| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS | REGISTER |
|
|
|
|
|||||||||||||
|
|
|||||||||||||||
 | 1 Introduction |
|---|
 TOP 1 Introduction 2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
Ideally, the definition would include the isolation of a responsible organism. However, it is apparent from many studies that a pathogen is not identified in a significant proportion of cases that otherwise meet the clinical definition (see Section 3 on Aetiology). As it is assumed that CAP is caused by infection, the presumption is that current techniques have insufficient sensitivity to detect all relevant pathogens. Treatment guidelines therefore have to assume that, where pathogens are isolated, they represent all likely pathogens. There is a clear need for better diagnostic methods.
In creating guidelines it is necessary to assess all available evidence with consideration of the quality of that evidence. This we have endeavoured to do. We have then produced key points and management guidelines based on the available evidence supplemented by consensus clinical opinion where no relevant evidence was found. A summary of the key points and a further summary prepared specifically for use in primary care are also available on the Thorax website (www.thoraxjnl.com) and the British Thoracic Society website (www.brit-thoracic.org.uk).
METHODS OF GUIDELINE DEVELOPMENT
Scope of guidelines
These guidelines address the management of CAP in infants and children in the United Kingdom. They do not include neonates or infants with respiratory syncytial virus bronchiolitis. The specific management of children with pre-existing respiratory disease or that of opportunistic pneumonias in immunosuppressed children is not addressed.
Guideline development group
The guideline development group was set up by the BTS Standards of Care Committee to produce guidelines for children in parallel with those being produced for adults. It comprised three paediatricians with a special interest in respiratory disease, a paediatrician with special interest in paediatric infectious diseases, a specialist registrar in paediatrics, a paediatric nurse, a general practitioner, and a guidelines methodologist. No external funding was obtained to support the development of the guidelines. Because of the breadth of scope of the topic, the guideline development was divided up into 12 sections and members were allocated to each. Eight of the 12 sections had at least two members allocated.
Identification of evidence
Search strategies were developed for each of the 12 sections (excluding guideline methodology) with the assistance of an information scientist. These search strategies (see Appendix 1) included MeSH and free text terms and had no language restrictions. They were run on Medline (Winspirs, Silverplatter) and the Cochrane Library (Issue 3, 1999). Where searches yielded more than 1000 citations, these were limited to English.
Assessing the literature
The sets of references generated by researchers were sifted for relevance to the clinical topic of the guidelines. Where two or more members were working on a section this was done independently. Initial sifting was on the basis of the title and abstract (as obtained from the specialist resource). Where there was doubt about whether a reference was relevant, the full publication was obtained. Studies from countries where the populations or clinical practices were very different from the UK were excluded unless they addressed questions which could be generalised to the UK (such as clinical assessment). The methodological quality of the publication was assessed using a checklist adapted from one previously developed for this purpose.1
Synthesising the evidence
Once individual papers were checked for methodological rigour and clinical relevance they were categorised according to study design.1 The evidence was synthesised by qualitative methods. The content of the identified papers was summarised into brief statements that were thought accurately to reflect the relevant evidence and the category of that evidence was indicated after each citation. The recommendations resulting from the evidence were graded according to the strength of that evidence (table 1
). The strength of each recommendation ([A], [B], [C], or [D]) was indicated after each recommendation. Where there was no identifiable evidence, and it was felt important to provide a recommendation within the guidelines, the lack of evidence was clearly stated and a recommendation based on a consensus from the group was provided (grade [D]).
|
Updating of guidelines
It was agreed that the guidelines should be reviewed for the content and evidence base no later than 3 years after completion.
Acknowledgements
The British Thoracic Society provided support for travel and telephone conference costs of the Working Group.
Conflicts of interest
PH is a consultant to Wyeth Vaccines on pneumococcal conjugate vaccines.
| Synopsis of main recommendations Aetiology and epidemiology
Clinical features
Radiological investigations
General investigations
Microbiological investigations
Severity assessment
General management
Antibiotic management
Complications
|
| 2 Incidence and mortality |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
Pathogen
Based on serological results, the authors of the first Finnish study were able to calculate the incidence of CAP by pathogen. For those less than 5 years of age, Streptococcus pneumoniae had an incidence of 8.6/1000/year, Mycoplasma pneumoniae 1.7/1000/year, and Chlamydia species 1.7/1000/year. For those aged 5–15 years the incidence figures were 5.4/1000, 6.6/1000, and 3.9/1000, respectively. The sex difference noted in those less than 5 years of age was mainly accounted for by S pneumoniae (11.2/1000 in boys and 5.7/1000 in girls). These figures include mixed infections2 [Ib].
Another population based prospective surveillance study of invasive pneumococcal disease (based on culture positive cases only) was performed in Southern California between 1992 and 1995, yielding an incidence of pneumococcal pneumonia of 17/100 000/year in children aged 2 years or less. There were no deaths5 [Ib].
Risk factors
Only one case control study of risk factors for CAP in a developed country has been performed. The cases were those identified in the Finnish prospective study.2 For children under 5 years of age, recurrent respiratory infections during the previous year, a history of wheezing episodes, and a history of acute otitis media treated by tympanocentesis before the age of 2 years were found to be significant in a multivariate model. For older children (5–15 years of age) a history of recurrent respiratory infections in the previous year and a history of wheezing episodes were found to be significant risk factors2 [III].
| 3 Aetiology and epidemiology |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
Studies of specific pathogens are summarised in table 2. All of these are prospective studies in which pneumonia was community acquired and where the case definition includes clinical findings compatible with pneumonia together with radiological changes. All constitute levels of evidence of [Ib] or [II] (indicated). In the columns the percentage indicates the percentage of all CAP cases in which that organism was the only isolate detected. Where other isolates were also isolated it was classified as mixed and indicated in a separate column. In some studies it was not possible to determine whether infections were single or mixed (as indicated). Bacterial isolates are not included if isolated from a sputum or upper respiratory tract specimen in the absence of other evidence of significance—for example, a rise in antibody concentrations.
|
A number of viruses appear to be associated with CAP, the predominant one being respiratory syncytial virus (RSV). Others isolated include: parainfluenza, adenovirus, rhinovirus, varicella zoster virus, influenza, cytomegalovirus, herpes simplex virus, and enteroviruses. Overall, viruses appeared to account for 14–35% of CAP cases in childhood [II].
Quantifying the proportion of CAP caused by bacteria is more difficult. Streptococcus pneumoniae is assumed to be the most common bacterial cause of CAP but is infrequently found in blood cultures. It is commonly found in routine cultures of upper respiratory specimens, yet is known to be a commensal in this setting. Antigen detection methods of urine are unreliable7 [II]. Serological testing is a promising non-culture technique but responses will be age related. Overall, blood or pleural fluid culture of S pneumoniae is positive in 5–10% of cases of CAP [Ib]. The proportion of CAP due to S pneumoniae increases to 16–37% where serological testing is used [II]. Other bacterial pathogens appear to be less frequent causes of CAP. Claesson et al11 assessed the antibody responses to non-capsulated Haemophilus influenzae and isolated it as the only pathogen from the nasopharynx of 43 of 336 children but a significant increase in IgG or IgM was shown in 16 (5% of all CAP) [II]. In the same study 3% also had a significant increase in antibodies to Moraxella catarrhalis, suggesting that it too is an uncommon cause of CAP in children.12 This was supported by another study by Korppi et al13 in which seroconversion to M catarrhalis was documented in only 1.5% of cases of CAP [II].
In these studies Mycoplasma pneumoniae accounted for 4–39% of isolates [II]. Where Chlamydia pneumoniae is sought, it appears to be a significant pathogen responsible for 0–20% of cases [II]. Biases which need to be considered in these reports include whether children with mycoplasmal (or chlamydial) pneumonia are over represented in hospital based studies because of failure of penicillin related antibiotic treatment in the community, or are over represented in community studies because they are less sick and therefore less likely to be referred to hospital.
Influence of age
Several generalisations are possible with respect to age. Evidence of specific aetiology tends to be more commonly found in older children14 [II]; viral infections (especially RSV) are more commonly found in younger children3, 6, 8, 10, 14 [II]; and Chlamydia and Mycoplasma species are more commonly found in older children6, 10, 14–16 [II]. For example, Harris et al16 found that patients over 5 years of age had a higher rate of M pneumoniae (42%) and C pneumoniae (20%) infections than those aged less than 5 years (15% and 9%, respectively) [II]. However, Block et al17 found the incidence of M pneumoniae and C pneumoniae infections to be comparable in all age groups between 3 and 12 years of age. In particular, the finding of a 23% incidence of M pneumoniae infection and 23% of C pneumoniae infection in children aged 3–4 years is higher than others [II] and raises questions about appropriate treatment in this age group. In most studies the incidence of S pneumoniae is less influenced by age3, 8, 10 [II].
Key points
| 4 Clinical features |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
Respiratory rate is difficult to count in healthy restless children. In children with moderate to severe pneumonia it is probably easier because they are more sick and quieter. The respiratory rate appears to be helpful in determining severity in infants under 1 year where a rate of >70 breaths/min had a sensitivity of 63% and specificity of 89% for hypoxaemia20 [II]. Between 12 and 36 months of age respiratory rates of >40 breaths/min were related to pneumonia, but in children aged >36 months tachypnoea and chest recession were not sensitive signs. Children can have pneumonia with respiratory rates of <40 breaths/min21 [II], and crackles and bronchial breathing were reported to have a sensitivity of 75% and a specificity of 57%20 [II]. High fever in both infants and children has been considered an important sign in the community, both in developed and developing countries22, 23 [III] [II]. Because clinical symptoms together contributed more than signs, it has been suggested that the WHO guidelines for diagnosing pneumonia should include breathlessness—that is, difficulty in breathing—which was found to be more helpful than breath count24 [II]. If all clinical signs (respiratory rate, auscultation, and work of breathing) are negative, the chest radiographic findings are unlikely to be positive.
A Medline search from 1982 to 1995 of studies which considered observer agreement of clinical examination suggested that observed signs (kappa 0.48–0.6) were better than auscultation signs (kappa 0.3)25 [Ia]. Wheezing is not a useful sign for determining severity in infants and young children18 [II]. Wheeze occurs in 30% of mycoplasma pneumonias and is more common in older children26 [IVb]. Because of this, the clinical diagnosis of mycoplasma pneumonia without radiography can be confused with asthma. Symptoms in older children may include abdominal pain (reflecting referred pain from the diaphragmatic pleura) and chest pain. The signs of bronchial breathing and pleural effusion are not present at the onset of symptoms.
Serious consideration should therefore be given to bacterial infection when the presentation is a fever of >38.5°C, recession, and tachypnoea. If wheeze is present, a primary bacterial pneumonia is very unlikely. If present, a viral or mycoplasmal infection should be considered or an underlying condition such as cystic fibrosis. Children with tuberculous pneumonia are severely ill and the radiographic appearances are suggestive. The features of bacterial, viral, and mycoplasma lower respiratory tract infection are shown in boxes 1, 2 and 3, respectively.
Box 1 Features of bacterial lower respiratory tract infection (LRTI)
|
Box 2 Features of viral lower respiratory tract infection (LRTI)
|
Box 3 Features of mycoplasma lower respiratory tract infection (LRTI)
|
Key points
CLASSICAL CLINICAL FEATURES
Pneumococcal pneumonia
Pneumococcal pneumonia starts with fever and tachypnoea. Since alveoli are poorly endowed with cough receptors, cough only occurs when lysis is present and debris is swept into the airways where cough receptors are plentiful. This accords with the many studies which emphasise the history of fever and breathlessness together with signs of tachypnoea, indrawing and unwell appearance ("toxaemia", "looks sick")18, 21, 24–27 [II]. This illness should therefore be considered in febrile tachypnoeic infants.
Staphylococcal pneumonia
Staphylococcal pneumonia is now rare in developed countries and, at the onset, is indistinguishable from pneumococcal pneumonia23 [IVb]. It is almost exclusively a disease of infants but can complicate influenza in older children.
Mycoplasma disease
Fever, arthralgia, headache, cough and crackles in a schoolchild would suggest mycoplasma infection26 [IVb], but again this can resemble pneumococcal and staphylococcal pneumonias as well as adenoviral illness if wheezing is prominent.
Others
Chlamydia trachomatis pneumonia is apparently not a fatal illness. The "staccato" cough is not specific, and crackles are described more frequently than wheeze. The only really significant clinical feature is a history of sticky eye in 50% of cases in the neonatal period.
It is unclear whether pertussis pneumonia is a primary pneumonia or is the result of aspiration28 [IVa]. It may co-exist with other pneumonias29 [III].
| 5 Radiological, general and microbiological investigations |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
39°C and white blood cell count of 20 000/mm3 or greater without an alternative major source of infection and with no additional clinical signs of pneumonia, radiographic signs of pneumonia were detected in about 25%30 [II]. This suggests that a chest radiograph should be undertaken in young children with a pyrexia of unknown origin. In a different study Heulitt et al31 reported a sensitivity and specificity for detecting radiographic pneumonia of 45% and 92%, respectively, for the presence of fever and tachypnoea in infants under 3 months. Only 6% of febrile infants had an abnormal chest radiograph in the absence of respiratory signs. The authors recommend that a chest radiograph should be obtained in febrile infants only when signs of respiratory distress are present31 [III]. However, the radiological features of segmental consolidation are not always easy to distinguish from those of segmental collapse, apparent in about 25% of children with bronchiolitis32, 33 [II] [II]. Taking these studies together, it would seem advisable to consider a chest radiograph in a child aged <5 years with a fever of >39°C of unknown origin unless classical features of bronchiolitis are present. One of the largest studies of the value of chest radiography was undertaken in children aged between 2 months and 5 years managed as outpatients with time to recovery as the main outcome34 [Ib]. Chest radiography did not affect the clinical outcome in these children with acute lower respiratory infection. This lack of effect was independent of clinicians' experience. There are no clinically identifiable subgroups of children within the WHO case definition of pneumonia who are likely to benefit from a chest radiograph. The authors concluded that routine use of chest radiography was not beneficial in ambulatory children aged over 2 months with acute LRTI. Antibiotic prescription was more frequent in those who underwent radiography (61% v 53%). This was the only trial identified in a recent Cochrane review.35
In a review of the value of chest radiography in infants with a clinical diagnosis of acute bronchiolitis, treatment was not altered by the radiographic signs (patchy collapse was evident in 25%). The authors concluded that the request for a chest radiograph in acute bronchiolitis should be made only when the need for intubation is being considered, where there has been an unexpected deterioration in the child's condition, or the child has an underlying cardiac or pulmonary disorder32 [II]. Lobar or segmental consolidation was observed more often in children aged <6 months infected with respiratory syncytial virus (RSV) than in older children with RSV infection36 [Ib].
Observer agreement on radiographic signs of pneumonia
Clinicians basing the diagnosis of lower respiratory infections in young infants on radiographic diagnosis should be aware that there is variation in intraobserver and interobserver agreement among radiologists on the radiographic features used for diagnosis. There is also variation in how specific radiological features are used in interpreting the radiograph. However, the cardinal finding of consolidation for the diagnosis of pneumonia appears to be highly reliable37 [II] and reasonably specific for bacterial pneumonia (74% of 27 patients with alveolar shadowing had bacterial proven pneumonia)38 [III]. Lateral radiographs are unhelpful39 [II].
Agreement between radiographic findings and clinical diagnosis
The evidence for having pneumonia without an abnormal chest radiograph is anecdotal. Anecdote suggests that fever and tachypnoea are present before classical consolidation is present. There is some evidence for an abnormal chest radiographic appearance without either fever or tachypnoea40 [III]. How this is managed depends on whether the patient is thought to have a developing illness or one which is resolving, as doubtless happened in the days before antibiotics.
Which pathogen?
Chest radiography is too insensitive to be useful in differentiating between patients with bacterial pneumonia and those whose pneumonia is non-bacterial41, 42 [II]. There is no radiological pattern pathognomonic for mycoplasmal pneumonia. Interstitial infiltrates, lobar consolidation, and hilar adenopathy have all been described. Pleural effusions are rare43 [III].
Because of poor observer agreement and appreciable false negative errors when viral and bacterial readings were compared with titre increases and positive bacterial cultures, respectively, radiographic findings are poor indicators of an aetiological diagnosis in ambulatory children with pneumonia and, of themselves, are an insufficient database for making therapeutic decisions44 [II]. There is no evidence that an experienced clinician would be more confident diagnosing non-bacterial pneumonia from the clinical features, age of the child, and the chest radiograph.
Follow up radiographs
Children with lobar collapse should probably all be followed up and reviewed with a radiograph. A follow up radiograph is also sensible for children with apparent round pneumonia to ensure tumour masses are not missed. Follow up radiographs after acute uncomplicated pneumonia are of no value where the patient is asymptomatic45, 46 [II] [III].
Key points
GENERAL INVESTIGATIONS
Community
There is no indication for any tests in a child with suspected pneumonia in the community.
In hospital
Pulse oximetry
Oxygen saturation (SaO2) measurements provide a non-invasive estimate of arterial oxygenation. The oximeter is easy to use and requires no calibration. However, it requires a pulsatile signal from the patient. When using paediatric wrap around probes, the emitting and receiving diodes need to be carefully opposed. It is also highly subject to motion artefacts. To obtain a reliable reading, (1) the child should be still and quiet; (2) a good pulse signal (plethysmograph) should be obtained; and (3) once a signal is obtained, the saturation reading should be watched over at least 30 seconds and a value recorded once an adequate stable trace is obtained.
In a prospective study from Zambia the risk of death from pneumonia was significantly increased when hypoxaemia was present20 [Ib].
Key point
Acute phase reactants
White cell count, total neutrophil count, C reactive protein (CRP), and erythrocyte sedimentation rate (ESR) are generally performed in the belief that they help distinguish bacterial from viral infections and the clinician will therefore find them helpful in deciding whether or not to prescribe antibiotics.
Recent prospective studies have examined the usefulness of acute reactants in distinguishing bacterial from viral pneumonia.47, 48 Nohynek et al48 studied 121 children admitted to hospital with acute lower respiratory infection. Using culture and serological techniques, they divided the children into four groups: those with a bacterial infection (n=30), those with a viral infection (n=30), those with mixed infections (n=24), and those of unknown aetiology (n=37). The distribution of ESR, full blood count, and CRP values was wide within each group and they could not identify cut off points that would reliably distinguish bacterial from viral infections or bacterial and mixed infections from viral infections [Ib]. Korppi et al,47 in a similar prospective study examined whether pneumococcal infection (n=29), could be distinguished from pneumococcal and viral infection (n=17) or viral infection alone (n=23). There was a statistically significant difference in the CRP level, ESR, and absolute neutrophil count between pneumococcal infection alone and viral infection alone, but specificity and sensitivity remained poor (white cell count >15 000: sensitivity 33%, specificity 60%; neutrophil count >10 000: sensitivity 28%, specificity 63%; CRP >60 mg/l: sensitivity 26%, specificity 83% for S pneumoniae versus a viral infection) [Ib].
Host response indices are best at detecting invasive infections. In children many acute lower respiratory infections may be less invasive and mucosal limited and cause less host response. Some viral agents, particularly adenovirus or influenza virus, are capable of causing invasive infection and hence may induce a host response very similar to that seen in invasive bacterial infections. Acute phase reactants do not therefore usually distinguish between bacterial and viral infection in children [Ib].
Acute phase reactants can also be measured as a baseline and may then only be useful if the patient does not improve on treatment as expected.
Key point
Urea and electrolytes
Investigation of urea and electrolytes to assess electrolyte imbalance is undertaken if the patient is severely ill or shows evidence of dehydration. Inappropriate secretion of antidiuretic hormone (ADH) is recognised in both children and adults with pneumonia. In adults it has been shown that there is a latent vasopressin dependent impairment of renal water excretion in acute pneumonia.49 A study of 264 children admitted to hospital with pneumonia in India showed hyponatraemia at admission in 27% and it was calculated that, in 68% of these children, the hyponatraemia was secondary to inappropriate ADH secretion.50 Treatment is with fluid restriction.
SPECIFIC MICROBIOLOGICAL INVESTIGATIONS
It can be difficult to determine the responsible pathogen in children with acute LRTI. The gold standard would be to take samples directly from the infected area of the lung, and the yield from bacterial growth from lung punctures in African children is high (79%).51 Most often in western countries, however, less invasive sampling measures are used for diagnosis.
Community
There is no indication for microbiological investigation of the child with pneumonia in the community.
In hospital
For patients admitted to hospital with pneumonia, it is important to attempt a microbiological diagnosis. It is clear from a number of studies that 10–30% of infections will have a mixed viral and bacterial aetiology8, 23 [II].
Bacterial pneumonia
Mycoplasma pneumonia
Complement fixation tests: a rise in paired titre is regarded as the gold standard for the diagnosis of M pneumoniae. IgM ELISA has been shown to reach a diagnostic level during the second week of the disease.57 Cold agglutinins are often used as an acute test but their value is limited. In children aged 5–14 years the positive predictive value for mycoplasma of a rapid cold agglutinin test was 70%.58
Viral pneumonia
Viral antigen detection in nasopharyngeal aspirates is highly specific for respiratory syncytial virus, parainfluenza virus, influenza virus, and adenovirus. Sensitivities of this test approach 80%, particularly in infants.55 Nasopharyngeal aspirates should also be cultured for viruses, although infections are also diagnosed by rises in titre in paired serum samples [II]. Nasal lavage can be substituted for nasopharyngeal aspirates.59
The results of viral detection tests are particularly useful for cohorting infected children during outbreaks and for epidemiological purposes.
A summary of the diagnostic value of specific microbiological investigations is shown in table 3.
|
| 6 Severity assessment |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
).
|
INDICATIONS FOR ADMISSION TO HOSPITAL
A key indication for admission to hospital is hypoxaemia. In a study carried out in the developing world, children with low oxygen saturations were shown to be at greater risk of death than adequately oxygenated children.20 The same study showed that a respiratory rate of 70 breaths/min or more in infants aged <1 year was a significant predictor of hypoxaemia. Oxygen saturation levels (SaO2) below 92%, other clinical signs of severe disease, and a family incapable of appropriate observation and supervision of the child are all indicators for admission to hospital.
Indicators for admission to hospital in infants:
92%, cyanosis; Indicators for admission to hospital in older children:
92%, cyanosis;
INDICATIONS FOR TRANSFER TO INTENSIVE CARE
Transfer to intensive care should be considered when:
| 7 General management (other than antibiotics) |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
GENERAL MANAGEMENT IN HOSPITAL
Oxygen therapy
Hypoxic infants and children may not appear cyanosed. Agitation may be an indication of hypoxia.
Patients whose oxygen saturation is less than 92% while breathing air should be treated with oxygen given by nasal cannulae, head box, or face mask to maintain oxygen saturation above 92%. There is no strong evidence to indicate that any one of these methods is more effective than any other. A study comparing the different methods in children under 5 years of age concluded that the head box and nasal cannulae are equally effective,60 but the numbers studied were small and definitive recommendations cannot be drawn from this study. It is easier to feed with nasal cannulae, but the maximum flow rate of oxygen recommended by the manufacturer by this method is 2 l/min. Alternative methods of delivering higher concentrations of humidified oxygen such as face mask or head box may be necessary.
Where the child's nose is blocked with secretions, gentle suctioning of the nostrils may help [D]. No studies assessing the effectiveness of nasopharyngeal suction were identified.
Key points
Fluid therapy
Children who are unable to maintain their fluid intake due to breathlessness or fatigue need fluid therapy. Studies on preterm infants or infants weighing <2000 g have shown that the presence of a nasogastric tube compromises respiratory status61, 62 [II]. Older children may be similarly affected, although potentially to a lesser extent because of their larger nasal passages, so although tube feeds offer nutritional benefits over intravenous fluids, they should be avoided in severely ill children. Where nasogastric tube feeds are used, the smallest tube should be passed down the smallest nostril.63 There is no evidence that nasogastric feeds given continuously are any better tolerated than bolus feeds (no studies were identified); however, in theory, smaller more frequent feeds are less likely to cause stress to the respiratory system.
Patients who are vomiting or who are severely ill may require intravenous fluids. These should be given at 80% of basal levels (once hypovolaemia has been corrected) and serum electrolytes should be monitored in the severely ill as inappropriate ADH secretion is a recognised complication.50, 64
Key points
Physiotherapy
Two randomised controlled trials65, 66 and an observational study67 conducted on adults and children showed that physiotherapy did not have any effect on the length of hospital stay, pyrexia, or chest radiographic findings in patients with pneumonia. There is no evidence to support the use of physiotherapy including postural drainage, percussion of the chest, or deep breathing exercises66 [Ib], 65, 67 [III]. There is a suggestion that physiotherapy is counterproductive, with patients who receive chest physiotherapy being at risk of having a longer duration of fever than the control group.65 In addition, there is no evidence to show that physiotherapy is beneficial in the resolving stage of pneumonia.68
A supported sitting position may help to expand lungs and improve respiratory symptoms in children with respiratory distress.
Key point
Management of fever and pain
Children with acute LRTI are generally pyrexial and may have some pain, including headache, chest pain, arthralgia (in the case of mycoplasma pneumonia), referred abdominal pain, and possibly earache from associated otitis media (see section on Clinical features). Pleural pain may interfere with depth of breathing and may impair the ability to cough. Antipyretics and analgesics can be used to keep the child comfortable and to help coughing. Minimal handling helps to reduce metabolic and oxygen requirements and this should be considered when planning and carrying out procedures, investigations, and treatments.
Key points
Monitoring
The frequency of monitoring—including heart rate, temperature, respiratory rate, oxygen saturation level, respiratory pattern including chest recession and use of accessory muscles—is determined by the child's condition. The sicker the child, the more likely that continuous oxygen saturation monitoring will be needed. Patients on oxygen therapy should have at least 4 hourly observations including oxygen saturation.
Key point
| 8 Antibiotic management |
|---|
|
TOP1 Introduction2 Incidence and mortality3 Aetiology and epidemiology4 Clinical features5 Radiological, general and...6 Severity assessment7 General management (other...8 Antibiotic management9 Complications and failure...10 Prevention and primary...Appendix 1: Search strategyAppendix 2: Community acquired...References |
|---|
There is a clear dearth of large pragmatic randomised controlled trials to provide the evidence necessary to make these decisions. Many of the studies identified by the searches were designed to support the licensing of new treatments which were compared with a "standard" antibiotic regimen. Although designed as randomised controlled trials, they frequently involved small numbers of patients and both the "new antibiotic" group and the "standard therapy" comparison group had a very high rate of complete recovery from pneumonia. These studies did not appear to be powered to show a difference in efficacy between the two regimes and certainly were not powered to demonstrate equivalence. Likewise, it was not possible to assess the differential safety of the various regimens which have been compared in these trials. The lack of evidence for treatment with antibiotics which has been identified in this review highlights the need for well designed randomised controlled trials to address the key questions concerning the management of CAP in children.
WHETHER TO TREAT WITH ANTIBIOTICS
One of the major problems in deciding whether to treat a child with CAP with antibiotics is the difficulty in distinguishing bacterial pneumonia (which would benefit from antibiotics) from non-bacterial pneumonia (which would not). This difficulty has been described in Section 3 on Aetiology. Resistance to antibiotics among bacterial pathogens is increasing and is of concern; an important factor in this increase is the overuse of antibiotics. Only one study was identified in which children with diagnosed pneumonia treated with antibiotics were compared with a group not treated with antibiotics69 [II]. This study was a randomised controlled trial of 136 young Danish children aged 1 month to 6 years. The diagnosis of pneumonia was based on the ausculatory findings of fine crepitating rales or radiographic appearances of pulmonary consolidation. Over half the children were diagnosed as having viral pneumonia on the basis of laboratory evidence, caused by RSV in most cases. The children had relatively mild signs and symptoms and those with severe breathing difficulty, cyanosis, suspected septicaemia, and pre-existing pulmonary or cardiac disease were excluded. The treatment group received either ampicillin or penicillin V, depending on their age. There were no differences in the course of the illness between the two groups but 15 of the 64 in the placebo group did eventually receive antibiotics. There are concerns about the generalisability of this study to a UK setting of children admitted to hospital with CAP as, in the UK, in about half the patients entered into the study bronchiolitis would have been the clinical diagnosis.
There is increasing concern about the inappropriate use of oral antibiotics and a recent commentary70 [IV] suggested that, by educating parents about the drawbacks of oral antibiotics, they may be empowered to question their doctors about antibiotic use.
Key point
CHOICE OF ANTIBIOTIC
It is clear that there is variation in medical prescribing which largely reflects custom and local practice. We have reviewed the relevant scientific evidence and provide recommendations based, where possible, on that evidence, but more frequently recommendations are based on judgements about what constitutes safe and effective treatment. In pneumonia in children the nature of the infecting organism is almost never known at the initiation of treatment and the choice of antibiotic is therefore determined by the reported prevalence of different pathogens at different ages and the associations between specific pathogens and certain clinical features.
Macrolides compared with other groups of antibiotics
In adults macrolide antibiotics have been shown to reduce the length and severity of pneumonia caused by Mycoplasma pneumoniae compared with penicillin or no antibiotic treatment.71 There are no similar studies in children. This class of antibiotics is also effective against a wide range of bacterial pathogens and thus has a number of advantages. Three randomised controlled trials which compared macrolides with other groups of antibiotics were identified10, 16, 72 [Ib]. Harris et al16 found no difference between treatment with a macrolide antibiotic and a penicillin based antibiotic [Ib]. Langtry and Balfour73 showed that azithromycin was slightly more efficacious than ceftibuten, but this was a small study and a review of azithromycin treatment of paediatric pneumonia showed that azithromycin was similar to all comparator antibiotics (co-amoxiclav, cefaclor, erythromycin or josamycin) [II]. There were no clinical significant differences in efficacy between azithromycin and these comparator agents at the end of treatment.
Macrolides compared with other macrolide antibiotics
Manfredi et al74 compared azithromycin for 3–5 days with erythromycin for 7 days and found no difference in efficacy [Ib], Block et al17 compared erythromycin with clarithromycin, each given for 10 days, and also found no differences in efficacy [II], and Ficnar et al75 compared a 3 day course of azithromycin with a 5 day course and showed no difference between the two treatment groups [II].
Cephalosporins compared with non-cephalosporins
Two randomised controlled trials were identified in this group76, 77 [II]. In a study by Amir et al76 parenteral treatment with ceftriaxone for 2 days was followed by either 8 days of treatment with co-amoxiclav or 8 days of treatment with cefixime. There were no differences between the groups. Klein77 compared cefpodoxime proxetil with amoxicillin clavulanate and also found no difference between the treatment groups.
Comparison of two cephalosporins
Three studies78–80 compared two cephalosporins, two of which were randomised controlled trials78, 80 [II]. No differences were detected between the groups in any of the three studies. In all the studies a third generation cephalosporin was evaluated against cefaclor.
Antibiotic resistance
Antibiotic resistance among pneumococci is increasing and is of concern because pneumococcus is an important cause of severe CAP in children and because penicillin and macrolide resistance are increasingly linked. Two studies addressed the response to antibiotic treatment of children with pneumonia caused by penicillin resistant S pneumoniae. The United States Pediatric Multicenter Pneumococcal Surveillance Study Group (consisting of eight children's hospitals) prospectively identified children with S pneumoniae; 257 episodes were included. 8% of isolates were intermediate and 6% were resistant to penicillin, and 3% were intermediate and 2% were resistant to cefotaxime. There was no difference in outcome between susceptible and resistant cases. Note that a high proportion received parenteral antibiotics: 80% of outpatients had an intravenous dose of a cephalosporin followed by a course of oral antibiotics and 17% received an oral ß-lactam course alone. 48% of inpatients had an intravenous course of a cephalosporin followed by a course of oral antibiotic, 20% had intravenous cephalosporin with intravenous penicillin, and 16% had intravenous cephalosporin and intravenous vancomycin. Of those with penicillin resistant organisms, all but one had at least one dose of an intravenous antibiotic81 [III].
In another study Friedland82 compared factors in 78 South African
