Someone with persistent diarrhoea, blood in the stool or unexplained abdominal pain may be offered several quite different stool tests, sometimes on the same sample. One looks for bacteria that can be grown. One looks for genetic material from a long list of organisms. One measures a protein released by inflamed intestinal tissue. One looks for traces of blood. They answer different questions, and the results can be discordant without any of them being wrong.
The most significant change in this area over the past decade has been the shift from culture to multiplex molecular panels. Culture took days, detected a narrow range of organisms, and missed a great deal. Panels return results the same day and detect a much wider range, including viruses and parasites that culture never addressed.
That gain has come with a genuinely difficult new problem. The panels are so sensitive that they detect organisms present in small numbers, including ones a person may carry without illness, and ones left over from an infection that resolved weeks ago. A positive result now means the organism’s genetic material is present. Whether it is the cause of the current symptoms is a separate question that the test cannot answer.
Key takeaways
- Culture detects living organisms and yields an isolate for susceptibility testing; molecular panels detect genetic material whether or not it is viable.
- Panels frequently return multiple positives, and deciding which one matters requires clinical judgement rather than the report alone.
- Faecal calprotectin distinguishes inflammatory from functional bowel disease well and does not identify a cause.
- Immunochemical occult blood tests are specific to human haemoglobin and do not require dietary restriction, unlike older chemical methods.
- Direct-to-consumer microbiome reports are not diagnostic tests and should not displace investigation of persistent symptoms.
What Traditional Stool Culture Detects
Stool culture is a targeted search, not a survey. The gut contains an enormous and dense bacterial population, the vast majority of it harmless, so plating a sample onto a general growth medium yields an unreadable lawn of normal flora. Culture works by using selective and differential media designed to suppress that background and reveal a specific short list of pathogens.
Selective media contain agents such as bile salts, dyes or antibiotics that inhibit the growth of ordinary gut bacteria while permitting the target. Differential media contain indicators, such as a sugar plus a pH dye, that make colonies of the target look visibly different from anything else that grows. A laboratory therefore runs several plates per sample, each aimed at a particular group, and incubates some under conditions the target prefers, such as reduced oxygen or higher temperature.
The consequence is that culture finds what it was asked to find. A routine stool culture in most laboratories covers a small number of common bacterial causes of infectious diarrhoea. Anything outside that list requires a specific request, and the laboratory has no way of knowing to look unless the clinical details prompt it. Requests that mention recent travel, contact with animals, shellfish consumption, immune suppression or a hospital stay change what is set up, which is why those details on the form are not administrative padding.
Culture retains two advantages that molecular methods cannot replicate. It yields a living isolate, which can be tested for antimicrobial susceptibility, an increasingly important consideration as resistance spreads. And that isolate can be typed for outbreak investigation, allowing public health laboratories to determine whether cases in different places share a source. For those reasons, culture has not been abandoned; it has been repositioned as a follow-up to a positive molecular result rather than the first-line test.
Multiplex PCR Panels and Their Speed
A multiplex panel extracts nucleic acid from the sample and simultaneously tests for many targets in a single reaction or cartridge, each target identified by its own primer and probe set with a distinguishable signal.
The practical difference is dramatic. A single panel typically covers bacterial pathogens, several diarrhoea-causing viruses, and the common protozoan parasites, returning results in a few hours rather than days. It replaces what would previously have required a bacterial culture, a separate viral test, and a microscopist examining concentrated stool for parasites, each with its own request, its own turnaround and its own chance of not being ordered.
Sensitivity is substantially higher across the board. Molecular detection does not require the organism to be alive, to grow on the chosen medium, or to be present in sufficient numbers for a microscopist to see it. This matters most for organisms that were always difficult: fastidious bacteria that grow poorly, viruses that most laboratories never tested for routinely, and parasites whose shedding is intermittent.
| Attribute | Culture | Multiplex PCR panel |
|---|---|---|
| Time to result | Two to four days typically | Hours |
| Range of organisms | Narrow, request dependent | Broad, fixed panel |
| Detects viruses and parasites | No | Yes |
| Requires viable organism | Yes | No |
| Yields isolate for susceptibility | Yes | No |
| Suitable for outbreak typing | Yes | Not directly |
| Multiple positives | Uncommon | Common |
| Distinguishes carriage from disease | Partially | No |
The last two rows are where the trouble lies, and they deserve a section of their own.
Detection Versus Causation in Molecular Panels
A molecular panel answers a narrow question: is the genetic material of this organism present in this sample above the assay’s detection threshold? Everything else is inference.
Several situations produce a positive result in someone whose symptoms have another cause. Asymptomatic carriage is common for several organisms on standard panels, particularly in young children, in people recently exposed in a household outbreak, and in some populations where certain protozoa are frequently carried without illness. Post-infection shedding continues for weeks after symptoms resolve for several pathogens, so a person whose diarrhoea last month was caused by an organism may still test positive for it while their current symptoms come from something else entirely. And nucleic acid from dead organisms is detected as readily as from live ones, so a treated infection can remain positive after successful treatment.
Multiple positives are the routine consequence. It is not unusual for a panel to report two or three organisms, and the report offers no ranking. Interpreting that requires knowledge the assay does not have: the clinical picture, the duration and character of symptoms, the person’s age, exposures and travel, immune status, and whether the detected organisms are ones that typically cause the observed illness. A detection of a classic invasive pathogen in someone with bloody diarrhoea and fever is straightforward. A detection of a commonly carried organism in someone with three months of bloating is not.
The particular case that has caused most difficulty in practice concerns toxin-producing bacteria acquired in healthcare settings, where colonisation without disease is common and molecular detection of the toxin gene does not establish that toxin is being produced. Laboratories have responded with algorithms combining a sensitive screening step with a test for the toxin protein itself, precisely because the molecular result alone over-diagnoses.
Three practical implications follow. Test people with symptoms rather than screening the asymptomatic. Do not use a panel as a test of cure, because it will remain positive after successful treatment. And read the report as a list of what was found, not a diagnosis.
Faecal Calprotectin and Inflammation
Calprotectin is a protein abundant in neutrophils. When neutrophils migrate into the intestinal wall and lumen during inflammation, calprotectin ends up in the stool, where it is unusually stable and can be measured by immunoassay. Its concentration correlates broadly with the degree of neutrophil traffic into the gut.
Its clinical value is as a discriminator. The common problem in gastroenterology is distinguishing inflammatory bowel disease from irritable bowel syndrome, which can produce similar symptoms but involves no mucosal inflammation. Calprotectin does this well. A low result makes significant inflammation unlikely and supports a functional diagnosis without endoscopy; a clearly raised result makes inflammation likely and supports proceeding to endoscopy. Given that endoscopy is invasive, uncomfortable and resource-intensive, a non-invasive test that reliably identifies who needs it has substantial value.
What calprotectin does not do is identify a cause. It rises in inflammatory bowel disease, in bacterial gastroenteritis, in inflammation caused by non-steroidal anti-inflammatory drugs, in diverticulitis, in colorectal cancer and in coeliac disease with mucosal damage. A raised result says the intestinal lining is inflamed and that further investigation is needed; it does not say why.
Interpretation has grey zones. Results are usually reported against a cut-off, with an intermediate band where the answer is genuinely uncertain and repeat testing after a few weeks is often more useful than acting immediately. Recent non-steroidal anti-inflammatory use raises values and should be asked about. Values are higher in infants for physiological reasons, so paediatric interpretation differs.
Occult Blood Testing Methods Compared
Occult blood means blood present in quantities too small to see. Detecting it is the basis of most population screening for colorectal cancer, on the reasoning that tumours and large polyps bleed intermittently long before they cause symptoms.
The older chemical method, the guaiac test, relies on the peroxidase-like activity of haem. A sample is applied to guaiac-impregnated paper, hydrogen peroxide is added, and haem catalyses a reaction turning the paper blue. It is cheap and simple, and it has two significant weaknesses. It detects peroxidase activity, not human blood specifically, so dietary haem from red meat and peroxidases from certain raw vegetables produce false positives, which is why the test required dietary restriction beforehand. And it detects blood from anywhere in the digestive tract, including the stomach and upper intestine, where the source is more often benign.
The immunochemical test, usually abbreviated as a faecal immunochemical test, uses antibodies against human haemoglobin. This solves both problems at once. It does not react to animal haem or plant peroxidases, so no dietary restriction is needed, and because haemoglobin from an upper gastrointestinal bleed is largely degraded during transit, the antibodies detect predominantly lower gastrointestinal bleeding, which is where screening is aimed.
| Feature | Guaiac chemical test | Immunochemical test |
|---|---|---|
| What it detects | Peroxidase activity of haem | Intact human haemoglobin |
| Dietary restriction | Required | Not required |
| Affected by animal blood in diet | Yes | No |
| Region of bleeding detected | Whole tract | Predominantly lower tract |
| Result type | Visual, positive or negative | Quantitative concentration |
| Samples typically needed | Several, over separate days | Usually one |
Two limitations apply to any occult blood test. Bleeding from tumours and polyps is intermittent, so a negative result on one occasion does not exclude a lesion, which is why screening is repeated at intervals rather than performed once. And a positive result is not a diagnosis of cancer; the great majority of positives are explained by haemorrhoids, polyps, inflammation or other benign causes, and the test’s role is to select who needs a colonoscopy.
Parasite Detection and Sample Timing
Classic examination for ova, cysts and parasites means a trained microscopist examining a concentrated, stained preparation for identifiable structures. It is skilled work, its sensitivity depends heavily on the examiner’s experience, and its greatest weakness is intermittent shedding. Many parasites are not released continuously; a person can be infected and shed nothing detectable on a given day. This is the reason for the traditional instruction to submit samples collected on separate days, usually three, rather than one large sample. Multiple samples across several days sample the shedding pattern rather than a single point in it.
Molecular detection of the common protozoa is now included on most gastrointestinal panels, and it substantially outperforms microscopy for those specific organisms, being less dependent on operator skill and less affected by low shedding. What it cannot do is detect anything outside its target list. Helminth eggs, less common protozoa and unexpected organisms are found only by someone looking down a microscope, which is why microscopy remains the correct request for a patient with relevant travel or exposure history and a negative panel.
For any of these, the practical instructions matter. Use the container provided, since the preservative differs by test. Do not contaminate the sample with urine or water from the toilet bowl, which dilutes it and may introduce free-living organisms that confuse interpretation. Collect before starting any antiparasitic or antimicrobial treatment. And note that some radiological contrast agents and some antacids interfere with microscopic examination for a period afterwards.
Microbiome Tests Sold Direct to Consumers
Alongside all of this sits a growing consumer market selling stool sequencing with a report describing the bacterial composition of the gut, often with dietary recommendations and a score of some kind. These are not diagnostic tests, and the distinction is important for anyone with symptoms.
The underlying measurement is generally sound. Sequencing a conserved bacterial gene, or sequencing all the DNA in the sample, genuinely characterises which organisms are present and in what proportions. The science of the gut microbiome is real, active and interesting, and associations between microbial composition and various conditions are well documented.
The interpretation layer is where the difficulty lies. There is no established definition of a healthy microbiome against which an individual can be scored, because composition varies enormously between healthy people according to diet, geography, age, medication history and genetics. Two people can have markedly different profiles and both be entirely well. Comparisons against a reference population therefore describe difference, not abnormality.
Most importantly, associations between composition and disease are population-level and largely correlational. Knowing that a group with a condition tends to have less of a particular organism does not establish that increasing it in an individual will help, and the direction of causation is frequently unresolved.
The practical risks are twofold. Unnecessary and sometimes restrictive dietary changes may be adopted on the strength of a report with no diagnostic validity, and a reassuring or explanatory-sounding microbiome result may delay proper investigation of symptoms that warrant it. Persistent change in bowel habit, blood in the stool, unintended weight loss, nocturnal symptoms, anaemia or a family history of bowel disease or cancer are reasons to see a clinician, and no consumer sequencing report addresses any of them.
Frequently asked questions
My PCR panel found two organisms. Which one is making me ill?
The panel cannot tell you, and the report is not ranked. Deciding requires the clinical picture: what your symptoms are and how long they have lasted, your age, recent travel, exposures, antibiotic use and immune status, and whether the organisms detected typically cause that pattern of illness. Some findings represent carriage or shedding left over from an earlier infection rather than current disease. This is exactly the judgement a clinician makes with the report in front of them, and it is why panels are interpreted rather than simply read.
Why did the laboratory culture my sample after the PCR was already positive?
Because the two tests provide different things. The molecular result identifies what is present quickly, but yields no organism. Culture grows the bacterium so that its antibiotic susceptibility can be determined and so that public health laboratories can type it and link cases to a common source during an outbreak. Where treatment choice depends on susceptibility, or where an outbreak is suspected, the extra days that culture takes are worth spending.
Does a normal faecal calprotectin rule out bowel disease?
It makes significant intestinal inflammation unlikely, which is genuinely useful, but it does not exclude everything. Calprotectin reflects neutrophil traffic into the gut, so conditions that do not produce that traffic can coexist with a normal result. Coeliac disease, bile acid malabsorption, some cases of limited or early inflammatory disease, and colorectal lesions that are not inflamed may all give low values. Symptoms that persist despite a normal result still warrant investigation.
Why does an occult blood test need no dietary preparation any more?
Because the immunochemical method uses antibodies specific to human haemoglobin rather than detecting the chemical activity of haem in general. The older guaiac method reacted to animal blood in food and to peroxidases in some raw vegetables, so dietary restriction was needed to avoid false positives. Antibodies do not recognise those substances, so ordinary eating has no effect on the result, which improves both accuracy and the proportion of people who complete the test.
Should I still collect three separate samples for parasites?
Follow whatever the laboratory instructs, because it depends on which test they will run. Where classical microscopy for ova, cysts and parasites is being performed, samples on separate days remain the standard approach, since shedding is intermittent and a single sample can easily miss an infection. Where a molecular panel covering the common protozoa is used, one sample is generally sufficient for those specific organisms, though microscopy on multiple samples may still be requested if the travel or exposure history points to something outside the panel.
The useful way to hold all of this together is to ask what question each test answers. Culture asks whether a growable pathogen is present and what will kill it. A molecular panel asks whether an organism’s genetic material is there, quickly and broadly, and leaves causation to be worked out. Calprotectin asks whether the bowel lining is inflamed, without saying why. Occult blood asks whether there is bleeding worth investigating. A consumer microbiome report asks none of these and answers none of them. Matching the test to the question, and providing the clinical details that determine what the laboratory sets up, does more for the eventual answer than any single result on the page.
This is education, not medical advice. Laboratory results only carry meaning alongside your symptoms, history and examination. Talk to a qualified clinician about your own results before changing anything about your care or supplements.
