How to Diagnose C. Diff: Tests, Steps, and Results

Diagnosing a Clostridioides difficile (C. diff) infection depends on stool-based laboratory tests, but no single test can confirm the diagnosis on its own. The most widely recommended approach uses a multistep algorithm that combines two or three different tests, each catching what the others miss. Getting from “I suspect C. diff” to “this patient has an active infection” involves clinical judgment about who should even be tested, choosing the right combination of lab assays, and interpreting results that sometimes point in different directions.

When Testing Should Happen

The first step in diagnosing C. diff is deciding whether to order a test at all. Guidelines consistently say testing should only be performed on unformed (liquid or soft) stool. If the stool holds its shape, the lab typically should not run a C. diff assay regardless of other symptoms. This principle matters because a substantial fraction of hospitalized patients carry C. diff in their gut without being sick from it, and testing formed stool from these patients would produce misleading positives that lead to unnecessary treatment.

Symptoms that warrant testing include watery diarrhea (usually three or more unformed stools in 24 hours), abdominal cramping or tenderness, fever, and sometimes nausea. Recent antibiotic use is the most well-known risk factor, though C. diff can also develop in people who haven’t taken antibiotics recently, particularly those who are older, hospitalized, immunocompromised, or on acid-suppressing medications. The clinical picture matters: a patient with a clear alternative explanation for diarrhea, such as laxative use or tube feeding, generally shouldn’t be tested unless other signs point toward infection.

The Main Laboratory Tests

There are several tests a lab can use to detect C. diff, each measuring something slightly different. Understanding what each one actually detects helps make sense of why results sometimes disagree and why doctors combine them.

Glutamate Dehydrogenase (GDH) Screening

GDH is an enzyme produced by all strains of C. diff, whether or not those strains make toxins. A GDH test is fast, cheap, and very sensitive. A meta-analysis comparing GDH to culture found sensitivity above 90% and a negative predictive value high enough to effectively rule out C. diff when the result is negative.¹ One large evaluation reported 100% sensitivity and about 93% specificity against culture.² The catch is that GDH cannot tell you whether the C. diff strain in the stool is actually producing toxins. It picks up both toxin-producing and harmless strains, which is why a positive GDH result always needs a follow-up test.

Toxin Enzyme Immunoassay (EIA)

Toxin EIAs detect the actual toxin proteins (toxin A, toxin B, or both) in a stool sample. Because these tests look for the disease-causing molecules themselves, a positive result is strong evidence of active infection. Early rapid EIAs targeted only toxin A, but newer assays detect both toxins A and B. In a multicenter evaluation, an EIA detecting both toxins achieved about 92% sensitivity and 100% specificity compared to the tissue culture reference test.³ Despite that, toxin EIAs are generally considered to have lower sensitivity than molecular tests, especially at low toxin concentrations. In some clinical settings, sensitivity for toxin EIA can fall into the 50% to 70% range, which means a negative toxin test alone doesn’t reliably rule out infection.⁴

Nucleic Acid Amplification Tests (NAAT/PCR)

PCR-based tests look for the gene that encodes toxin B (called tcdB). They are the most sensitive single tests available. Commercial NAATs generally achieve sensitivity above 92% and specificity above 99%.⁵ One evaluation found PCR had the highest negative predictive value of all tests studied, meaning a negative PCR result is very reassuring.⁶ A separate study reported PCR sensitivity and specificity of about 97.5% and 99.7%, respectively, with same-day turnaround.⁷

The limitation of NAAT is significant, though. Detecting the toxin gene only proves that a toxin-producing strain is present in the stool. It does not prove the strain is actively making toxins and causing disease. A person who carries toxigenic C. diff without symptoms will also test positive by PCR. Used alone, NAAT can lead to overdiagnosis.⁵

The Older Reference Methods

Two tests still serve as benchmarks against which newer tests are measured, even though most clinical labs no longer use them routinely. The cell cytotoxicity assay exposes cultured cells to filtered stool and watches for cell death caused by toxin B; it achieves roughly 93% sensitivity and 100% specificity but requires 48 to 96 hours. Toxigenic culture grows C. diff from stool and then tests the isolate for toxin production; it has excellent accuracy but the same slow turnaround.⁸ Both are impractical for everyday clinical use, so they mainly appear in research settings and as the “gold standard” comparators for validating faster tests.

The Multistep Algorithm

Because every test has blind spots, guidelines from major infectious disease societies recommend combining tests rather than relying on any one alone. The 2017 SHEA/IDSA guideline specifically recommends using a stool toxin test as part of a multistep algorithm rather than NAAT by itself. Three approved combinations are outlined: GDH plus toxin EIA; GDH plus toxin EIA with NAAT to resolve discrepancies; or NAAT plus toxin EIA.⁹

In practice, the most common workflow starts with GDH as a rapid screen. If GDH is negative, C. diff is effectively ruled out and no further testing is needed. If GDH is positive, the lab runs a toxin EIA. If both GDH and toxin are positive, the diagnosis is confirmed. The tricky scenario is when GDH is positive but the toxin EIA comes back negative. That discrepancy gets resolved by running a NAAT: if the PCR is also positive, the patient likely has toxigenic C. diff (though clinical context still matters); if the PCR is negative, infection is unlikely.²

What Discordant Results Mean

The scenario where the organism is detected (by GDH or PCR) but free toxin is not (by EIA) creates a genuine clinical gray zone. A study comparing outcomes in these two groups found that patients with a positive toxin EIA had more severe disease, with about a third presenting with severe or complicated forms, compared to roughly a fifth of those who tested positive only by PCR. Recurrences were also far more common in the toxin-positive group, at about 26% versus 7%.¹⁰ Death rates, however, were the same between the groups. The takeaway for clinicians is that a GDH-positive, toxin-negative, PCR-positive result usually signals a milder situation, but complications can still occur. Treatment decisions in that gray zone hinge on individual symptoms and risk factors, not the lab result alone.

Colonization Versus Infection

A positive lab test does not automatically equal an infection that needs treatment. A meaningful proportion of hospital patients carry toxigenic C. diff without any symptoms. A meta-analysis found that about 8% of patients admitted to hospitals were colonized with toxigenic strains, with that figure reaching 10% in North American studies.¹¹ These colonized patients face about six times the risk of developing active infection compared to non-carriers, but most of them will not get sick during their hospital stay.

This is precisely why testing should be reserved for patients with compatible symptoms. Screening asymptomatic patients inevitably finds carriers and can trigger unnecessary antibiotic courses that paradoxically increase the risk of full-blown C. diff disease. It also creates infection-control headaches: a positive test in a patient without diarrhea might lead to isolation precautions that aren’t needed and that reduce quality of care for the patient.¹²

What Not to Do With Testing

One of the most common mistakes in C. diff diagnosis is repeating the test. Once a properly collected stool sample has been tested with an appropriate algorithm and the result is negative, retesting the same episode of diarrhea within seven days adds almost no useful information. The same goes for “test of cure” after treatment. C. diff organisms and even free toxin can persist in stool for weeks after successful treatment while the patient is improving clinically. Retesting a patient who is getting better will often come back positive and can trigger unnecessary retreatment.¹³ The decision to treat a recurrence should be based on return of symptoms, not on re-running the lab test.

Blood Markers and Severity Assessment

While stool tests confirm the diagnosis, blood and fecal biomarkers help determine how serious the infection is. An elevated white blood cell count (above about 15,000 per microliter) has been associated with longer hospital stays and a roughly three-fold higher risk of in-hospital death in confirmed C. diff patients.¹⁴ Low serum albumin and elevated fecal lactoferrin have both been linked to more severe disease.¹⁵ Procalcitonin, a blood marker often used to gauge bacterial infections, also tracks with C. diff severity: a level above 0.2 ng/mL had about 81% sensitivity and 72% specificity for identifying clinically severe cases, with a negative predictive value around 90%.¹⁶ None of these markers diagnose C. diff on their own, but they help clinicians decide between oral antibiotics, intravenous therapy, or surgical consultation.

When the Scope Goes In

Endoscopy is not a routine part of C. diff diagnosis, but it has a role in specific scenarios. If stool tests come back negative but clinical suspicion remains high, or if the patient is deteriorating rapidly and the team needs an answer before stool results return, a flexible sigmoidoscopy or colonoscopy can provide direct visualization of the colon. The hallmark finding is pseudomembranes: yellowish-white raised plaques scattered across the mucosal surface. These plaques, composed of fibrin, mucus, and inflammatory debris, sit on top of damaged tissue and are highly specific for C. diff colitis when present. Their absence doesn’t rule out the infection, though. Mild cases can show only redness and swelling without pseudomembranes, and the plaques can also appear in the right colon beyond the reach of a sigmoidoscope. Endoscopy also carries risks in patients with severe colitis, including perforation, so it’s generally reserved for cases where the stakes of delayed or missed diagnosis outweigh those risks.

Sample Handling Matters

How a stool sample is stored between collection and testing can affect test accuracy, especially for toxin-based assays. Research shows that C. diff toxins are reasonably stable in stool kept at refrigerator temperature for the first 24 hours, but degrade over time. In patient samples, toxin levels stayed stable within 24 hours under transport conditions.¹⁷ Beyond that window, problems accumulate. Repeated freezing and thawing of samples lowered toxin levels significantly within days.¹⁸ There is also some evidence that toxin A degrades faster than toxin B under storage stress, which could bias results from assays that only target toxin A.¹⁹ PCR-based tests are less vulnerable to these handling issues because DNA is far more stable than protein, but getting samples to the lab promptly remains the simplest way to ensure accuracy across all test types.

The C. Diff Diagnostic Puzzle in Inflammatory Bowel Disease

Patients with inflammatory bowel disease (IBD) face a particularly frustrating diagnostic overlap. The symptoms of a C. diff infection and an IBD flare are nearly identical: worsening diarrhea, abdominal pain, and sometimes fever. Bloody diarrhea skews more toward IBD, but that alone doesn’t settle the question. Making things harder, C. diff is actually one of the most common triggers of an IBD flare, so both conditions can be happening simultaneously.⁴

Because PCR has such high sensitivity, a positive result in an IBD patient doesn’t necessarily mean C. diff is the cause of their symptoms. The patient might simply be a colonizer whose flare has a completely different cause. And because symptoms can’t reliably distinguish flare from infection, the positive predictive value of PCR alone drops in this population. The two-step approach (GDH followed by toxin EIA) is particularly valuable here, because a positive toxin EIA provides stronger evidence that C. diff is actively contributing to the patient’s illness.²⁰ Clinicians managing IBD patients are generally advised to test for C. diff any time symptoms worsen, even if the flare seems “typical.”

Pediatric Testing Pitfalls

Testing children for C. diff comes with a unique wrinkle: very young children carry C. diff at high rates without getting sick from it. Colonization rates in infants can be strikingly high, possibly because their immature intestinal lining lacks the receptors that C. diff toxins bind to in adults. The same tests used in adults are used in children, but the interpretation changes. A positive NAAT in a baby under 12 months old almost certainly reflects colonization rather than disease, which is why most guidelines recommend against testing in that age group unless there’s no other explanation for symptoms. Even in toddlers and young children, the high background colonization rate means positive tests need to be weighed carefully against the clinical picture.²¹ Limited data exist on how the performance characteristics of specific assays change in pediatric populations, so clinicians often extrapolate from adult validation studies.

Hypervirulent Strains and Molecular Typing

Not all C. diff strains behave the same way. The strain known as BI/NAP1/027 (sometimes called ribotype 027) has been linked to outbreaks with higher infection rates and more severe disease, and some treatments may work differently against it.²² This strain produces higher levels of toxins and sometimes produces an additional toxin (called binary toxin) that standard diagnostic tests don’t specifically look for.

Routine clinical testing does not typically identify which strain is involved. Strain typing requires specialized molecular methods and is usually done by reference laboratories for surveillance or outbreak investigation rather than for individual patient care. Newer PCR assays are being developed that can detect the 027 strain specifically, including one that uses a unique gene target to identify it directly from stool samples.²³ Whether routine strain identification would change day-to-day treatment decisions is still debated, but knowing you’re dealing with 027 can prompt more aggressive monitoring and earlier escalation of therapy.

Emerging Diagnostic Approaches

Researchers are working on tests that could make C. diff diagnosis faster and more accessible, particularly in settings without full microbiology labs. Point-of-care PCR systems that healthcare workers can run directly on hospital wards have already been validated, achieving greater than 98% agreement with central laboratory results and cutting turnaround time dramatically.²⁴

An even more experimental approach involves analyzing volatile organic compounds, the chemicals responsible for the distinctive smell of C. diff stool. A proof-of-concept study used mass spectrometry to detect VOC patterns not only in stool but also in breath and blood plasma of C. diff patients, achieving area-under-the-curve values of 93%, 86%, and 91% for breath, stool, and plasma, respectively.²⁵ Electronic nose devices paired with machine-learning algorithms are being explored to turn these smell patterns into practical diagnostic tools.²⁶ A different approach called multiple cross displacement amplification (MCDA) can detect toxigenic C. diff DNA in about 16 minutes on average, matching PCR sensitivity while being simple enough for resource-limited settings.²⁷ None of these technologies have replaced the standard multistep algorithm yet, but they represent genuinely different ways of solving the same problem: getting an accurate answer faster, with less infrastructure.