This recommendation is consistent with guidelines produced by the British Thoracic Society [British Thoracic Society, 2004].

• Immediate empirical treatment with antibiotics is considered essential to reduce mortality, length of illness, severity of symptoms, and the likelihood of complications.
• Although the evidence from randomized controlled trials to support antibiotic use is limited, this is likely to be because the benefit of antibiotics is beyond doubt in this group. This would make placebo controlled trials unethical.
• Antibiotics may not be appropriate during the terminal phase of life, see the CKS topic on Palliative cancer care - cough.

These recommendations are based on those of the British Thoracic Society (BTS) [British Thoracic Society, 2004; British Thoracic Society, 2009] and the Scottish Intercollegiate Guidelines Network (SIGN) [SIGN, 2002].

• Empirical treatment is necessary, as sputum samples take too long to process to identify a causative pathogen of community-acquired pneumonia and practically influence initial management.
• CKS found no evidence from randomized controlled trials to support the use of one antibiotic over another in the treatment of pneumonia. The choice of antibiotic used is therefore dependent on the known susceptibility of the causative pathogens of bacterial pneumonia, taking into account resistance patterns.
• Amoxicillin is the preferred first-line choice in most cases [British Thoracic Society, 2009], as it has a broad spectrum of activity and most of the common pathogens involved are still susceptible to it [Bush, 2003].
• The majority of cases of community-acquired pneumonia are caused by Streptococcus pneumoniae or Haemophilus influenzae, which are still susceptible to amoxicillin if a large enough dose is given (see Dosing regimen of amoxicillin).
• Other atypical pathogens may not be sensitive to amoxicillin. However, in reality, these tend to be rare causes of community-acquired pneumonia, and will often elicit symptoms severe enough for referral or admission.
• Clarithromycin has a similar spectrum of activity to amoxicillin and is an appropriate alternative if the person has an allergy to penicillin.
• Clarithromycin is recommended as the macrolide of choice (instead of erythromycin) by the BTS guidelines, on the basis of improved gastrointestinal tolerance and an easier dosing schedule [British Thoracic Society, 2009].
• In exceptional cases where a person has severe pneumonia but is not admitted, the addition of clarithromycin to amoxicillin should be considered to maximize effectiveness and provide a broad range of cover. This recommendation is based on clinical experience rather than published studies [SIGN, 2002].
• Doxycycline is recommended as an alternative to amoxicillin and amoxicillin plus clarithromycin by the BTS guidelines because pneumococci have less resistance to it, and it has greater activity against atypical pathogens, including M. pneumoniae [British Thoracic Society, 2009]. SIGN recommend that a tetracycline should be used first-line when M. pneumoniae infection is suspected because tetracyclines have more activity against this pathogen than amoxicillin [SIGN, 2002].
• Fluoroquinolones, such as levofloxacin and, more recently, moxifloxacin, have greater activity than amoxicillin against some of the pathogens involved in pneumonia, including S. pneumoniae. They are not generally recommended for use in primary care because of rising levels of resistance to fluoroquinolones observed in countries with widespread prescribing of these antibiotics in the community. However, they may be recommended by a specialist if the person has contraindications to penicillins and macrolides.

The recommendations are consistent with those made by the British Thoracic Society in Guidelines for the management of community-acquired pneumonia in adults [British Thoracic Society, 2004].

• Analgesics: there is limited evidence from controlled trials that both paracetamol and ibuprofen reduce some symptoms of cold, such as pain and temperature. However, they have no effect on symptoms such as cough. Ibuprofen has more contraindications than paracetamol (which may be especially relevant to older people with severe infection), but has the advantage of probably being more effective at reducing temperature and requiring less frequent dosing.
• Aspirin and other nonsteroidal anti-inflammatory drugs are not recommended, as they are more likely to cause serious adverse effects.
• In particular, aspirin should be avoided in older people who are more prone to its adverse effects, or who are taking concomitant nonsteroidal anti-inflammatory drugs, or aspirin for cardiovascular purposes.
• Preventing dehydration:
• In clinical practice dehydration is a commonly observed problem for people who are unwell with a chest infection.
• It occurs because they have increased fluid losses from sweating, and/or a reduced fluid intake due to general malaise.
• It may not be recognized by someone who is unwell because many of the symptoms of dehydration such as headache, dry mouth, and general malaise, may be wrongly attributed to their infective illness rather than dehydration.
• For someone who is unwell dehydration may be most easily recognized by observing urine colour and output, and may be relieved by increasing fluid intake appropriately.
• Cough medicines to suppress a productive cough are not recommended because they suppress the natural mechanism that keeps the airway clear. In principle, suppression of an unproductive cough is unlikely to cause harm [SIGN, 2002]. However:
• Cough medicines that are available over the counter are largely ineffective, and may have associated adverse effects [Morice et al, 2006].
• They may have a useful placebo effect. Because they have little effect at suppressing cough, there is probably no harm in someone continuing to use them even for a productive cough if they have already bought them.
• Opioids in doses higher than recommended in over-the-counter preparations may suppress cough, but also have significant adverse effects and are therefore not recommended.
• Smoking cessation is widely advocated because:
• Smoking cessation reduces irritation to the bronchial tree already inflamed due to infection, and theoretically may reduce coughing.
• Smoking is a risk factor for chest infections [Gutzwiller et al, 1989]. Smoking cessation reduces the risk of further episodes of chest infections, in addition to conferring many other health benefits.

Basis for arranging admission for people that deteriorate on treatment

• People that deteriorate on treatment are at an increased risk of death from pneumonia.
• They require alteration to their treatment and close observation to ensure that these alterations are effective.
• For most people hospital admission is the appropriate way to deliver this care.

Basis for management options for people that are not improving on treatment

• In people who have not responded to an initial course of amoxicillin, switching to, or adding on a macrolide, is recommended by the British Thoracic Society [British Thoracic Society, 2001].
• Empirical treatment with a macrolide and/or amoxicillin is based on clinical experience rather than published studies [SIGN, 2002].
• The rationale is to provide antibiotic coverage for as many pathogens involved in the infection as possible. Up to 10% of community-acquired pneumonia may be caused by two organisms.
• If the person was not initially taking amoxicillin because it was contraindicated, CKS recommends getting expert advice from a microbiologist. Other antibiotics such as fluoroquinolones may be recommended.

The British Thoracic Society recommend a follow-up X-ray at 6 weeks based upon evidence that:

• People with pneumonia who smoke are at high risk of having lung cancer.
• People with pneumonia and persistent or slowly resolving symptoms and signs of pneumonia are at high risk of having an underlying malignancy.
• Chest X-rays taken at presentation will not detect all people with an underlying malignancy.
• A chest X-ray will detect an underlying malignancy in most people after 6 weeks of treatment for pneumonia.

[British Thoracic Society, 2001]

Basis for recommending smoking cessation

• Smoking cessation is widely advocated because:
• Cross-sectional studies show that smoking is a risk factor for chest infections [Gutzwiller et al, 1989]. Smoking cessation reduces the risk of further episodes of chest infections, in addition to conferring many other health benefits.
• Advice and treatment to stop smoking have been shown to have a small but significant effect in reducing smoking.

Basis for recommending pneumococcal vaccination

• Streptococcus pneumoniae is one of the pathogens known to cause acute bronchitis [Macfarlane et al, 2001].
• The effectiveness of pneumococcal vaccination in preventing morbidity and mortality from S. pneumoniae has been demonstrated in a meta-analysis of 14 randomized controlled trials (n = 48,837) [Cornu et al, 2001].

Basis for recommending influenza vaccine

• The influenza virus is one of the pathogens known to cause acute bronchitis [Macfarlane et al, 2001].
• A number of randomized controlled trials and cohort studies have demonstrated that influenza immunization is effective for:
• Reducing the incidence of influenza.
• Reducing morbidity and mortality from secondary bacterial infections following influenza, particularly in at-risk groups.