When I got some upper respiratory infection the doctor always prescribed broad-spectrum antibiotics like amoxicillin. Is there any test to identify bacteria, and if so, why doesn't the doctor order one? Is it prohibitively expensive (or slow)?

Actually they always do some stupid blood test (just some cell count), which reveals nothing about the cause of infection (they told me they were not sure. I'm in China and correct me if I'm wrong).

2 Answers 2


First, I will explain what tests DO tell you the bacteria, and how they are used. Secondly, I will explain why clinicians rarely do that for respiratory infections.

To test for bacteria, a sample can be taken of the body tissue/fluid that is infected - blood for bacteremia, urine for bladder/kidney infection, pus for abscess, spinal fluid for meningitis, sputum for respiratory infection, etc.

Usually, soon after taking the sample, treatment is started with "broad spectrum antibiotics," rather than waiting for test results while the infection grows. This is called "Empiric Therapy." Then, when results come back, the antibiotic choice can be "narrowed" in to target the likely bacteria.

  • Gram Stain can give a hint to the bacteria type, which is sometimes enough to assume what it is.

  • Bacterial cultures not only tell you the specific bacteria (usually), it can test for which exact antibiotics kill it the best. That is called antibiotic sensitivities.

Commonly, the first test to come back is the Gram Stain, which can be minutes to hours. For example, "gram positive cocci" highly suggests Staphylococcus or Streptococcus. In this case, we could narrow the antibiotic spectrum to something that kills those types of bacteria.

Then, 24-72 hours later is usually the earliest that culture results are available. For example, it would tell us (1) Staphylococcus aureus and (2) Resistant to Methicillin but Sensitive to Vancomycin, Clindamycin... (other representative antibiotics). That would define this bacteria as Methicillin-Resistant Staph Aureus (MRSA). Then we could further narrow the antibiotic to Vancomycin, Clindamycin, etc depending on what kind of infection it is and a bunch of other factors.


The respiratory tract (including your mouth) is colonized with many different kinds of bacteria. The bacteria in the phlegm from your lungs will show up in a gram stain or culture, but so will a lot of the bacteria from your mouth. How then would you know which is causing the infection?

Instead, usually, a clinician will go straight to prescribing an antibiotic that kills the most common bacteria that cause your specific signs/symptoms. Empiric therapy. Then if it doesn't work, they may investigate further.

Clinicians often take cultures for (1) serious infections like blood infections (2) suspected uncommon or dangerous pathogens (like tuberculosis) that AREN'T oral colonization bacteria or (3) sometimes after a common antibiotic has failed.


I found some resources on Laboratory Diagnosis of Infectious Disease, from Merck Manual. In practice, there is urgent need to start treatment before identification of the cause, so lab test will not be widely carried out (together with consideration on cost and potential benefits).

  • Whether treatment needs to begin before identification depends on the patient's circumstances. That's not always true as you state.
    – Carey Gregory
    Commented Feb 19, 2017 at 23:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.