THE SHIT LIST: The Green Phlegm That Demanded an Antibiotic

The Claim

For immunocompetent adults with an acute cough of less than about three weeks, normal oxygenation, no focal lung findings, and no abnormal vital signs suggesting pneumonia, routine antibiotics do not improve the overall clinical course of uncomplicated acute bronchitis. Green or yellow sputum is not a secret bacterial telegram. It can reflect neutrophils and inflammation in a viral illness.

The claim is deliberately narrow. It does not apply to suspected pneumonia, a COPD exacerbation, confirmed or strongly suspected pertussis, significant immunocompromise, or a patient who is systemically very unwell. It is also not an instruction to diagnose bronchitis from a browser. The useful clinical work is ruling out the conditions that are not bronchitis.

The Number That Should End The Argument

The 2017 Cochrane review included 17 trials and 5,099 participants. Antibiotics did not produce a statistically clear improvement in the proportion of patients who were clinically better at follow-up: risk ratio 1.07, 95% confidence interval 0.99 to 1.15, across 3,841 participants. Mean cough duration fell by 0.46 days. Adverse effects rose by 20%, with one additional harmful outcome for roughly every 24 people treated.

Half a day is not nothing. It is also not much of a return for a treatment that can cause rash, nausea, diarrhoea, allergy, Clostridioides difficile infection, drug interactions, and antimicrobial resistance.

A double-blind randomised trial made the point more theatrically. Among 220 adults with acute bronchitis and no underlying lung disease, five days of azithromycin was no better than low-dose vitamin C. When the antibiotic cannot beat the control chosen largely to keep the blinding tidy, the prescription pad might consider a quiet lie-down.

Why This Habit Persists

A productive cough looks bacterial, sounds bacterial, and can turn a tissue the colour of a 1970s kitchen. The visual drama is persuasive. The CDC is blunt: coloured sputum does not indicate bacterial infection.

Time pressure matters too. A prescription can feel like a conclusion. Explaining that cough often lasts two or three weeks, checking red flags, and arranging sensible return precautions takes longer than clicking amoxicillin. Clinicians also remember the one pneumonia they nearly missed more vividly than the hundreds of bronchitis cases that resolved without ceremony.

Patients are frequently blamed for demanding antibiotics, but clinicians often overestimate that demand. Many patients want an examination, an explanation, relief where possible, and a clear point at which to worry. A confident no-antibiotic plan is treatment. A shrug is not.

The Sophisticated Objection

The strongest objection is not that antibiotics secretly cure viral bronchitis. It is that real primary-care patients are untidy. Frail older adults, people with several chronic illnesses, and those who present atypically were under-represented in many trials. The Cochrane authors explicitly allowed that some frail, elderly people with multimorbidity might derive a modest benefit that existing evidence has not measured well.

Fair enough. Uncertainty should trigger better assessment, not a universal macrolide tax. Abnormal vital signs, focal consolidation, hypoxaemia, pleuritic pain, haemoptysis, significant breathlessness, clinical deterioration, or a high-risk host should change the diagnostic pathway. So should a cough compatible with pertussis, because antibiotics there are used mainly to reduce transmission rather than to perform miracles on an established cough.

The other objection is stewardship fatigue: one prescription will not collapse modern medicine. True. Nor will one dropped cigarette reverse lung cancer risk. Population problems are built from individual acts that each look conveniently small.

What The Evidence Does And Does Not Say

NICE concludes that antibiotics do not improve the overall clinical condition in acute bronchitis and shorten cough by about half a day on average. The CDC recommends no routine antibiotic treatment regardless of cough duration after pneumonia has been excluded.

Trials of delayed prescribing add nuance. In a pragmatic study of 405 adults with uncomplicated respiratory infections, no-prescription and delayed-prescription strategies reduced antibiotic use and weakened beliefs that antibiotics were effective. A larger British trial of 889 patients found that no prescription or delayed strategies produced similar symptom outcomes to immediate antibiotics, with fewer than 40% using antibiotics.

That does not mean every patient should leave with a just-in-case prescription. NICE notes that routine back-up prescriptions may imply that antibiotics will eventually be needed. For a clearly uncomplicated case, a clear no-antibiotic plan is the cleaner message. Delayed prescribing is a compromise for selected uncertainty, not a participation trophy for the prescription pad.

The evidence is strongest for short-term symptoms and adverse effects. Trials were heterogeneous, and many did not capture rare harms or the downstream ecological cost of resistance. They also do not prove that every symptomatic remedy works. Acute bronchitis is an excellent place to discover that doing less can be evidence-based without pretending that doing less makes the cough pleasant.

Clinical Exceptions And Safety Red Flags

Seek prompt medical assessment for significant or worsening shortness of breath, oxygen saturation below the person's normal range, chest pain, coughing blood, confusion, fainting, persistent high fever, severe weakness, dehydration, blue lips, or rapid deterioration.

Pneumonia becomes more concerning with a heart rate at or above 100 beats per minute, respiratory rate at or above 24 breaths per minute, oral temperature at or above 38 C, low oxygen saturation, or focal consolidation, egophony, or fremitus on examination. These are assessment clues, not a home diagnostic score.

This article does not cover COPD exacerbations, bronchiectasis, cystic fibrosis, active cancer treatment, substantial immune suppression, recent hospitalisation, pregnancy-specific decisions, or suspected pertussis, influenza, or COVID-19 when testing or specific treatment may matter. Children need their own assessment and guidance.

What We Do About It

• Define the population before applying the rule: otherwise healthy adult, acute cough, no pneumonia pattern, no chronic-lung-disease exacerbation, and no systemic instability.
• Explain the expected course. A cough lasting two to three weeks can be ordinary even when it is deeply irritating.
• Say plainly that coloured sputum does not prove bacterial infection.
• Use symptom relief selectively and honestly. Hydration, rest, honey for appropriate adults, and individualised analgesia may help comfort, but no syrup should be promoted as an eviction notice for bronchitis.
• Test for influenza, COVID-19, or pertussis when the clinical context makes the result actionable.
• Give specific return precautions and a time for reassessment if symptoms worsen or fail to follow the expected course.
• Reserve antibiotics for a defined bacterial diagnosis or a genuinely high-risk clinical decision, and document the reason.

What Would Change My Mind

A large, modern randomised trial in well-defined adults with uncomplicated acute bronchitis showing a meaningful reduction in hospitalisation, pneumonia, prolonged functional impairment, or severe symptoms, without offsetting adverse effects or resistance, would change the conclusion. A greener tissue would not.

The humane version of stewardship is not “go away, it is viral.” It is “I have looked for the dangerous alternatives, here is what this usually does, here is what may help, and here is exactly when I want to see you again.”

Sources and receipts

1. Smith and colleagues, Cochrane Review: Antibiotics for acute bronchitis, 17 trials and 5,099 participants — Source cited in the reviewed article
2. Evans and colleagues, randomised double-blind trial of azithromycin versus vitamin C in 220 adults — Source cited in the reviewed article
3. CDC, Outpatient Clinical Care for Adults: acute uncomplicated bronchitis — Source cited in the reviewed article
4. NICE NG120, Cough (acute): antimicrobial prescribing recommendations — Source cited in the reviewed article
5. de la Poza Abad and colleagues, randomised clinical trial of prescription strategies in 405 adults — Source cited in the reviewed article
6. Little and colleagues, pragmatic randomised trial of delayed prescribing strategies in 889 patients — Source cited in the reviewed article
7. Stuart and colleagues, individual-patient-data meta-analysis of delayed antibiotic prescribing — Source cited in the reviewed article
8. Kinkade and Long, Acute Bronchitis: Rapid Evidence Review — Source cited in the reviewed article