Pneumonia: diagnosis and management

Why the committee made the recommendations

The evidence compared home-based care with inpatient care for people with low- to moderate-severity community-acquired pneumonia. The committee agreed that home-based care covers a variety of different care models, such as hospital at home service, virtual wards and same-day emergency care (SDEC) units. They noted that the evidence was limited to hospital at home models but agreed that the findings could be extrapolated to other types of home-based care.

The studies were small, had methodological limitations and included participants not considered to be representative of those who would potentially use virtual wards or hospital at home services. It showed that patients treated at home tended to do no worse than patients treated as inpatients in hospital in terms of hospital re-admission, antibiotic duration, symptom improvement and adverse events. There was insufficient evidence to determine the impact of home-based care on mortality or length of stay.

Patients treated at home were more satisfied with the quality and location of their care. The committee agreed other potential benefits of home-based care such as avoiding deconditioning, reducing the risk of hospital-acquired infection, freeing up hospital beds and reducing demand on acute inpatient services. They noted that home-based care could help avoid hospital admission for people attending primary care or emergency departments.

The committee discussed the existing recommendations on using CBR65 or CURB65 scores together with clinical judgement, to inform shared decisions about place of care. They agreed they should be retained but added the option to refer to a virtual ward or SDEC unit or hospital at home service. They also identified a list of factors that would need to be discussed with the person when making a shared decision about using home-based models of care. They noted that family and carers should be consulted, where appropriate, for example, if they will be involved in supporting the person at home.

There was no evidence for children, so the committee was not able to make any recommendations about home-based models of care for children and young people.

How the recommendations might affect practice

Most trusts now have established virtual wards or hospital at home services, so where those services already exist, the recommendations are likely to have a positive impact on reducing hospital admissions and length of hospital stay. This would reduce pressure on hospital inpatient care and potentially free up beds for more severely unwell people who require hospital admission. Early discharge of people who are improving could also improve inpatient capacity.

Return to recommendations

Why the committee made the recommendation for research

The evidence was reviewed on risk assessment tools and clinical prediction tools for identifying children and young people attending primary care who may be at risk of deterioration. Recommendations were unable to be developed in this area.

Many children and young people presenting to primary care with respiratory tract infection symptoms will not require antibiotics and can be safely cared for at home. A small number can deteriorate and require secondary care. Identifying those most at risk is important to ensure quick and appropriate referral as well as preventing over-referral to secondary care, so the committee made a recommendation for research on prediction tools for assessing under 18s in primary care with suspected pneumonia.

Why the committee made the recommendation for research

The evidence was reviewed on the prognostic accuracy of the national early warning score 2 (NEWS2) and paediatric early warning system (PEWS) for people with pneumonia. Recommendations were unable to be developed in this area.

The committee noted that CURB65 is a validated tool that can be used, alongside clinical judgement, to assess the severity of community-acquired pneumonia in hospital. The committee made a recommendation for research to identify or develop tools or methods for stratifying people with hospital-acquired pneumonia according to disease severity. This could include validating an existing model.

Why the committee made the recommendation

Evidence showed that lung ultrasound is a good diagnostic tool for people with suspected community-acquired pneumonia, showing high accuracy for both ruling in and ruling out the diagnosis in symptomatic people. The accuracy of lung ultrasound did not depend on the setting within the hospital where it was used. The evidence was limited to people with community-acquired pneumonia, but the committee agreed that it could be extrapolated to people with hospital-acquired pneumonia.

The committee discussed the benefits of lung ultrasound compared with chest X-ray, particularly that it avoids radiation exposure and can be used to rapidly consider diagnosis of pneumonia and promptly start antibiotic treatment. Lung ultrasound can be performed at the bedside or point of care during other clinical examinations, which in some situations may be more efficient and less time consuming than transporting someone to the X-ray department and subsequently accessing the images. As such, the committee agreed lung ultrasound may be particularly helpful for initial assessment of a sick or deteriorating person with suspected pneumonia or where several possible causes are being considered for the presenting symptoms.

Lung ultrasound can take longer to perform than chest X-ray and it can be more difficult to visualise the entire lungs, particularly in people who are lying down or in unwell children who are not able to sit still for the time required to complete the examination. Chest X-ray may detect other potential illnesses or problems with the lungs, such as tumours or mediastinal abnormalities. The committee was also concerned about the storage and audit of lung ultrasound images, specifically that it may not be possible to save diagnostic images for later review by other healthcare professionals or for monitoring changes. There was also a discussion about the current lack of trained operators with sufficient experience to perform diagnostic lung ultrasound for pneumonia. The committee agreed that lung ultrasound can be used as a diagnostic tool but should not replace chest X-ray for confirming a diagnosis of pneumonia.

How the recommendation might affect practice

The recommendations may increase the use of lung ultrasound beyond current levels, particularly point-of-care use in emergency departments or critical care areas to help inform a diagnosis of pneumonia and promptly start antibiotic treatment. This may mean a quicker diagnosis, starting treatment earlier and enabling assessment of complications.

To respond to the potential increased use of lung ultrasound, more clinicians will require training and accreditation in this procedure and time to build up experience of this imaging method, which is potentially an implementation issue.

Return to recommendation

Why the committee made the recommendations

The evidence showed a link between elevated biomarker levels (C-reactive protein [CRP], procalcitonin [PCT] and neutrophil to lymphocyte ratio [NLR]) and adverse outcomes in people with pneumonia. The evidence also showed that patients treated for pneumonia in hospital whose CRP, PCT or NLR levels remained high or increased between admission and day 3 or 4 were more likely to experience adverse clinical outcomes than people whose biomarkers decreased. Although NLR showed a similar pattern of findings to CRP and PCT, there was less evidence on NLR than CRP or PCT and it is not routinely used (despite being easily obtained through routine blood samples), so the committee agreed to focus on the use of CRP and PCT.

The committee considered extending the existing recommendation on measuring CRP on admission to hospital for adults with community-acquired pneumonia to also include PCT. They concluded that adding admission PCT would not give substantially more information than CRP alone so this would be an unnecessary extra test, with additional cost.

The committee agreed that CRP or PCT should be measured 3 to 4 days after starting treatment where there is concern about treatment failure. This will allow clinicians to monitor the person's CRP or PCT levels at this time and any changes to their levels if a baseline level has been taken. Low or decreasing levels may help rule out complications or poor prognosis, while high levels, or levels that remain elevated, may help identify when senior clinical review is needed. The committee noted that CRP levels that fail to halve in 3 to 4 days are a cause for concern. They highlighted that both the absolute level of CRP or PCT, as well as any change in levels from baseline to repeat assessment, can be useful. They agreed that patients who are responding well to treatment and whose clinical condition is improving are unlikely to need follow-up biomarker testing.

Additional evidence compared antibiotic treatment using a PCT-guided algorithm with standard antibiotic treatment and showed that patients treated using the PCT-guided approach took antibiotics for a shorter time and experienced fewer side effects. The reduction in duration of antibiotic treatment did not affect rates of pneumonia recurrence, re-hospitalisation, intensive care unit (ICU) admission or mortality. However, the committee was concerned that the average standard duration of antibiotic treatment used in the trials of 10 to 12 days was much longer than the UK current recommended practice of 5 days. Furthermore, the trials did not give the committee any information on whether it was safe or effective to use PCT levels to reduce antibiotic treatment below 5 days, so they were not able to make a recommendation about this.

There was no evidence linking biomarkers to other aspects of de-escalating care such as discharge from ICU or discharge home, so the committee was unable to make any recommendations about these aspects of care for people with pneumonia.

How the recommendations might affect practice

The recommendations could increase the use of CRP and PCT testing in people in hospital with pneumonia who are not responding to treatment after 3 or 4 days, but this could help identify complications or allow clinicians to amend the treatment plan to be more effective. Although there may be an increase in testing, it would not require significant changes in procedures.

Return to recommendation 1.4.3

Return to recommendations 1.11.7 and 1.11.8

Why the committee made the recommendations

The committee discussed the evidence on use of microbiological tests for people presenting to secondary care with suspected community-acquired pneumonia. They agreed that the evidence supported the existing recommendation to not routinely offer microbiological tests to adults with low-severity community-acquired pneumonia, so chose to retain it. They also agreed to add children with non-severe community-acquired pneumonia to this recommendation.

Noting the lack of evidence identified in this area, the committee discussed when to consider blood cultures for adults with moderate- and high-severity community-acquired pneumonia and children with severe community-acquired pneumonia if there are additional clinical indications such as suspected sepsis. Where sepsis is suspected, clinicians are advised to follow NICE's guideline on sepsis.

The committee noted that blood samples should ideally be taken before starting antibiotics because antibiotics may impact blood culture results.

No evidence was identified for hospital-acquired pneumonia. The committee noted the existing recommendation on sending microbiological samples for those with hospital-acquired pneumonia. They recognised that this would mean a different approach when compared with community-acquired pneumonia but agreed this was reasonable given the possible differing microbiology findings for community-acquired pneumonia and hospital-acquired pneumonia.

Noting the lack of evidence identified in this area, and the existing recommendations, the committee discussed and agreed a recommendation to consider sputum culture for adults, and for children if age appropriate as they acknowledged that sputum samples from younger children may be difficult to obtain.

The committee agreed that infection with Legionella pneumophila is a relatively rare cause of pneumonia. It is mostly seen in people who have been exposed to stagnant water. Exposure to legionella can also occur at the workplace, for example among people who frequently work on air conditioning units. Therefore, they recommended that legionella urinary antigen tests should only be considered if there are risk factors for legionella infection.

For adults assessed via clinical judgement and CURB65 score as having moderate- or high-severity community-acquired pneumonia, there was evidence to support the use of pneumococcal urinary antigen tests to aid decision making around the selection of the most appropriate antibiotic. This will benefit people with the condition, help improve antimicrobial stewardship and could support de-escalation to a narrower antibiotic spectrum.

The evidence on pneumococcal urinary antigen tests for babies and children was very limited and mostly not directly applicable to the UK. Based on their expertise, the committee discussed the implications of using urinary antigen tests in babies and children. They made a consensus recommendation not to routinely use these tests for this population as, in their experience, they are not useful in practice.

There was no evidence identified on mycoplasma testing. The committee discussed this and noted there is limited demand for this test and a positive result may not impact usual antibiotic prescribing decisions, so they did not make a recommendation about this.

The committee acknowledged that overuse of antibiotics is associated with antimicrobial resistance and is a national and global priority. However, the evidence on how microbiological tests may help reduce rates of empirical prescribing and support more directed antibiotic therapy remains limited. Therefore, they made a recommendation for research into which tests could safely reduce inappropriate antibiotic prescribing in people with suspected pneumonia.

How the recommendations might affect practice

Blood and sputum cultures and urinary antigen tests were previously recommended and therefore will not require additional resource to implement.

The new recommendations have the potential to reduce the number of blood and sputum cultures taken in adults with moderate- and high-severity community-acquired pneumonia, potentially leading to cost savings for the NHS. Only using legionella urinary antigen tests for adults where needed may also be a cost-saving strategy.

Urinary antigen tests are occasionally used in babies and children. With the new recommendation, their use will potentially decrease, leading to fewer negative outcomes and potential cost savings for the NHS.

Return to recommendations

Why the committee made the recommendations

There was evidence that for babies and children (up to 11) with community-acquired pneumonia, a 3‑day course of antibiotics was as effective as a 5‑day course. There were overall limitations to the evidence found, particularly around the applicability of the populations included, with only 1 UK-based study identified. A difference was not identified in the adverse effects outcomes between the shorter or longer duration courses, this included with the frequency of re-admission for further antibiotics. The committee discussed the lower age and agreed that for a child under 3 months they would have concerns about reducing the duration of antibiotics and agreed to keep this as 5 days for this age group. The committee further noted that shorter courses of antibiotics are important as part of antimicrobial stewardship.

The committee discussed that antibiotics are usually less effective or may be ineffective in children with a cough or lower respiratory tract infection not caused by pneumonia, so it is important that the diagnosis is community-acquired pneumonia.

The committee noted that not all community-acquired pneumonia resolves as expected, and longer courses of antibiotics may be needed in some babies and children. This should be guided by clinical judgement.

Symptoms of pneumonia can last a long time and having symptoms after stopping antibiotics does not mean that the antibiotics have not worked. The committee agreed that this was an important point to convey to parents and carers, so this is reflected in the recommendations in this area.

How the recommendations might affect practice

The recommendations will reduce duration of antibiotic use for treating community-acquired pneumonia in babies and children. They should contribute to antimicrobial stewardship aims.

Return to recommendations

Why the committee made the recommendations

The evidence on use of antibiotics plus corticosteroids, compared with antibiotics alone, for treating community-acquired pneumonia in adults showed it reduced mortality rates and time spent in hospital and ICU, particularly for adults with high-severity pneumonia.

The evidence showed an increased risk of hyperglycaemia with use of corticosteroids and, for less severe pneumonia, an increase in the risk of secondary infections. The committee noted that these risks are not specific to pneumonia but are among the known side effects of corticosteroids. They concluded that the benefit in terms of reduced mortality outweighed the risk of adverse effects for people with high-severity pneumonia, but not for low- and moderate-severity pneumonia. The committee agreed high-severity community-acquired pneumonia in hospital to mean a CURB65 score of 3 to 5 combined with clinician judgement of high disease severity, which may include physiological instability, shock, profound hypoxia, or need for mechanical ventilation.

The evidence suggested that intravenous (IV) hydrocortisone may be more effective than other corticosteroids, though a direct comparison was not available. The committee discussed this and agreed that a recommendation without any indication of the type of steroid to use would not be helpful to clinicians. Therefore, they included in the recommendation to consider starting treatment with IV hydrocortisone. The committee did not review any evidence on the most effective dose, so were not able to include this in the recommendation.

The committee was made aware of evidence from a newly published trial (REMAP-CAP) after completion of the evidence review. The trial was stopped early because the results did not show a pre-specified large mortality benefit of steroids for people with severe community-acquired pneumonia. The committee agreed that this did not rule out the possibility of there being a smaller effect of steroids on mortality in line with that suggested by the evidence review, and noted methodological approaches with the trial that led them to be cautious about the findings. Further results on the use of steroids for community-acquired pneumonia are expected in ongoing studies.

There was a study in the evidence for children. This study did not report the main outcomes of interest, except for adverse events. The committee reflected on the evidence for the use of corticosteroids in adults and discussed the implications of any extrapolation to children. However, given the weak evidence base and limited applicability of the included trials, the committee agreed that this did not currently support a recommendation on the use of corticosteroids for children with pneumonia.

The committee agreed that further research was needed on use of adjunctive corticosteroid antibiotics in people hospitalised with community-acquired pneumonia or hospital-acquired pneumonia, including for babies, children and young people. They made a recommendation for research on the effectiveness of corticosteroids including which type, dose and route of administration is most effective and whether effectiveness varied depending on the type of pathogen being treated.

How the recommendations might affect practice

The evidence suggests use of corticosteroids is likely to improve outcomes for adults with high-severity community-acquired pneumonia. The additional use of corticosteroids incurs a small cost, but this is likely to be outweighed by the potential benefits such as reduced stay in ICU or overall time spent in hospital, freeing up resources and service capacity.

Return to recommendations

Why the committee made the recommendations

The evidence on use of high-flow nasal oxygen (HFNO), continuous positive airways pressure (CPAP) and non-invasive ventilation (NIV) compared with standard oxygen therapy for people with pneumonia was limited. The committee noted that the populations studied were small and only partially applicable with not all participants having pneumonia. They also noted the evidence showed lack of adverse effects and no impact on mortality at 30 days.

Based on the evidence and their expertise and experience, the committee agreed that HFNO was their preferred option because it is less invasive and better tolerated than NIV and CPAP, has fewer safety concerns, and allows the person to eat and drink.

The committee also agreed the importance of considering the person's clinical trajectory in multidisciplinary team decisions about trialling HFNO. Lay members emphasised the importance of respecting people's preferences, describing the discomfort of NIV and CPAP compared with HFNO.

The ability of different locations within the hospital to deliver HFNO, CPAP or NIV was discussed. A number of factors were identified that should be taken into account when making decisions around this, including the possible need for escalation of care.

The committee agreed that some people with pneumonia and a coexisting condition, such as type 2 respiratory failure in a person with chronic obstructive pulmonary disease or acute pulmonary oedema in a person with heart failure, may benefit from a trial of NIV or CPAP for respiratory support.

How the recommendations might affect practice

The recommendations align with current practice in the UK. No cost implications are expected as hospitals are likely to already have the resources to deliver non-invasive ventilation or CPAP and will already routinely use HFNO in acute care areas.

Return to recommendations

Why the committee made the recommendations

The committee agreed that it is important to give parents and carers of children with community-acquired pneumonia advice and information on the usual timeframe for symptom improvement. They agreed that although there can be variation in the time to symptom resolution, for most otherwise healthy children, their symptoms will steadily improve after starting treatment. The committee acknowledged that some symptoms take longer to resolve than parents or carers may expect, particularly cough, which can contribute to unnecessary repeat visits to the GP, so they wanted to reassure parents and carers that cough may persist for up to 4 weeks.

The committee discussed symptoms that may indicate more serious illnesses or complications. These symptoms include persistent fever, increased work of breathing, and reduced fluid intake or fatigue. They agreed parents or carers should seek further advice if their child continued to present with those symptoms.

Given the frequency and severity of hospital-acquired pneumonia it would be useful to be able to offer people and their families similarly informed advice. However, searches revealed no evidence to support advice on recovery trajectories in hospital-acquired pneumonia, so no recommendation was made for hospital-acquired pneumonia for parents and carers.

How the recommendations might affect practice

The committee agreed that providing parents and carers with information about expected symptom duration in children could reduce unnecessary visits to GPs and other services about symptoms that will resolve with time without the need for further treatment or testing. They also noted that by outlining symptoms that may indicate complications or a deterioration in their child's condition, parents and carers may be better able to identify when they should reconsult a healthcare professional and this may mean their child is seen earlier. This could reduce downstream costs of treatment and resource use.

Return to recommendations

Why the committee made the recommendations

The evidence looked at follow-up chest X-rays following discharge from hospital for detecting cancers and other lung conditions and to check whether the pneumonia has resolved. The X-rays took place 4 to 8 weeks after discharge from inpatient care. The evidence showed a cancer detection rate of around 2% and moderate to high rates of pneumonia resolution.

The committee agreed that the evidence suggested that radiological changes may persist after symptoms of pneumonia have resolved and these do not always indicate a need for further investigation or treatment. As such, offering routine follow-up chest X-rays to all people hospitalised with pneumonia is not always useful for checking if pneumonia has resolved. It was also unclear if it would provide timeliness of cancer detection. The committee noted that the severity of hospital-acquired pneumonia and the frequent association with comorbidities implied that follow-up radiological assessment may be beneficial, but no evidence was available.

The evidence suggested that people whose follow-up chest X-ray detected cancer were older and had a history of smoking. The committee agreed that smoking or being aged over 50 years are risk factors for lung cancer and other underlying respiratory disease, so follow-up chest imaging should be considered. They agreed that clinicians should also consider follow-up chest X-rays for people with unresolved symptoms or who have unexplained weight loss because these can also be indications of cancer or other underlying conditions.

The committee discussed that people with hospital-acquired pneumonia may have had recent chest imaging during their admission for non-pneumonia reasons, so there may not be a requirement to perform further chest imaging in this group. The committee agreed that the result of recent chest imaging should be considered when deciding whether to request a follow-up chest X-ray.

The committee acknowledged that some people may not want to attend for further investigations. The committee agreed that people's preferences and medical factors should be discussed so a shared decision can be made about follow-up chest X-ray.

The evidence only looked at cancer detection in people who received a follow-up chest X-ray at 6 weeks; there was no information on the cancer detection rate in people who did not receive a follow-up chest X-ray, so the committee made a recommendation for research on the clinical and cost effectiveness of follow-up chest imaging for adults discharged from hospital after treatment for pneumonia.

The committee noted that information on long-term survival rates of people diagnosed with cancer as a result of routine follow-up chest X-rays and stage of cancer detected would allow better understanding of the benefits of cancer detection using this method. The committee agreed that this research should focus on adult populations only, because the evidence showed that follow-up chest X-rays are not clinically useful for children and young people.

How the recommendations might affect practice

These recommendations should reduce the number of chest X-rays required, although a large proportion of people with pneumonia will have the risk factors listed. This would prevent unnecessary investigations and reduce demand on imaging services and associated administration.

Return to recommendations