 74 percent of ST131 cases versus 31 percent of the rest. Every one of those escalations spends a little of the last line of defense. The study's title says it plainly: this one strain is driving carbapenem use — and carbapenem use is the selection pressure that breeds carbapenem resistance.
Michigan State's contribution to this science is the deepest of all. Richard Lenski's laboratory at MSU has run the world's longest evolution experiment with E. coli since 1988. In a 2021 study, Kyle Card, Lenski, and colleagues sequenced E. coli clones that had evolved antibiotic resistance from different genetic starting points and found that the path to resistance depended on the strain's history — 'historical contingency can alter both genotypic and phenotypic pathways to antibiotic resistance.' In plain terms: resistance evolves readily, but not predictably. There is no single mutation to watch for and no single drug design that closes the question. That unpredictability is itself an argument for prevention — you cannot count on forecasting the enemy, so you stop him at the wall.
And resistance does not stay in hospitals. University of Michigan School of Public Health researcher Joseph Eisenberg and colleagues followed twenty-one villages in coastal Ecuador across five years and 2,210 samples, tracking E. coli resistant to the common antibiotic pairing of ampicillin and sulfamethoxazole. Resistant E. coli was significantly more prevalent in villages along a new road than in remote villages — individual antibiotic use alone could not explain the pattern; transmission through the community did. Resistant bacteria move through people, animals, and the environment — and water is one of the oldest roads they travel.
Meanwhile, the pipeline meant to save us runs thin. University of Michigan's Life Sciences Institute describes researchers struggling 'to stay a step ahead of growing antibiotic resistance given the trickle of new alternatives entering the market,' noting that 'resistance now threatens to eclipse some of our best lines of defense.' Michigan's universities are fighting back on several fronts at once — U-M teams are hunting natural products that starve pathogenic E. coli of the iron it needs to survive, and U-M biomedical engineers have built an AI model, validated against E. coli, that predicts which antibiotic combinations will work against resistant pathogens with the least toxic side effects. That is good science, and it will take years to reach a pharmacy shelf. Your well is in the ground today.
Why a resistant infection is a different kind of emergency
When the first antibiotic fails, three things happen. Treatment is delayed while doctors culture the infection and find what still works — and with bloodstream infections, hours matter. The drugs that remain are broader, harsher, and more likely to be given intravenously in a hospital. And the failure rate climbs. Prevention inverts every one of those problems: an infection that never occurs needs no antibiotic, applies no selection pressure, and risks nothing. Every glass of contaminated water your treatment system stops is, in a small but literal sense, an antibiotic that still works tomorrow because nobody had to spend it today.
The firewall: what water treatment actually does to bacteria
Here is the point the drug companies cannot match: antibiotic resistance is a defense against drugs — and water treatment is not a drug.
- Ultraviolet (UV) disinfection damages the DNA of bacteria and viruses directly, so they cannot reproduce. An ESBL enzyme that shreds penicillin does nothing against ultraviolet light. Resistance genes are simply irrelevant to a physical mechanism.
- Reverse osmosis is a physical barrier: the membrane's pores pass water molecules and reject bacteria outright. The germ is not poisoned; it is stopped and flushed away.
- Fine filtration removes sediment and the particles bacteria ride on and hide behind — which also matters because UV only works well in clear water. Barriers work as a system: clarify, then disinfect.
- A softener alone is not disinfection — and anyone who tells you otherwise is selling, not teaching. Softening treats hardness minerals. Bacteria control is a separate job, done by disinfection and physical removal, chosen because a test said it was needed.
That last discipline — test first — is the whole Hambly method, and it is the same discipline the universities preach. MSU Extension tells private well owners that yearly water testing is how you ensure the safety of water a private well delivers, and its Agrifood Safety guidance shows what careful sampling looks like: test for generic E. coli, keep the sample cold, and get it to the laboratory within six hours, because E. coli can grow or die off in the bottle and change the result. Good data first. Then treatment built for your water — never a box off a shelf.
The detection side is getting faster, too. MSU biosystems engineer Evangelyn Alocilja's team has built nanoparticle biosensors that can pull E. coli and its fellow waterborne bacteria out of a sample and identify them in two to four hours instead of one to two days — and the same technology is now being field-tested to detect antimicrobial-resistant organisms themselves. Faster answers mean faster treatment — in a clinic, and at a kitchen tap.
Your family's checklist
1. Test your well water for bacteria every year — total coliform and E. coli — and after flooding, well work, a new pump, or any sudden change in taste, odor, or clarity.
2. Protect the wellhead. Keep septic systems maintained and properly separated, keep manure and chemicals away from the well, and make sure the cap and casing are sealed against surface water.
3. If a test ever comes back positive for E. coli, stop drinking the water until it is resolved — then disinfect the system, find the pathway, retest, and put a permanent barrier in place if the risk is ongoing.
4. Match treatment to the test. Bacteria risk calls for disinfection (UV) and physical barriers (filtration, RO at the drinking tap) — sized and installed for your water, and maintained so the barrier never quietly lapses.
5. Retest after treatment. A firewall you never inspect is a hope, not a firewall.
The bottom line
E. coli is getting harder to treat. Michigan's own universities have documented the enzymes, the gene-sharing, the globe-trotting resistant clones, and the thin pipeline of replacements. None of that is a reason for fear — it is a reason for a barrier. Antibiotics treat sick people; water treatment keeps families from getting sick. As the medicine gets weaker, the firewall matters more.
We test water free at your kitchen tap — about 20 minutes, no pressure, just the results. Call or text (616) 240-0025. Hambly Water Conditioning, Holt, Michigan — serving rural Michigan.
Notes
1. Sara Talpos, 'Stumping Iron,' University of Michigan Life Sciences Institute, 2017.
2. Mobley-Pearson-Anderson Research Laboratory, 'Virulence and Fitness Factors of Gram-Negative Pathogens in Bacteremia,' University of Michigan Medical School, accessed November 18, 2026.
3. Bethany Mauger, 'What It Takes to Supply a 60,000-Person Campus With Water,' Michigan State University Water Alliance, accessed November 18, 2026.
4. Natalie A. Mackow et al., 'Escherichia coli ST131 Drives Carbapenem Use for E. coli Bloodstream Infections,' Clinical Infectious Diseases 83, no. 1 (2026): e27–e37.
5. Kyle J. Card et al., 'Genomic Evolution of Antibiotic Resistance Is Contingent on Genetic Background Following a Long-Term Experiment with Escherichia coli,' Proceedings of the National Academy of Sciences 118, no. 5 (2021): e2016886118.
6. Joseph N. S. Eisenberg et al., 'In-Roads to the Spread of Antibiotic Resistance: Regional Patterns of Microbial Transmission in Northern Coastal Ecuador,' Journal of the Royal Society Interface 9, no. 70 (2012): 1029–1039.
7. Talpos, 'Stumping Iron.'
8. 'Researchers Develop AI Model to Target Safer, More Effective Drug Combinations for Antimicrobial Therapy,' University of Michigan Biomedical Engineering, March 9, 2026.
9. Bindu Bhakta, 'Protect the Water Quality of Your Water Well,' Michigan State University Extension, June 2, 2024.
10. Michigan State University Extension Agrifood Safety Workgroup, 'Agrifood Safety Minute,' accessed November 18, 2026.
11. Michigan State University, 'Faster Methods for Detecting Viruses in Food, Water,' MSUToday, August 2025.
Bibliography: Bhakta (MSU Extension 2024); Card et al. (PNAS 2021); Eisenberg et al. (J Royal Society Interface 2012); Mackow et al. (Clinical Infectious Diseases 2026); Mauger (MSU Water Alliance); MSU MSUToday Aug 2025; MSU Extension Agrifood Safety; Mobley Lab (U-M Medical School); Talpos (U-M LSI 2017); U-M Biomedical Engineering Mar 9 2026. (Full URLs/DOIs are in the source manuscript — include the Notes list above verbatim; do not drop the notes/bibliography.)
Keep reading




Comments