e-learning | Diagnosis and management of pneumonia

NICE first published guidance on pneumonia in 2014, and updated 2019. At that time there was a push to use the CRB65 score in general practice and also CRP as an aid to deciding on when to use antibiotics. There was also excellent information on how long patients should expect symptoms to last for. I was also very surprised then - and still am - by the statistics behind pneumonia, so linked those in.

NICE has just published new guidance on the use of antibiotics in community acquired pneumonia. I am therefore updating this summary with that information. The most useful additional information from this is how long to use antibiotics for (i.e. most patients will need a five day course, but some will need longer).

NICE has also published new guidance on the use of antibiotics in hospital acquired pneumonia (i.e. that which starts 48 hrs or more after hospital admission). I have not included that in this summary as it is less useful for us. They do suggest considering following the community acquired pneumonia guidelines if the pneumonia is diagnosed between three and five days after admission.

The scale of the problem.

  • 0.5 to 1% of adults are affected each year
  • 5 to 12% of patients seen in GP with LRTI have pneumonia
  • 22 to 42% of patients diagnosed are admitted to hospital
  • 5 to 14% in hospital mortality
  • 1.2 to 10% of those admitted end up on ITU
  • More than half of deaths occur in those aged 84 and older

C-reactive protein CRP

Where pneumonia is diagnosed clinically, treat it. If you are in doubt about whether antibiotics are needed, they advise a point of care CRP test. I haven't seen these used anywhere and would be interested to know if others have. Presumably a normal CRP could be done, but the time scales would make it less useful.

  • 19 or less - do not routinely use antibiotics
  • 20 to 100 - consider delayed antibiotics
  • 100 or more - give antibiotics

CRB65 score

This gives a risk of mortality associated with the pneumonia. Score 1 for each of:

  • Confusion (MMT 8 or less, or new confusion in time, person or place)
  • RR 30 or more
  • Low BP (90/60)
  • 65 or more

0 - low risk (1% mortality) - consider home care
1 or 2 - intermediate risk (1 to 10% mortality)
3 or 4 - high risk ( > 10% mortality)

Consider admission if there is a score of 2 or more

When should we send samples for microbiology?

Do not routinely send-off samples in low severity pneumonia. For more severe pneumonia testing may be needed, but this would be in secondary care (e.g. blood cultures and specialist tests).

If the patient has not improved following antibiotic treatment, then send a sputum sample for testing if it hasn’t already been done.

Be aware of the possibility of non-bacterial causes, e.g. flu if patients aren’t improving.

When choosing an antibiotic, what should be considered?

  • Severity in adults as per CRB65.
  • Severity in children and young people as per clinical judgement.
  • Risk of complications, e.g. relevant comorbidity (lung disease, immunosuppression).
  • Local resistance patterns and surveillance data (e.g. flu or Mycoplasma Pneumoniae – this latter comes in outbreaks every 4 yrs or so and is more common in school age children).
  • Recent microbiological results.

What antibiotics are advised?


  • Low risk - amoxicillin for five days (longer if they're not improving by day three). Doxycycline, clarithromycin or erythromyin can be used as alternatives or if atypical pathogens are suspected.
  • Intermediate risk - consider amoxicillin + (if atypical pathogens are suspected) a macrolide for five days. Doxycycline or clarithromycin can be used alone if they are penicillin allergic.
  • High risk - consider co-amoxiclav + macrolide (or levofloxacin).

Children > 1m

NB < 1m – refer to a specialist.

  • Non–severe symptoms and signs: amoxicillin for five days. Clarithromycin or erythromycin or doxycycline (if > aged 12) can be used as alternatives or if atypical pathogens are suspected.
  • Severe. Co-amoxiclav + (if atypical pathogen is suspected) a macrolide.

How long should we continue antibiotics for?

Five days, unless:

  • Microbiological results suggest a longer course is needed OR
  • Patient isn’t clinically stable, eg has had fever in the past 48 hrs or more than 1 sign of instability:
    • sBP < 100
    • HR > 100
    • RR > 24

Evidence suggests that shorter courses are as effective as longer courses for most patients.

What advice should we give patients?

  • Possible adverse effects of antibiotics.
  • How long symptoms are likely to last (as below).
  • Seek medical help if:
    • Symptoms worsen significantly or rapidly
    • Symptoms do not start to improve in three days
    • The person becomes systemically very unwell.

What should you advise patients about recovery times?

  • 1w - fever should have resolved
  • 4w - chest pain and sputum should have substantially reduced
  • 6w - cough and breathlessness should have substantially reduced
  • 3m - most symptoms should have resolved, though may have fatigue
  • 6m - most will be back to normal

No Comments Yet.

Leave your comments