How to Test for Bacterial Infection: Blood, Culture & More

Testing for a bacterial infection usually starts with simple blood work and progresses to more targeted methods depending on where the infection is suspected. A complete blood count can flag that your immune system is fighting something, inflammatory markers like C-reactive protein can help distinguish bacterial from viral causes, and cultures remain the gold standard for identifying exactly which bacterium is responsible. But the testing landscape extends well beyond these basics, and understanding what each test does and does not tell you can make a real difference in how quickly you get the right treatment.

Blood Tests That Signal Infection

When your doctor suspects a bacterial infection, a complete blood count is usually one of the first orders. White blood cells, especially neutrophils, ramp up production in response to bacteria. During an active bacterial infection, neutrophil counts go through a characteristic pattern: an initial dip in the first hours as cells rush to the infection site, followed by a rebound where counts climb above normal, then a gradual return to baseline as the infection resolves.¹ A key feature labs look for is the “left shift,” meaning immature neutrophils are being released from bone marrow faster than usual. While other conditions can nudge immature cells into circulation, the sudden and dramatic shift seen in bacterial infection is fairly distinctive.²

A white blood cell count alone cannot tell you whether an infection is bacterial or viral. That is where inflammatory biomarkers come in. C-reactive protein (CRP) is a protein your liver produces in response to inflammation, and it rises sharply during bacterial infections. Procalcitonin (PCT) is another marker that tends to climb with bacterial but not viral illness. In a large study across Southeast Asia, CRP outperformed procalcitonin for distinguishing bacterial from viral infections, with CRP at a threshold of 20 mg/L catching about 86% of bacterial cases while correctly ruling out about two-thirds of non-bacterial ones.³ A systematic review and meta-analysis, however, found procalcitonin had slightly higher sensitivity for picking up bacterial infections, around 92% compared to 86% for CRP, with similar specificity for both.⁴ In pediatric patients, procalcitonin showed a specificity of 93% when a higher cutoff was used, meaning very few children with viral infections had falsely elevated levels.⁵

The takeaway is that neither CRP nor procalcitonin is a definitive answer on its own. They are screening tools that help your doctor decide whether to order more targeted testing or start empiric antibiotics while waiting for culture results.

Blood Cultures and Why the Bottles Matter

If your doctor suspects bacteria have entered your bloodstream, blood cultures are the definitive test. A nurse draws blood from a vein (sometimes from two separate sites to improve accuracy) and injects it into special bottles that encourage bacterial growth. These bottles go into an automated incubator that monitors them continuously, flagging any that turn positive. Most bacteria will be detected within 24 to 48 hours, though some slower-growing organisms take longer.

You will typically have blood drawn into at least two types of bottles: one aerobic (designed for bacteria that thrive in oxygen) and one anaerobic (for bacteria that grow without it). This pairing matters more than you might expect. In one evaluation of over 158,000 bottles, the anaerobic bottles recovered Staphylococcus aureus at a statistically higher rate than aerobic bottles and detected it faster. If only aerobic bottles had been collected, over 2,000 fewer positive cultures would have been detected.⁶ Meanwhile, aerobic bottles are better at recovering certain other organisms, including most fungi and some gram-negative bacteria like Pseudomonas.⁷ The two bottle types complement each other, and skipping either one means missing infections that would otherwise be caught.

Gram Stain and the Value of a Quick Look

Once a blood culture or other specimen turns positive, or when a sample arrives from a wound, sputum, or body fluid, a Gram stain is often the first thing performed. The technique, which uses a pair of dyes to classify bacteria by the structure of their cell walls, takes only minutes and divides organisms into two broad camps: gram-positive (which retain a violet stain) and gram-negative (which pick up a pink counterstain). That single piece of information can narrow the list of possible culprits dramatically.

For critically ill patients, the Gram stain’s speed is its greatest asset. In ventilator-associated pneumonia, for instance, Gram stain results can guide initial antibiotic choices before culture results return, helping prevent the use of inappropriate drugs during the hours or days it takes for a full identification.⁸ It is not a replacement for culture, which tells you the exact species and what antibiotics will kill it. But it buys time and narrows the guesswork.

Urine Dipsticks Versus Urine Culture

For suspected urinary tract infections, the quickest screening tool is the urine dipstick. It checks for two things associated with bacteria: leukocyte esterase (an enzyme released by white blood cells fighting infection) and nitrite (a byproduct of certain bacteria). In older adults, dipsticks performed reasonably well, with a combined sensitivity above 90% in one study.⁹ But the dipstick has a serious blind spot: among ambulatory women with suspected UTIs, nearly one in five urine samples with significant bacterial growth tested negative on the dipstick.¹⁰

That is why a urine culture remains the gold standard when the dipstick is negative but symptoms are convincing, or when the infection keeps coming back, or when there is concern about antibiotic resistance. The culture identifies the exact bacterium and tells your doctor which antibiotics it responds to. A positive dipstick with classic symptoms may be enough to start treatment, but relying on the dipstick alone to rule out infection is risky.

Cerebrospinal Fluid and Other Site-Specific Samples

When bacterial meningitis is suspected, doctors need a sample of the cerebrospinal fluid that surrounds the brain and spinal cord. A lumbar puncture, where a needle is inserted into the lower spine, is the standard way to get it. The fluid is then analyzed for white blood cell counts, protein and glucose levels, Gram stain, culture, and increasingly, molecular panels that can identify pathogens and resistance genes in about an hour.¹¹ Lumbar puncture remains central to the diagnosis of meningitis.¹²

Timing matters enormously here. After antibiotics are given, cerebrospinal fluid cultures can turn negative surprisingly fast. In meningococcal meningitis, cultures sterilized within two hours of receiving a third-generation cephalosporin. Pneumococcal cultures took a bit longer, with most turning negative within four to ten hours.¹³ This is why, whenever possible, cultures should be drawn before the first dose of antibiotics. The same principle applies to other site-specific samples: joint fluid, wound swabs, abscess drainage, and sputum all yield the best results before treatment begins.

How Antibiotics Affect Test Results

Starting antibiotics before cultures are collected reduces the chance of growing the responsible organism. A retrospective study found that patients who had received antibiotics before blood cultures were drawn had roughly half the odds of a positive culture compared to those who had not been pre-treated.¹⁴ This does not mean you should delay lifesaving antibiotics for the sake of a culture. In serious infections like meningitis and sepsis, antibiotics should be given immediately. But if there is time to draw cultures first, even minutes before the first dose, it substantially improves your chances of getting a diagnosis.

When the Culture Grows a Contaminant

Not every positive blood culture means you have a bloodstream infection. Skin bacteria, particularly coagulase-negative staphylococci (CoNS), are the most common contaminants in blood cultures. In one study, CoNS were isolated from about 9% of blood samples, but only about a quarter of those represented a true infection. The vast majority, around 69%, were contaminants picked up during the blood draw.¹⁵ Another study found a contamination rate of nearly 73% among CoNS-positive blood cultures.¹⁶

Doctors weigh several factors before treating a CoNS-positive culture: Is the same organism growing in multiple bottles drawn from different sites? Does the patient have an indwelling catheter or prosthetic device that CoNS commonly colonize? Does the clinical picture fit a bloodstream infection? If only one set of cultures is positive and the patient looks well, contamination is far more likely than true infection. This is one reason many hospitals draw blood cultures from two separate venipuncture sites.

Rapid Molecular Panels

Waiting 24 to 72 hours for a culture result is a long time when someone is critically ill. Multiplex PCR panels have compressed the identification step dramatically. One widely used system can identify more than 25 common pathogens and four antibiotic resistance genes from a positive blood culture bottle in about an hour.¹⁷ Similar panels exist for respiratory infections, meningitis, and gastrointestinal pathogens.

The clinical impact is tangible. At one center, using a newer-generation blood culture identification panel cut the time to optimal antibiotic therapy from about 29 hours to about 17 hours.¹⁸ A separate study found that pairing rapid diagnostics with an antimicrobial stewardship team reduced the time to effective therapy from roughly 90 hours down to about 32 hours in patients with antibiotic-resistant gram-negative infections.¹⁹ These panels have also been evaluated for detecting resistance genes directly from surgical drainage fluids, showing strong consistency with conventional susceptibility testing.²⁰

PCR panels do have limitations. They can only detect what they are designed to look for, so an unusual organism not on the panel will be missed. And because PCR amplifies DNA, it can detect dead bacteria or residual genetic material after the infection has cleared, making interpretation trickier in patients already on antibiotics.

MALDI-TOF Mass Spectrometry

Another technology that has quietly transformed microbiology labs is MALDI-TOF mass spectrometry. Once a colony grows on a culture plate, a small smear of it is placed on a target slide, hit with a laser, and the resulting pattern of protein fragments is matched against a database. The whole process takes minutes and identifies the species with high accuracy. It is faster and cheaper per test than traditional biochemical identification methods, which could take a day or more.²¹

Researchers are now pushing MALDI-TOF beyond identification into susceptibility testing. A rapid assay that exposes bacteria to antibiotics for six hours and then uses MALDI-TOF to check for growth showed 96% agreement with standard methods across multiple drug classes.²² If these approaches become routine, same-day susceptibility results from a positive culture could become the norm rather than the exception.

Point-of-Care Antigen Tests

For certain infections, a simple antigen test can provide an answer at the bedside. Urinary antigen tests for Streptococcus pneumoniae and Legionella are the best-known examples in bacterial diagnostics. These work by detecting pieces of the bacterial cell wall that are excreted in urine during infection. For pneumococcal pneumonia, the clinical sensitivity of these tests ranges from about 88% to 94%, with specificity above 96%.²³ They are fast, non-invasive, and can confirm a diagnosis even after antibiotics have been started, since antigen shedding can persist for days to weeks.

The trade-off is narrow scope. These tests look for one or two specific pathogens and nothing else. A negative Legionella antigen test does not rule out other causes of pneumonia. And most urinary antigen tests detect only Legionella serogroup 1, missing the less common serogroups. They are useful additions to the diagnostic toolbox, not replacements for culture or molecular testing.

Serological Testing and When It Is Necessary

Some bacterial infections are difficult or impractical to diagnose by growing the organism. Lyme disease is a classic example: the spirochete Borrelia burgdorferi is slow-growing and present in low numbers in blood, making direct detection unreliable. Diagnosis therefore relies on serological testing, which measures your immune system’s antibody response to the bacterium rather than detecting the bacterium itself.²⁴ Syphilis, brucellosis, and certain rickettsial infections are diagnosed similarly.

The downside of serology is timing. Antibodies take days to weeks to develop after exposure, so testing too early in the illness can produce a false negative. And antibodies can linger long after the infection has cleared, making it hard to distinguish a past infection from a current one. For this reason, doctors sometimes order paired samples a few weeks apart to look for a rising antibody level, which is more convincing evidence of an active infection.

Antimicrobial Susceptibility Testing

Identifying the bacterium is only half the job. Knowing which antibiotics it is susceptible to is equally critical, especially with rising antibiotic resistance. The standard approach is to expose the cultured organism to a panel of antibiotics and see which ones inhibit its growth. The most familiar version is the disk diffusion test, where antibiotic-impregnated paper disks are placed on an agar plate seeded with the bacterium, and zones of no growth around each disk are measured. Broth microdilution, which tests a range of antibiotic concentrations in liquid, is considered the reference method for determining the minimum concentration needed to stop growth.²⁵

Newer rapid phenotypic platforms are aiming to shrink the timeline. One such system delivered susceptibility results with a potential median time savings of about 42 hours compared to standard methods for gram-negative bloodstream infections, with roughly half of tested patients seeing a potential change in antibiotic management as a result.²⁶ Getting susceptibility data faster means patients can be switched from broad-spectrum antibiotics to targeted ones sooner, which reduces side effects and helps slow the development of resistance.

Special Challenges in Infants and Children

Diagnosing bloodstream infections in young children is harder than in adults for a straightforward reason: you can draw less blood. Babies have small total blood volumes, and the amount that can safely be taken for cultures is limited. The problem is that children frequently have low concentrations of bacteria circulating in their blood. In one study, about 60% of true bloodstream infections in children involved fewer than 10 colony-forming units per milliliter, and nearly a quarter had concentrations so low that only one colony-forming unit per milliliter was present.²⁷ That same study found that roughly 59% of septic episodes were associated with at least one false-negative culture from one of the devices used, underscoring how easy it is to miss real infections when blood volumes are small.

The sensitivity of a blood culture is directly tied to the volume of blood inoculated, and low-level bacteremia in children is more common than many clinicians assume.²⁸ Current recommendations call for culturing a higher proportion of total blood volume in pediatric patients than in adults, but in practice, drawing enough blood from a screaming, dehydrated infant remains one of the most persistent challenges in pediatric infectious disease.

Colonization Versus True Infection

A positive culture from a non-sterile site, such as the skin, throat, or respiratory tract, does not always mean you have an infection. Many bacteria live on and in you without causing disease. Conventional diagnostic methods often struggle to tell the difference between harmless colonization and an actual infection causing symptoms, a problem that contributes to unnecessary antibiotic use.²⁹ This is why clinical context matters so much. A throat swab growing Staphylococcus aureus in someone without symptoms is colonization; the same organism in a blood culture from someone with a fever and low blood pressure is a very different story.

Emerging Approaches on the Horizon

Two newer diagnostic strategies are gaining traction. Host gene expression profiling takes a different approach entirely: instead of looking for the pathogen, it measures how your own immune system is responding. Researchers have identified small sets of human genes whose activity patterns differ depending on whether an infection is bacterial, viral, or absent. One machine-learning model using 100 host genes achieved a sensitivity of 93% and specificity of 96% for identifying bacterial infections, with comparable performance for viral infections.³⁰ In hospitalized children, a two-gene signature discriminated bacterial from viral infections with strong accuracy across both discovery and validation groups.³¹

Metagenomic sequencing takes a complementary path, reading all the DNA in a clinical sample to identify every organism present without needing to grow anything in culture. This approach proved especially useful in febrile urinary tract infections where standard urine cultures came back negative: metagenomic sequencing identified the same pathogen that had been found in corresponding blood cultures in the majority of culture-negative urine samples.³² Neither host gene expression panels nor metagenomic sequencing has replaced conventional culture in routine practice yet, but both represent genuine shifts in how the field is thinking about diagnosis, particularly for patients whose infections refuse to grow in a bottle.