ORSANCO’s 2009 screening found several pharmaceutical or lifestyle markers in Ohio River samples near Louisville, including caffeine, carbamazepine, and gemfibrozil. The study sampled river water, including an effluent-influenced location—not household tap water—and did not establish a human-health effect from those detections.
- Mainstem sample 17: Ohio River mile 600.5
- Effluent-influenced sample 18: river mile 612.2
- Caffeine: 105 and 167 ng/L
- Carbamazepine: 11.1 and 14.5 ng/L
- Gemfibrozil: 9.52 and 12.8 ng/L
“Prescription drugs found in the Ohio River” is the kind of headline that stops people cold. It is also a headline that can mislead unless it answers four questions: which compounds, how much, where was the sample collected, and was it river water or finished drinking water? A major ORSANCO screening study gives Louisville-area homeowners real numbers—but the study did not test household taps and did not conclude that its detections caused human illness.
ORSANCO, the Ohio River Valley Water Sanitation Commission, collected samples in 2009 and published its report in 2010. The work screened for 118 pharmaceutical and personal-care compounds, 27 hormones and sterols, and 13 perfluorinated compounds across the Ohio River Basin. Near Louisville, sample 17 came from the Ohio River mainstem at river mile 600.5. Sample 18 came from river mile 612.2 and was specifically labeled effluent influenced, meaning the location was selected to help characterize the influence of a discharge.
What the Louisville-area samples found
Several compounds were reported in both nearby samples at concentrations measured in nanograms per liter—parts per trillion. Caffeine was measured at 105 ng/L in sample 17 and 167 ng/L in sample 18. Carbamazepine, an anticonvulsant and mood stabilizer, was reported at 11.1 and 14.5 ng/L. Gemfibrozil, a lipid-lowering medication, was reported at 9.52 and 12.8 ng/L.
The study also found other signals that help distinguish the two locations. Ibuprofen was below the method reporting limit in sample 17 but measured at 15.6 ng/L in the effluent-influenced sample 18. Naproxen followed a similar pattern: below the reporting limit in sample 17 and 13.3 ng/L in sample 18. Cocaine itself was below the reporting limit in sample 17 and measured at 0.434 ng/L in sample 18. Benzoylecgonine, a cocaine metabolite, was measured at 2.05 and 4.46 ng/L.
Those values are not doses of medicine. A therapeutic tablet contains amounts millions or billions of times larger. The detections are environmental markers showing that human-used chemicals can survive or bypass parts of the wastewater pathway and enter surface water at trace levels. Their presence is scientifically important without being equivalent to swallowing a prescription.
How pharmaceuticals reach waterways
Most pharmaceuticals do not enter the environment because someone empties a bottle directly into the river. After a person takes a medication, the body absorbs and transforms some of it and excretes the rest. Those compounds and metabolites enter sewer systems. Hospitals, nursing facilities, manufacturing, animal operations, septic systems, and improper disposal can add additional pathways.
Conventional wastewater plants are designed primarily to remove solids, oxygen-demanding material, nutrients, and pathogens—not every molecule of every modern pharmaceutical. Some compounds are degraded, some attach to sludge, some are transformed, and some remain in treated effluent. The result depends on the compound and the plant process.
Improperly flushing unused medication is avoidable. Drug take-back programs reduce the amount sent directly into plumbing and also reduce diversion and accidental poisoning risks. The gain is modest for any one household but meaningful when many households use take-back instead of the toilet.
Why sample type changes the meaning
The ORSANCO results came from river water. Sample 18 was intentionally “effluent influenced.” That makes it useful for studying where emerging contaminants enter the river, but it is not representative of every location or every day. It is also not a Louisville Water finished-water result.
A drinking-water plant adds treatment barriers between source water and the distribution system. Louisville’s Crescent Hill and B.E. Payne facilities use different source and treatment configurations. Finished water is monitored under drinking-water programs, while wastewater and river studies are designed for different purposes. Collapsing all three into “your tap” is inaccurate.
The age of the study matters as well. These samples were collected in 2009. They show that the pathway existed and document what was measured under the methods available then. They do not tell us the exact concentration today. Wastewater treatment, prescribing patterns, laboratory methods, river flow, industrial activity, and regulations can all change.
What the study said about human health
ORSANCO explicitly cautioned that the study was a preliminary screening effort designed to document occurrence and guide future work. The report stated that, at the time, limited evidence was available to conclude that environmental occurrences of these compounds posed a causal risk to human health. That caution remains essential.
Detection is the first step, not the last step, in risk assessment. Scientists also need toxicity information, exposure duration, concentration, sensitive populations, mixture effects, and the amount that reaches people after drinking-water treatment. A laboratory reporting limit is not a health threshold. A compound can be detectable far below a level expected to cause an effect, and a non-detection can still mean a trace amount below the method’s threshold.
In April 2026, EPA released nonregulatory human-health benchmarks for 374 pharmaceuticals that may occur in source water or treated drinking water. The agency provides separate benchmark values for the general population and infants. EPA stresses that these benchmarks are not enforceable limits; they are screening tools that can help determine whether a detected level warrants further investigation.
Should a Louisville homeowner test for pharmaceuticals?
A broad pharmaceutical panel can be expensive and may not answer the question a homeowner actually has. Start with the water source and objective. A Louisville Water customer should review current utility reporting and ask whether the utility has conducted targeted emerging-contaminant monitoring. A private-well owner has a different risk profile and should prioritize common well concerns such as bacteria, nitrate, arsenic where relevant, and local land-use risks before ordering an exotic panel without a reason.
Targeted testing may be reasonable when there is a documented source concern, a research objective, or a need to verify a treatment claim. The laboratory should explain which analytes are included, reporting limits, sample containers, preservation, turnaround, and whether the method is appropriate for drinking water.
Can home treatment reduce pharmaceutical residues?
Activated carbon and reverse osmosis can reduce selected organic compounds, but performance varies widely. NSF/ANSI 401 provides reduction claims for selected emerging compounds on specific products. It does not mean every product with a filter cartridge removes every prescription drug. NSF/ANSI 53 covers specific health-effect claims, and NSF/ANSI 58 applies to reverse-osmosis systems, again based on listed claims.
The exact model matters. So do capacity, flow, water chemistry, cartridge age, membrane condition, and bypass. A small pitcher, an under-sink carbon block, a reverse-osmosis system, and a whole-house carbon tank are not interchangeable simply because they all contain carbon or advertise “advanced filtration.”
Read the ZeroTrace Water guide to PFAS and pharmaceutical filter claims and the point-of-use versus whole-house comparison before buying.
The gain: understand the pathway and choose a measurable response
The real story is unsettling: chemicals designed to change human biology can be measured in the environment after use. The equally real context is that a river-screening detection is not proof that Louisville tap water contains the same concentration or that the detected level caused harm.
A strong response is specific. Dispose of medication responsibly. Follow current source and utility data. Test only when the result will change a decision. Select treatment by verified claim, not by a frightening product label. Then maintain the system so the performance you purchased has a chance to continue. That turns an alarming headline into an accountable water plan.
This article uses source-specific language. River, wastewater, raw-water, finished-water, and tap-water results are not treated as interchangeable. Detection does not automatically establish a health effect, and a regulatory change does not erase toxicology. Information is educational and not medical advice.
Primary sources and further reading
- ORSANCO: Screening Study of Emerging Contaminants in the Ohio River Basin
- ORSANCO study PDF
- U.S. EPA: 2026 Human Health Benchmarks for Pharmaceuticals
- U.S. EPA: Impact of Pharmaceuticals Released to the Environment
Frequently asked questions
Did ORSANCO find prescription drugs in Louisville tap water?
No. The cited study analyzed Ohio River surface-water samples, including an effluent-influenced sample. It was not a test of finished water delivered to homes.
Do detections in nanograms per liter prove a health danger?
No. Detection establishes that a method found a compound above its reporting threshold. Risk assessment also requires toxicology, exposure concentration, duration, route, and uncertainty.
Why should homeowners care if the study is old?
It documents the pathway and shows why modern monitoring matters. EPA’s 2026 benchmarks for 374 pharmaceuticals give agencies a newer screening tool, but those benchmarks are nonregulatory.
