Pathophysiology

Bacterial infection of the lung parenchyma results in an inflammatory response from the host, with an outpouring of neutrophils and exudate into the alveolar spaces, resulting in consolidation. This compromises the oxygen exchange because of the ventilation‐perfusion mismatch and results in type 1 respiratory failure. Inflammation of the pleura results in pleuritic chest pain and the development of a parapneumonic pleural effusion in a third of those with CAP. An empyema may develop in a certain percentage as discussed in Chapters 10 and 12. Table 8.1 lists the causes, typical clinical and radiological features, and the treatment of the common pneumonias.

Diagnosis of CAP

The diagnosis of pneumonia is made on the clinical symptoms and signs, and a chest X‐ray (CXR) confirming new parenchymal shadowing. Those presenting with symptoms suggestive of a CAP should have a comprehensive history, examination, and investigations to confirm the diagnosis of CAP and to determine the severity. Investigations should include blood tests, CXR, urinary pneumococcal and legionella antigens and sputum for microbiological analysis. A CURB‐65 score should be calculated in every patient presenting with a CAP; this guides management and has prognostic implications. Box 8.3 describes how the CURB‐65 score is calculated.

A full blood count will usually show an elevated white cell count, with a raised neutrophil count. Leukopenia, with a white cell count <4 × 10⁹ l⁻¹, is associated with a poorer outcome. Mycoplasma pneumonia can be associated with cold agglutinins and a haemolytic anaemia. An elevated eosinophil count should raise the possibility of an eosinophilic pneumonia and will warrant further investigations, such as a bronchoalveolar lavage (BAL), a lung biopsy, and a HRCT. Eosinophilic pneumonias are discussed in Chapter 7.

A raised urea signifies a worse prognosis. Hyponatraemia secondary to the syndrome of inappropriate anti‐diuretic hormone (SIADH) can be associated with CAP, particularly legionella pneumonia. The CRP is always elevated in bacterial CAP, with a sensitivity of 73% and a specificity of 65%, although there is usually a lag, so it may not be raised at the onset of symptoms. Serial CRP measurements can be useful in monitoring response to treatment. Pro‐calcitonin (PCT), a peptide precursor of calcitonin that is released by cells in response to bacterial toxins, can also be used to distinguish between bacterial and non‐bacterial causes of pneumonia. PCT levels may also correlate with the severity of pneumonia.

Abnormal liver function tests (LFT), particularly a raised alanine transaminase (ALT) level, can occur with legionella or mycoplasma pneumonia. Deranged LFTs can also occur after treatment with intravenous antibiotics, particularly macrolides.

Microbiological diagnosis of CAP

Patients with a low‐severity CAP do not routinely require sputum analysis. However, sputum should be sent in those presenting with moderate or severe CAP and when there is cavitation on the CXR. If the patient has a productive cough, then a deep cough sputum sample should be sent for Gram staining, culture, and sensitivity prior to starting antibiotics. There is huge variation in getting a positive sputum result, ranging from 10–80%. Some infections, such as Streptococcus aureus, are easily cultured, whereas Streptococcus pneumonia and Haemophilus influenzae are more difficult to culture, so false negative results can occur. Culture results are reported according to the amount of growth, with moderate or heavy growth indicating a true pathogen, whereas a light growth may indicate colonisation.

Bronchoscopy and bronchoalveolar lavage (BAL) samples are not routinely required, but should be considered in the immunocompromised patient, those who have an abnormal CXR suggestive of malignancy, and in those not improving on empirical antibiotic therapy. BAL samples should have Gram staining, Ziehl‐Neelsen staining for acid‐fast bacilli (AFB), silver staining for pneumocystis jerovici, and stains for fungi. Cultures can take several days to weeks.

Blood cultures should be taken in those presenting with fever and other symptoms of sepsis, even in the absence of fever. Blood cultures are positive in 12% of hospitalised patients, with two‐thirds growing Streptococcus pneumonia. Care should be taken when collecting blood for culture as there is a 10% rate of contamination with MRSA from the skin.

If Streptococcus pneumonia or legionella pneumophila infections are suspected, then urinary antigen testing is more sensitive and specific than blood cultures or sputum cultures, and gives a result more rapidly than cultures of sputum or blood. Urinary antigen testing has the advantage that the antigen will remain positive even after starting antibiotics, but as no pathogens are available, it is not possible to determine sensitivities. Therefore, it is recommended that sputum or BAL samples are also sent for microscopy, culture, and sensitivity.

Polymerase chain reaction (PCR) diagnostic kits are available to use on sputum samples which can give a result within a few hours for chlamydophila pneumoniae and mycoplasma pneumonia, although care must be taken to limit contamination from upper airway flora. If no sputum is available, then a throat swab for mycoplasma pneumonia PCR is recommended.

Complement fixation test (CFT), or paired serological test, can be used to diagnose legionella and mycoplasma infections, especially during outbreaks, and will show a fourfold rise in antibody titres. CFT is also recommended in any patient under 40 years presenting with pneumococcal pneumonia. The consultant microbiologist and public health consultant will usually be able to give advice about outbreaks and put in place public health safety measures, such as closure of infected hotels.

CMV serology can be requested if CMV pneumonia is suspected. Human immunodeficiency virus (HIV) test is recommended in any patient with an atypical presentation and an abnormal CXR.

Microbiological diagnosis is made in less than 40% of patients presenting with CAP. However, empiric antibiotic treatment results in outcomes as good as if the pathogen were detected, with only 1% treated in the community for CAP requiring hospitalisation because of treatment failure. The choice of initial antibiotic therapy is therefore made on the likelihood of the infecting organism.

Radiological diagnosis of CAP

The chest X‐ray (CXR) is an essential investigation in making the diagnosis of CAP. NICE and British Thoracic Society (BTS) guidelines recommend that a CXR should be done in a timely way so that antibiotics can be prescribed within 4 hours of the patient presenting to hospital. The CXR may appear normal very early on or show interstitial infiltrates which are better seen on HRCT (Figure 8.1). In established CAP, the CXR will show consolidation in one or more lobes or show patchy consolidation of bronchopneumonia (Figure 8.2, Figure 8.3). Radiographic changes do not correlate with the pathogen, although Streptococcus aureus pneumonia and Klebsiella pneumonia present with cavitating lesions, the differential diagnoses for which include Mycobacterium tuberculosis infection, lung cancer, and vasculitis.

The differential diagnosis of CAP includes an exacerbation of COPD, exacerbation of asthma, acute bronchitis, pulmonary oedema, pulmonary embolus, adenocarcinoma in situ, eosinophilic pneumonia, hypersensitivity pneumonia, cryptogenic organising pneumonia or diffuse parenchymal lung disease. If there is no clinical improvement with appropriate antibiotics, then further investigations will be required.

Management of CAP

Patients presenting to their General Practitioner (GP) with symptoms and signs of a CAP should have a CXR and blood tests, including CRP measurement, to guide management. Pulse oximetry may be helpful in determining whether a patient needs to be admitted to hospital.

NICE Guidelines recommend that a five‐day course of a single antibiotic should be given to those with a CRP of greater than 100 mg l⁻¹ but with a CURB‐65 score of less than 2. The choice of antibiotic will depend on local antibiotic prescribing guidelines and any antibiotic allergy that the patient may have. NICE recommends amoxicillin or tetracycline, and a macrolide or tetracycline for those with penicillin allergy. If symptoms do not resolve within three days, then the duration of antibiotic therapy should be increased. For those with a CRP between 20 and 100 mg l⁻¹, a delayed antibiotic prescription should be considered, and the patient should be instructed to take the antibiotic if symptoms worsen. Patients with a non‐severe CAP should expect clinical improvement within a week and complete resolution of symptoms by 6 weeks, although many report feeling fatigued for up to three months.

Radiological resolution usually takes up to six weeks after antibiotic treatment is completed. It is recommended that a CXR is done after 6 weeks to ensure complete resolution of changes.

Patients seen in hospital with a CURB‐65 score of 2 or less and with no other adverse prognostic features should be started on oral dual antibiotic therapy with amoxicillin and a macrolide for 7–10 days. Those who are allergic to penicillin should be given either a macrolide alone or a tetracycline. If there are no adverse features, they could be discharged home with follow‐up by the GP. The patient should be advised to rest, have adequate hydration, and seek medical help if their symptoms do not improve.

Patients with a diagnosis of CAP who have a CURB‐65 score > 2 and those who have adverse prognostic features should be hospitalised and receive intravenous antibiotics, intravenous fluids, oxygen, and thromboprophylaxis to prevent pulmonary emboli. Adverse features include fever, respiratory rate > 24 breaths per minute, tachycardia, systolic blood pressure < 90 mmHg, oxygen saturation < 90% on room air, and confusion.

The aim should be to maintain oxygen saturation in the range of 94–98%, ensuring that there is no evidence of CO₂ retention. Empirical intravenous antibiotic therapy should be started without delay once appropriate samples have been sent for microbiological analysis. These patients should be given nutritional support, a mucolytic agent, and physiotherapy for sputum clearance. Patients with COPD and asthma may benefit from regular nebulised bronchodilators.

The choice of empiric antibiotic therapy for CAP is based on the likelihood of a specific pathogen, and the local antibiotic prescribing guidelines which are based on resistance patterns. Microbiological classification prior to treatment is not practical as the organism is not always identified and waiting for identification may result in treatment delay.

As Streptococcus pneumonia is the commonest cause of CAP, it is recommended that intravenous amoxicillin or benzylpenicillin, together with a macrolide, is given. This combination has been shown to decrease mortality and length of hospital stay. Patients who are allergic to penicillin should be commenced on a macrolide. Macrolides can cause prolongation of the QT interval and can interact with other drugs (discussed in Chapter 3). The prevalence of macrolide‐resistant Streptococcus pneumonia is on the increase. Fluoroquinolones, such a levofloxacin and moxifloxacin, are also active against Streptococcus pneumonia so could be used as monotherapy. However, these drugs can also cause prolongation of the QT interval and ventricular arrhythmias in the elderly. Vancomycin and Teicoplanin are reasonable options in those who have travelled abroad where penicillin and macrolide resistance are high.

The antibiotic regime should be reviewed after 48 hours. If the patient is improving clinically, the fever has settled, and the CRP is coming down, then the intravenous antibiotics should be changed to oral antibiotics. There is no trial evidence regarding the optimal duration of antibiotic therapy, but treatment for 5–10 days is usually given as this results in improvement while minimising the risk of antibiotic‐associated clostridium difficile infection.

Patients with a CURB‐65 score of 3 or 4 with evidence of sepsis may develop hypotension and severe hypoxaemia requiring inotropic support, intubation, and ventilation on ICU.

A parapneumonic effusion occurs in 30–50% of patients with a CAP and will usually resolve without any intervention (Figure 8.4). In some cases, the effusion can progress to an empyema, the diagnosis and management of which are discussed in Chapter 12. Certain organisms, such as Staphylococcus aureus and Klebsiella pneumonia, can predispose to the development of a lung abscess, which is discussed in Chapter 12.

Prognosis with CAP

Predictors of mortality include the CURB‐65 score and the Pneumonia Severity Index (PSI) which is based on the patient’s gender, age, co‐morbidities (diabetes, cardiac failure, renal failure), results of clinical examination, blood test results, and CXR findings. Additional adverse features include hypoxaemia (SaO₂ <92% or PaO₂ <8 kPA), white cell count >20 × 10 9 l⁻¹ or <4 × 10 9 l⁻¹, multilobe involvement and positive blood culture. Leukopenia, thrombocytopenia, and a raised serum glucose concentration in a non‐diabetic patient, are also predictors of mortality.

A CT thorax should be considered when the CXR shows no improvement in the radiological changes, when there is a cavitating lesion, possible adenopathy, or clinical features of malignancy. These patients may also require a bronchoscopy to see if there is an obstructing lesion.

Mortality from CAP ranges from 5.1–13.6% (for all CURB‐65 scores) in the community, but increases to 36% in patients who require admission to ICU. Young patients with a CURB‐65 score of 0 have a good prognosis with a mortality of <1%. The majority will return to full health within a few weeks. The morbidity and mortality are greater in those over the age of 65 and those with co‐morbidities, such as diabetes and COPD. Patients with a CURB‐65 of 2 have a ninefold increase in risk of death. Those who survive complain of persistent fatigue, cough, and breathlessness. There is an increased risk of death in survivors over the next three years, with a one‐year mortality of 27%, as hypoxia and the acute inflammatory response due to CAP are associated with death due to acute cardiac events. Box 8.4 lists the mortality associated with CURB‐65 score.