Every year, doctors worldwide write roughly 7.7 billion courses of oral antibiotics — and by the most conservative estimates, at least 30 percent of those prescriptions are unnecessary. In the United States alone, the CDC reports that approximately 47 million unnecessary antibiotic courses are prescribed annually, mostly for respiratory conditions that antibiotics cannot treat: common colds, most sore throats, influenza, acute bronchitis, and many sinus infections.

The consequences are not abstract. Every unnecessary course disrupts the gut microbiome, increases the risk of drug-resistant infections, and exposes the patient to side effects ranging from diarrhea and yeast infections to serious complications like Clostridioides difficile colitis. A 2022 analysis in The Lancet estimated that bacterial antimicrobial resistance was directly responsible for 1.27 million deaths worldwide in 2019 and contributed to 4.95 million more.

This article is not about the science of resistance — we cover that in our article on antibiotic resistance and superbugs. Instead, this is a practical guide: when do you actually need an antibiotic, when are you taking one for nothing, and how should you take it correctly?

Bacteria vs. Viruses: Why the Distinction Is Everything

Antibiotics kill bacteria or stop them from multiplying. They do absolutely nothing to viruses — not partially, not a little. Antibiotics target structures that exist only in bacterial cells: cell wall synthesis, bacterial ribosomes, DNA gyrase. Viruses lack all of these. They hijack human cells to replicate, and no antibiotic can touch them.

This matters because the vast majority of acute infections that send people to the doctor are viral. A 2018 systematic review in the British Medical Journal estimated that 70 to 80 percent of upper respiratory tract infections in primary care are caused by viruses. These infections produce the exact symptoms people associate with needing antibiotics: sore throat, congestion, cough, fever, and fatigue. The problem is that many bacterial infections produce similar symptoms, making it difficult — without testing — to distinguish the two. This ambiguity drives overprescription: when in doubt, doctors and patients default to "just in case," a habit that consistently does more harm than good.

When Antibiotics Will NOT Help

Guidelines from the CDC, NICE, and the European Society of Clinical Microbiology are consistent on which common conditions do not benefit from antibiotics.

The common cold. Caused by over 200 viruses, most commonly rhinoviruses. A 2013 Cochrane review of nine trials (2,249 participants) found no benefit of antibiotics over placebo — no shorter symptoms, no fewer complications, but significantly more side effects. Green or yellow mucus does not indicate bacterial infection; it reflects normal immune activity during viral illness.

Influenza. A viral illness. Antibiotics are useless. The appropriate antivirals (oseltamivir, baloxavir), when indicated, should be started within 48 hours of symptom onset. Antibiotics matter only if secondary bacterial pneumonia develops.

Most sore throats. Approximately 85 to 95 percent of adult sore throats are viral, according to a 2016 review in JAMA Internal Medicine. The exception is Group A Streptococcus (5 to 15 percent of adult cases, up to 30 percent in children), which a rapid test can identify in minutes. Prescribing antibiotics for sore throats without testing is not evidence-based.

Acute bronchitis. Almost always viral. A 2014 Cochrane review (17 trials, 3,936 participants) found antibiotics reduced cough duration by roughly half a day while significantly increasing adverse events.

Most sinus infections. A 2012 IDSA guideline estimated only 2 to 10 percent of acute sinusitis episodes are bacterial. A 2012 trial in JAMA found amoxicillin was no better than placebo for clinically diagnosed sinusitis at day three. Antibiotics are warranted only for severe symptoms (high fever with purulent discharge for three-plus days), worsening after initial improvement, or symptoms persisting without improvement beyond ten days.

When You DO Need Antibiotics

Overuse concerns should not obscure the fact that antibiotics save lives daily when used appropriately.

Confirmed strep throat. A positive rapid strep test warrants a 10-day course of penicillin or amoxicillin — not primarily to speed recovery, but to prevent rheumatic fever, which can permanently damage heart valves. A 2020 review in The Lancet Infectious Diseases confirmed this reduces rheumatic fever risk by approximately 70 percent.

Urinary tract infections. Uncomplicated UTIs (most commonly E. coli) benefit from short antibiotic courses: nitrofurantoin for five days, trimethoprim-sulfamethoxazole for three days, or single-dose fosfomycin. Untreated UTIs can ascend to the kidneys, causing pyelonephritis.

Bacterial pneumonia. Prompt antibiotic therapy is essential — delays are directly associated with increased mortality. Empiric therapy follows ATS/IDSA guidelines based on severity and risk factors.

Skin infections. Cellulitis, wound infections, and abscesses caused by Staphylococcus aureus or Streptococcus pyogenes require antibiotics tailored to local resistance patterns.

Bacterial STIs. Chlamydia, gonorrhea, and syphilis all require antibiotic treatment to resolve infection and prevent transmission.

How to Take Antibiotics Correctly

How you take an antibiotic matters as much as whether you take it.

Follow food instructions precisely. Some antibiotics (ampicillin, certain erythromycin formulations) must be taken on an empty stomach because food reduces absorption. Others (metronidazole, amoxicillin-clavulanate, nitrofurantoin) should be taken with food. A 2019 review in Clinical Pharmacokinetics found food-timing effects varied antibiotic absorption by 40 to 70 percent — this is clinically significant, not optional.

Maintain consistent intervals. "Three times a day" means every eight hours, not with meals. "Twice a day" means every twelve hours. A 2017 study in the Journal of Antimicrobial Chemotherapy showed that irregular intervals create periods of sub-therapeutic concentration — exactly the condition that selects for resistant bacteria. WatchMyHealth's medication tracking can set precise interval reminders (say, 7 AM, 3 PM, and 11 PM) and log each dose, creating a record to share with your doctor if symptoms persist.

Handle missed doses correctly. Take a missed dose as soon as you remember, unless it is nearly time for the next one — then skip the missed dose and continue on schedule. Never double up.

The "Complete the Course" Debate

For decades, patients were told to always finish the full course to prevent resistance. In 2017, infectious disease experts published a landmark analysis in the BMJ challenging this orthodoxy. Their argument: resistance is driven primarily by total antibiotic exposure, not by stopping early. For many infections, randomized trials show shorter courses work equally well.

A 2016 trial in JAMA Internal Medicine found 5 days of treatment for community-acquired pneumonia matched 10 days. A 2015 trial in the New England Journal of Medicine found 4 days for intra-abdominal infections equaled 8 days. UTI treatment has trended toward 3-day and even single-dose regimens.

But this does not apply universally. Tuberculosis requires months of multi-drug therapy — incomplete treatment drives drug-resistant TB. Strep throat needs a full 10-day course to prevent rheumatic fever. Bone infections and endocarditis require prolonged courses to penetrate poorly vascularized tissue.

The practical rule: never stop early on your own initiative, but know that "longer" does not automatically mean "better." The correct length is what your doctor prescribes based on current guidelines for your specific infection.

Side Effects: The Collateral Damage

Antibiotics are not precision weapons. They destroy beneficial gut bacteria alongside pathogens.

Gut microbiome disruption. A 2016 study in mBio found that a single antibiotic course reduced gut microbial diversity by 25 to 50 percent within days. A 2018 study in Nature Microbiology found some changes persisted over a year. The most immediate consequence is antibiotic-associated diarrhea, affecting 5 to 35 percent of patients depending on the drug. Amoxicillin-clavulanate and clindamycin are the worst offenders.

C. diff infection. The most dangerous consequence of microbiome disruption. C. diff thrives when competing bacteria are eliminated, causing severe colitis and approximately 12,800 US deaths per year (CDC data). A 2020 meta-analysis in The Lancet Infectious Diseases identified fluoroquinolones, clindamycin, and broad-spectrum cephalosporins as highest-risk.

Yeast infections. When antibiotics kill protective lactobacilli, Candida overgrows. A 2018 review in the Journal of Antimicrobial Chemotherapy found broad-spectrum antibiotics approximately doubled vaginal yeast infection risk.

Allergic reactions. Penicillin allergies range from mild rashes (1 to 5 percent) to anaphylaxis (roughly 1 in 5,000 courses). Crucially, a 2019 study in the Journal of Allergy and Clinical Immunology: In Practice found over 90 percent of reported penicillin allergies are false — leading to unnecessary use of broader, more harmful alternatives.

Tendon damage. Fluoroquinolones (ciprofloxacin, levofloxacin) carry an FDA-boxed warning for tendon rupture. A 2019 meta-analysis in the British Journal of Clinical Pharmacology found nearly threefold increased risk, higher in adults over 60 and those on corticosteroids.

If you notice unusual symptoms during an antibiotic course — persistent diarrhea, new joint pain, rash — logging them in WatchMyHealth's symptom tracker alongside your medication doses creates a precise timeline your doctor can use to determine whether the antibiotic is the cause.

Drug Interactions to Know

Alcohol. Metronidazole and tinidazole interact with alcohol to produce a disulfiram-like reaction (severe nausea, vomiting, flushing) because they inhibit aldehyde dehydrogenase. This can occur up to 48 hours after the last dose. Most other antibiotics do not produce this specific reaction, but alcohol impairs immune function and worsens GI side effects.

Dairy and minerals. Calcium, magnesium, iron, and zinc bind to tetracyclines and fluoroquinolones in the gut, reducing absorption by up to 85 percent (per a 1992 study in Antimicrobial Agents and Chemotherapy). Take these antibiotics at least two hours before or four to six hours after dairy, supplements, or antacids.

Oral contraceptives. Only rifampin (and other rifamycins) has a well-documented interaction with oral contraceptive pills. For other antibiotics, population studies show no significant increase in contraceptive failure — though severe vomiting or diarrhea from any cause can impair pill absorption.

Blood thinners. Metronidazole, trimethoprim-sulfamethoxazole, and fluoroquinolones potentiate warfarin's effects. A 2014 study in JAMA Internal Medicine found trimethoprim-sulfamethoxazole increased GI bleeding risk by 38 percent in warfarin users. INR should be monitored more frequently during antibiotic courses.

Probiotics: What the Evidence Actually Shows

A 2017 Cochrane meta-analysis of 31 trials (8,672 participants) found probiotics reduced antibiotic-associated diarrhea by about 40 percent. The best-studied strains are Lactobacillus rhamnosus GG and Saccharomyces boulardii (a yeast unaffected by antibiotics). However, a large 2013 trial in The Lancet (PLACIDE, nearly 3,000 hospitalized patients) found no significant C. diff reduction from probiotics.

Practical advice: taking a well-studied probiotic during and for one to two weeks after an antibiotic course is reasonable and low-risk. Take it at least two hours apart from the antibiotic dose. But do not expect probiotics to fully prevent microbiome disruption.

Five Common Myths Corrected

"Green mucus means I need antibiotics." Mucus color reflects neutrophil activity, not pathogen type. A 2011 study in the European Respiratory Journal found no reliable correlation between color and bacterial infection.

"Antibiotics help me get over a cold faster." The 2013 Cochrane review confirms they do not. The placebo effect is real; the pharmacological effect on viruses is zero.

"Resistance means my body becomes immune to antibiotics." Resistance develops in bacteria, not in people. One person's unnecessary course can create resistant bacteria that spread to others. A 2018 study in The BMJ found community prescribing rates directly correlated with community resistance rates.

"Newer or more expensive antibiotics work better." For most common infections, older narrow-spectrum drugs are equally effective and often preferable — they cause less microbiome damage. Amoxicillin (approved 1972) remains first-line for strep throat, dental infections, and ear infections.

"I should save leftover antibiotics for next time." Leftover antibiotics mean the course was not completed correctly, or the wrong drug was prescribed. Using partial courses exposes bacteria to sub-therapeutic levels — the exact condition that breeds resistance. Dispose of leftovers properly.

What to Ask Before Accepting a Prescription

The CDC's "Be Antibiotics Aware" campaign recommends five questions:

  1. "Do I definitely have a bacterial infection?" If the answer is "probably not, but just in case," discuss watchful waiting.
  2. "Is there a test to confirm?" Rapid strep tests take minutes. Urine cultures confirm UTIs.
  3. "What side effects should I watch for?" Know what is expected and what should prompt a callback.
  4. "Are there interactions with my other medications?" Especially relevant for warfarin, oral contraceptives, and mineral supplements.
  5. "What is the shortest effective course?" Some doctors prescribe longer courses out of habit rather than current evidence.

If prescribed an antibiotic, log it in WatchMyHealth's medication tracker with the correct frequency and duration. If symptoms are not improving within 48 to 72 hours, worsen after initial improvement, or you develop severe diarrhea, rash, hives, or breathing difficulty — contact your doctor immediately. Logging symptoms alongside doses creates a precise timeline that replaces guesswork at your follow-up appointment.

The Bottom Line

Antibiotics save millions of lives when used correctly — and cause measurable harm when used for infections they cannot treat.

  • Most colds, flu, sore throats, coughs, and sinus infections are viral. Antibiotics will not help. Ask for a test before accepting a prescription.
  • When truly needed, take them exactly as prescribed. Right dose, right intervals, right duration — use a medication tracker to stay consistent.
  • Expect and manage side effects. Gut disruption is the norm. Probiotics may help. Severe diarrhea, rash, or new symptoms warrant an immediate call to your doctor.
  • Watch for interactions. Metronidazole with alcohol, fluoroquinolones with dairy, several antibiotics with warfarin.
  • Never take antibiotics "just in case." If the infection is not bacterial, the antibiotic is not helping you — it is harming you and contributing to a global resistance crisis that threatens everyone.

For more on why resistance is accelerating and what it means for the future of medicine, read our companion article on antibiotic resistance and superbugs.