Professor Mark Woodhead (pictured) and Dr Celia Pincus discuss NICE CG191 on pneumonia and the importance of accurate diagnosis, severity assessment, and appropriate antibiotic treatment

woodhead mark

Read this article to learn more about:

  • the key recommendations relevant to the management of pneumonia in primary care
  • the use of CRP testing to guide antibiotic prescribing decisions
  • a 'time scale of improvement' to share with patients to manage their expectations.

Key points

GP commissioning messages

Lower respiratory tract infections (LRTIs) are one of the commonest reasons for GP consultations and will be an increasing burden on general practice as the population ages.1 The term LRTI encompasses overlapping clinical syndromes of varying severity. Pneumonia accounts for between 5% and 12% of GP consultations for LRTI.2 While the majority of such cases can be managed at home by GPs, pneumonia is often at the more severe end of the severity spectrum in LRTI with between 22% and 42% of diagnosed cases requiring hospital admission.2 Of those admitted to hospital, as many as 20% may die,3 emphasising the need for early accurate diagnosis and appropriate treatment, including choice of place of care.


The majority of LRTIs are believed to be either bacterial or viral, although non-infectious illnesses (e.g. asthma) may be misdiagnosed as LRTI due to symptom overlap. While viruses can cause pneumonia, most pneumonia is bacterial in origin and it is accepted that antibiotic therapy is appropriate once the diagnosis of pneumonia is made. However, here lies the problem—while some patients present with a clinically obvious pneumonia (i.e. respiratory symptoms with new focal signs on chest examination), the clinical features of pneumonia are often non-specific, especially in the elderly, and overlap with other, usually non-bacterial LRTI syndromes.

A chest radiograph, the gold standard for pneumonia diagnosis, is not usually quickly available to GPs, which leaves them with a difficult decision about when to prescribe an antibiotic for LRTIs. The rising tide of bacterial antimicrobial resistance provides a powerful argument for more limited, but appropriate, antibiotic prescription,4 but at the same time antibiotic treatment failure (defined as switching antibiotics within 30 days) for LRTIs appears to be increasing.5 How should GPs decide when and what to prescribe, and who to refer to hospital?

NICE guidelines

NICE has recently published its Clinical Guideline (CG) 191 on Pneumonia: diagnosis and management of community and hospital-acquired pneumonia in adults, which has been awarded the NICE Accreditation Mark (see Box 1, below). Its recommendations for hospital management are not described in this article. The key recommendations relevant to the management of pneumonia in primary care include:2

  • using C-reactive protein (CRP) (as a point-of-care [POC] test) to guide the antibiotic prescribing decision in people without a clinical diagnosis of pneumonia
  • using a severity assessment tool to help identify patients with a clinical diagnosis of community-acquired pneumonia, whose outcome will be improved by hospital referral
  • antibiotic choice and duration
  • the expected natural history of the condition to inform patients.
Box 1: NICE Accreditation Mark
NICE Accreditation Mark

NICE Clinical Guideline 191 on Pneumonia: diagnosis and management of community- and hospital-acquired pneumonia in adults has been awarded the NICE Accreditation Mark.
This Mark identifies the most robustly produced guidance available.
See for further details.


Point-of-care testing

The guideline development group (GDG) for CG191 evaluated the use of both the chest radiograph and biomarkers to assist in the antibiotic prescribing decision. No suitable studies were identified to make a recommendation about chest radiographs.

Studies performed on two blood biomarkers used as POC tests in general practice, CRP and procalcitonin, were assessed. A significant reduction in antibiotic prescriptions, without evidence of significant harm from lack of antibiotics, was found using either biomarker. No advantage of procalcitonin over CRP was found. A non-significant increase in hospital admission and re-consultation rates was found in the CRP groups, but it is not known whether these changes were appropriate in relation to patient care or not.6


NICE therefore recommends that a CRP test should be considered at point of care if a diagnosis of pneumonia cannot be made following clinical assessment and if there is doubt as to whether antibiotics should be prescribed. It recommends that:6

  • antibiotic therapy should not routinely be offered if the CRP concentration is less than 20 mg/l
  • a delayed antibiotic prescription (i.e. a prescription for use if symptoms worsen) if CRP concentration is between 20 mg/l and 100 mg/l should be considered
  • antibiotic therapy should be offered if CRP concentration exceeds 100 mg/l.

These recommendations build on those set out in NICE CG69 for antibiotic management of self-limiting RTIs in adults and children.7 The GDG for CG69 acknowledged that GPs would identify a group of patients with self-limiting RTIs who would not benefit from antibiotics. They also recognised that GPs would identify some patients as having a clinical diagnosis of pneumonia, and in these cases antibiotics should always be prescribed. However, this would leave a group of patients with RTI where the antibiotic prescribing decision would be uncertain and it is only in these patients that considering CRP testing is recommended.7

While POC CRP testing is part of routine GP practice in some Northern European countries, the GDG for CG191 recognised that this recommendation poses the biggest challenge for implementation and practice change in the UK.6

Severity assessment in primary care

The NICE pneumonia guideline relied on evidence from studies conducted in patients with pneumonia in hospital, which showed robustness of the CRB65 mortality risk assessment tool (see Box 2, below).6 One study performed in general practice showed similar findings in patients with pneumonia. There was insufficient evidence to make a recommendation about severity scoring tools for LRTI.6 It is important to note that CRB65 is a mortality risk score that should be used to aid clinical judgement rather than to replace it. It is not a perfect tool and occasional patients with low scores may be severely ill. The GDG recognised that all patients aged 65 years or over would automatically score 1.6 Older patients are at statistically higher risk, but this does not mean that all of those aged over 65 years should be referred to hospital. The GDG felt there was insufficient evidence to make an age-specific recommendation.6

Box 2: CRB65 score for mortality risk assessment in primary care2

  • CRB65 score is calculated by giving 1 point for each of the following prognostic features:
    • confusion (abbreviated Mental Test score 8 or less, or new disorientation in person, place, or time)
    • raised respiratory rate (30 breaths per minute or more)
    • low blood pressure (diastolic 60 mmHg or less, or systolic less than 90 mmHg)
    • age 65 years or more
  • Patients are stratified for risk of death as follows:
    • 0: low risk (less than 1% mortality risk)
    • 1 or 2: intermediate risk (1%–10% mortality risk)
    • 3 or 4: high risk (more than 10% mortality risk).

NICE (2015) CG191. Pneumonia: Diagnosis and management of community- and hospital-acquired pneumonia in adults. Available at: Reproduced with permission.


Recommendations on severity assessment in primary care are based on the CRB65 score (see Box 2, above) and state that:2

  • it should be used to assess low, intermediate, or high risk of death when a diagnosis of community-acquired pneumonia has been made in primary care
  • it should be combined with clinical judgement to decide whether patients need hospital assessment as follows:
    • home-based care should be considered if there is a CRB65 score of 0
    • hospital assessment should be considered for other patients, especially if they have a CRB65 score ≥2.

Choice and duration of antibiotic

Clinical Guideline 191 found few studies that compared different durations of antibiotic treatment. Those that did, found similar outcomes with 5 days compared with longer treatment, regardless of antibiotic type.6 Symptom improvement and the need for extension of the antibiotic course or alternative treatment should be based on expected symptom resolution as per the recommendation for patient information below.

Despite the availability of many studies comparing single antibiotic use, only three were conducted in the UK and none of these included amoxicillin as used in the UK as a comparator. No advantage, indicated by clinical cure rates, was found for any antibiotic. The decision to recommend amoxicillin was based on familiarity, a desire to favour narrow spectrum antibiotics, and a suggestion of increased harms with some other antibiotic classes. No advantage of dual antibiotic therapy in non-severe pneumonia was found.6

Studies on moderate- or high-severity pneumonia were conducted in hospital. Such patients would usually be referred to hospital in the UK and it might be unusual for a GP to need to initiate treatment in such cases, but the recommendations are included for completeness.6


Inlow-severity community-acquired pneumonia, NICE CG191 recommends:2

  • offering a single antibiotic for 5 days
  • considering amoxicillin rather than a macrolide or a tetracycline— in cases of penicillin allergy a macrolide or tetracycline may be considered
  • that if symptoms still persist after 3 days, consideration should be given to extending the course of antibiotic for longer than 5 days
  • explaining to patients (or families/carers) that they should seek further medical advice if there is no improvement in symptoms within 3 days, or if symptoms get worse
  • that a fluoroquinolone or dual antibiotic therapy should not routinely be offered.

If a patient is diagnosed with moderate-or high-severity community-acquired pneumonia, NICE recommends considering a 7- to 10-day course of dual antibiotic therapy with:2

  • amoxicillin and a macrolide for patients with moderate-severity community-acquired pneumonia
  • a beta-lactamase stable beta-lactam and a macrolide for patients with high-severity community-acquired pneumonia.

Patient information

Patient expectations of the speed of symptom resolution will be one determinant of re-consultation with their GP. Patients often expect symptom resolution to be quicker than it is in reality. Studies of the natural history of symptom resolution after pneumonia were used to develop this recommendation for GPs to share with their patients.


As well as explaining to patients with community-acquired pneumonia that symptoms should steadily improve after starting their course of treatment, with rates of improvement varying according to severity, the NICE guideline on pneumonia suggests a timescale for improvement (see Box 3, below). The guideline recommends to clinicians that they should advise their patients with community-acquired pneumonia to request a further appointment with their doctor if they feel their condition is getting worse or if it is not improving as expected.2

Box 3: Expected schedule for improvement of pneumonia symptoms following commencement of treatment2

  • 1 week: fever should have resolved
  • 4 weeks: chest pain and sputum production should have substantially reduced
  • 6 weeks: cough and breathlessness should have substantially reduced
  • 3 months: most symptoms should have resolved, but fatigue may still be present
  • 6 months: most people will feel back to normal.

NICE implementation tools

NICE has developed the following tools to support implementation of Clinical Guideline 191 (CG191) on Pneumonia: diagnosis and management of community- and hospital-acquired pneumonia in adults.

NICE support for service improvement systems and audit

Baseline assessment toolNICE support for service improvement systems and audit icon

The baseline assessment is an Excel spreadsheet that can be used by organisations to identify if they are in line with practice recommended in NICE guidance, and to help them plan activity that will help them meet the recommendations.

NICE support for commissioners

Costing statement NICE support for commissioners icon

A costing statement has been produced because of wide variation in practice, therefore a national resource impact would be challenging to estimate. The statement has been prepared in consultation with experts working in this area and has been approved for publication by NICE.

Key to NICE implementation icons

Commissioning.epsNICE support for commissioners

  • Support package for commissioners and others for quality standards
  • NICE guide for commissioners
  • NICE cost impact support for guidance (selection from national report/local template/costing statement, dependent on topic)

Audit.epsNICE support for service improvement systems and audit

  • Forward planner
  • 'How to' guides (generic advice on processes)
  • Local government briefings (with Centre for Public Health Excellence)
  • Baseline assessment tool for guidance
  • Audit support including electronic data collection tools
  • E-learning modules (commissioned)

Education.epsNICE support for education and learning

  • Clinical case scenarios
  • Learning packages including slide sets
  • Podcasts
  • Shared learning and other local best practice examples


The key priorities identified by the GDG for implementation of NICE CG191 in primary care are: use of POC CRP testing to aid diagnosis; offering a 5-day course of antibiotics; and provision of information to patients. Many clinical questions in relation to pneumonia remain that cannot be robustly answered from the current research evidence base. NICE CG191 provides the most methodologically robust and up-to-date assessment of the available evidence. Hopefully this will improve pneumonia and LRTI care and contribute to antibiotic stewardship.

Key points

  • Lower respiratory tract infections, including pneumonia, are an increasingly common reason for GP consultation
  • Prescribe antibiotic treatment when the diagnosis of pneumonia is made
  • CRP testing may assist prescribing where a diagnosis has not been made and the antibiotic prescribing decision is initially uncertain
  • Use CRB65 with clinical judgement to assess illness severity and care setting
  • Offer 5 days of antibiotic therapy to those with low-severity pneumonia
  • Amoxicillin is the first choice antibiotic for low-severity pneumonia
  • Inform patients of the likely natural history of their symptoms.

Back to top

GP commissioning messages

 written by Dr David Jenner, NHS Alliance GMS contract/PBC Lead

  • NICE CG191 on POC CRP testing will prove controversial with regard to how the testing kits are funded:
    • the use of the POC CRP test does not seem to reduce hospital admissions but does reduce prescribing of antibiotics (which are of low acquisition cost). The POC CRP test is therefore likely to act as another cost pressure in a financially burdened NHS
    • GPs may not choose to implement this guidance unless the kits are funded or are reimbursable through personally administered prescription schemes
  • Commissioners:
    • may wish to balance the cost of lab-based CRP tests against POC tests as the former can usually give a result within 24 hours. The POC tests are only recommended where the clinical diagnosis of pneumonia is in doubt, so a short delay is unlikely to be critical
    • will need to consider updating their formularies to identify amoxicillin as the first-line antibiotic for CAP and liaise with local microbiologists where other drugs are currently recommended as first-line (e.g. doxycycline) to ensure that local antibiotic resistance trends are recognised and addressed
  • Use of the CRB65 tool and oxygen saturation monitoring should be encouraged in primary care settings (e.g. GP practices, out-of-hours, and urgent care centres) as an adjunct to clinical judgement.

POC=point of care; CRP=C-reactive protein; CAP=community-acquired pneumonia

Back to top


  1. Millet E, Quint J, Smeeth L et al. Incidence of community-acquired lower respiratory tract infections and pneumonia among older adults in the United Kingdom: a population based study. PLoS ONE 2013; 8 (9): e75131. doi:10.1371/journal.pone.0075131
  2. NICE. Pneumonia: diagnosis and management of community- and hospital-acquired pneumonia in adults. Clinical Guideline 191. NICE, 2014. Available at:
  3. Lim W, Rodrigo C. British Thoracic Society: adult community acquired pneumonia audit 2012/13. audit-reports/bts-adult-community-acquiredpneumonia- audit-report-201213/.
  4. Department of Health. Annual Report of the Chief Medical Officer. Volume 2, 2011. Infections and the rise of antimicrobial resistance. London: Crown Copyright, 2013. Available at: chief-medical-officer-annual-report-volume-2
  5. Currie C, Berni E, Jenkins-Jones S et al. Antibiotic treatment failure in four common infections in UK primary care 1991–2012: longitudinal analysis. Br Med J 2014; 349: g5493
  6. National Clinical Guideline Centre. Pneumonia: diagnosis and management of community- and hospital-acquired pneumonia in adults. Clinical Guideline 191, full guideline. National Clinical Guideline Centre, 2014. Available at:
  7. NICE. Respiratory infections—antibiotic prescribing: prescribing of antibiotics for selflimiting respiratory tract infections in adults and children in primary care. Clinical Guideline 69. NICE, 2008. Available at: