0
Reviews |

The Value and Complications of Percutaneous Transthoracic Lung Aspiration for the Etiologic Diagnosis of Community-Acquired Pneumonia* FREE TO VIEW

J. Anthony G. Scott, BM, BCh; Andrew J. Hall, BM, BS
Author and Funding Information

Affiliations: *From the Infectious Disease Epidemiology Unit (Dr. Hall), London School of Hygiene and Tropical Medicine, London, UK; and Wellcome Trust/Kenya Medical Research Institute (Dr. Scott), Centre for Geographic Medicine Research–Coast, Kilifi, Kenya.,  Currently at Center for Disease Control and Prevention, Atlanta, GA.

Correspondence to: J. Anthony G. Scott, BM, BCh, Wellcome Trust/Kenya Medical Research Institute, Centre for Geographic Medicine Research-Coast, PO Box 230, Kilifi, Kenya; e-mail: Ascott@kilifi.mimcom.net



Chest. 1999;116(6):1716-1732. doi:10.1378/chest.116.6.1716
Text Size: A A A
Published online

Diagnosis of pneumonia is frustratingly incomplete, with most large studies failing to identify a causative organism in 33 to 45% of patients.13 Diagnosis by blood culture is highly specific, but the sensitivity of blood cultures for bacterial pneumonia is < 25%. An etiology can be obtained in undiagnosed cases by culture of sputum or from serologic diagnoses, but these techniques lack specificity or produce an unconvincing description with multiple pathogens.46 The most convincing evidence of etiology in observational studies of acute pneumonia is the demonstration of a single pathogen in the diseased tissue of the patient and the absence of this same organism from equivalent tissue in healthy control subjects. Percutaneous transthoracic aspiration of “lung juice” is the most acceptable and uncontaminated means of describing the microbiology of pneumonic lung antemortem, but the technique is not widely practiced because of the potential for adverse events. Here, we review the existing data on both the etiologic yield and the frequency of complications in order to evaluate the role of lung puncture in modern diagnostic studies.

Pertinent studies were identified by a search of the National Library of Medicine MEDLINE database examining all fields for the phrases “lung aspiration,” “lung puncture,” “lung suction,” and “lung tap,” and extending retrospectively through secondary, tertiary, and quaternary citations. At the outset, a distinction was drawn between lung aspiration, in which a small, fine-gauge needle is used to diagnose pneumonia, and needle biopsy, in which a longer, thicker needle is used to diagnose malignant or inflammatory focal lung lesions under fluoroscopic or CT guidance.7

After sterilizing the skin, a 20- or 25-gauge needle is attached to a 10-mL syringe and inserted immediately above a rib to the depth of the parietal pleura. While the patient suspends respiration, the needle is pushed rapidly 2 to 3 cm into the consolidated lung, and suction is applied as the needle is steadily withdrawn. The procedure is so quick that it is unnecessary to anesthetize the skin or pleura.8 Nonetheless, local anesthetic is used by some authorities to prevent rapid movements of the chest wall in response to pleuritic pain,9and premedication with atropine is practiced by others.1011

The aspirate, also known as lung juice, lung fluid, or pulmonary exudate, usually has a volume of < 0.5 mL and is expelled by in-and-out movements of the syringe plunger after the needle tip is immersed in broth. A common variant of the technique is to prime the syringe with 0.5 mL of sterile isotonic saline solution and inject this into consolidated lung to increase the volume of the aspirate, although this has not been proven to increase diagnostic yield.9 Because preservatives used in clinical supplies of saline solution or buffer may inhibit the growth of bacteria, the results may be improved by using an alternative lavage such as sterile culture broth.12 Occasionally, the patient will have a small laminar pleural effusion, and a larger volume of aspirate will be obtained as the tip of the needle is withdrawn through the pleural space.

Most commonly, a 21-gauge, 35-mm venipuncture needle is used,13but needles as short as 20 mm14or as long as 150 mm15have been used on children. Some investigators insert a guide needle to the pleura,1619 while others use a 22-gauge lumbar puncture needle with a stylet.20 Normally, only one pleural puncture is attempted, but several investigators have repeatedly withdrawn and replunged the needle inside the lung parenchyma to increase the volume of lung juice to 1 to 3 mL,2021 and others have conducted prolonged suctions over ≥ 30 s.1011

Pulmonary consolidation is localized by radiographs in two planes, and is refined by clinical examination, selecting the area of most pronounced bronchial breathing.22 Some authors use the same “safe” locations for disease in any given lobe, eg, the seventh intercostal space in the midscapular line for a lower lobe puncture.,9,23Safe surface markings have been described for the whole thorax in children by radiology of punctured cadavers.24With longer needles (> 4 cm), there is less certainty of the course of the needle in the thorax, and material from liver or spleen has occasionally been extracted.25

The first accounts of transthoracic percutaneous lung aspiration were published in 1883 in Germany, by Leyden26and Günther,27and in France, by Talamon.28 Leyden found the same organisms in lung aspirate and blood of one patient with pneumonia but found no organisms in aspirates from two other patients. The technique was popularized in England in 1909 by Horder,29 who diagnosed five out of six cases of pneumonia or lung abscess by aspiration,2930 and in the United States in the 1920s by Thomas and Parker22 and Lyon,31 who together published > 100 cases.

Lung aspirate studies played an important part in elucidating the pathogenesis of pneumococcal pneumonia. Tchistovitch32 demonstrated that lung juice obtained from patients after crisis contained pneumococci virulent for mice and rabbits; crisis was, therefore, not dependent on the clearance of pneumococci from the lungs. From a series of 48 lung punctures at different stages of pneumonia, Rosenow13 deduced that the probability of finding pneumococci decreased with time in those who survived but increased with time in those who died. Lister33 used pneumococci derived from lung aspirates to demonstrate the agglutinating power of convalescent serum from the same patient, paving the way to a strategy of vaccination with capsular antigens.

Without antibiotics or serum therapy, early lung aspirate studies were only of academic interest. Nonetheless, episodes of pneumonia were“ treated” with lung puncture, with “good results”34and, in the days before controlled clinical trials, several authors noted that lung puncture itself hastened the resolution of pneumonia.3537 With the advent of serum therapy, it became apparent that specific antiserum was most effective when given early in the course of pneumonia.38 Serotyping the infecting pneumococcus became a medical emergency,9 and lung aspirates were used to ensure the timely and economic use of the available sera.36,3940 During the 1930s, lung aspiration became a commonplace medical procedure and was reported in a wide number of scientific publications from Europe,3637,39,4143 the United States,9,15,40,4446 and Africa.24,4748

In the discovery of etiologic agents, lung aspirate studies have a scientific authority that has shaped our current understanding of pneumonia etiology. Oval diplococci, presumably pneumococci, were identified in lung juice separately in 1882 by Leyden26 and Günther27 in Germany. In France, Netter49established that Hemophilus influenzae was responsible for many bronchopneumonias in children, using lung aspirates during the influenza epidemic of 1918. Staphylococcus aureus was defined as a pneumonia pathogen by lung aspirate studies performed shortly after death,50 and Neisseria meningitidis by both postmortem51 and antemortem lung aspirates.52Mycobacterium tuberculosis,52respiratory syncytial virus,53Pneumocystis carinii,54 and Nocardia55have all been defined as pneumonia pathogens by antemortem lung aspirate studies. Idiopathic pulmonary hemosiderosis of children, previously only found at postmortem, was also first described in life by lung puncture.56

Postmortem lung aspiration was thought to provide uncontaminated lung material without introducing the risk of adverse events during life and was practiced widely in the early part of the 20th century.28,5760 Axton and Giles61used this technique to investigate the etiology of pneumonia in malnourished Rhodesian children. Specimens were taken within 10 min of death and were dominated by Gram-negative cultures. Overgrowth of the lungs by Gram-negative organisms is a potential hazard of postmortem material, but subsequent antemortem lung aspirate studies of malnourished African children have revealed a similar shift in the spectrum of pathogens.6263 The value of postmortem cultures has diminished in the antibiotic era, where the causative organism may be obliterated by treatment, leaving only postmortem opportunistic organisms,64but specimens taken after death have been successfully used recently to implicate a viral etiology in children with acute respiratory infection.65 The practical difficulty is in obtaining consent in a tactful and timely manner.

After the introduction of antibiotics, interest in lung aspiration waned. However, two factors led to its revival in recent decades: the development of antibiotic resistance and the recognition of the wide spectrum of potential respiratory pathogens. The additional diagnostic yield of lung aspiration has been used to enhance sensitivity in etiologic studies of pneumonia and to obtain live organisms for susceptibility testing.6667 It has also evolved a role in the evaluation of individual patients with unresponsive or complex pneumonias,16,19,25,6869 and in the detection of disease-specific end points in vaccine efficacy trials for adults70and children.71

The advantage of lung aspirate samples is that neither the instrument nor the specimen has to pass through the upper respiratory tract, trachea, or bronchi, all of which may be colonized by potential respiratory pathogens. For example, while a single pathogen is normally obtained from lung aspirate cultures, sputum specimens from the same patient frequently culture a mixture of organisms with an excess of Gram-negative bacilli.5,72In healthy individuals, bacteria are detected with decreasing frequency and concentration as the respiratory tract is descended.7374 Using a transtracheal approach to avoid upper respiratory tract contamination, Kalinske et al75were unable to culture bacteria from the tracheas of 13 healthy volunteers; Berman et al76 found only scanty cultures in 4 of 15 tracheal aspirates from 12 normal volunteers.

Lung aspirates in 25 healthy anesthetized dogs yielded positive bacterial cultures in 3 animals, despite shaving and sterilizing the skin, giving a specificity estimate of 0.88.77In humans, lung tissue excised at pulmonary lobectomy has been shown to be sterile, unless the indication for surgery was an abscess.78Aspirates of the lung in normal healthy control subjects support this conclusion. Bacteria were cultured from none of 13 lung aspirates from normal healthy Chilean infants, but from 228 of 505 aspirates (45%) from infants with pneumonia.79 In 10 healthy South African adults subjected to lung puncture, respiratory pathogens were absent from all cultures, although 4 yielded scanty“ contaminants.”25

Patients with other diseases provide additional evidence. Of 16 aspirates from Spanish adults with suspected pneumonia who later received diagnoses of pulmonary embolism, pulmonary edema, or malignancy, only 1 specimen cultured a microorganism, Candida parapsilosis, a likely opportunistic pathogen in an immunocompromised patient.80Cultures of 10 lung aspirates from children with suspected tuberculosis (TB) were negative for conventional respiratory pathogens.81 Taken together, the evidence suggests that the normal lung rarely contains sufficient organisms to produce a significant culture in lung aspirates, and the specificity of the technique in pneumonia is likely to be very high.

A false-positive result implies culture of an organism in a lung aspirate that is not responsible for the episode of pneumonia being investigated. There are three obvious sources of false-positive results: (1) contamination of the aspirate cultures; (2) detection of infection confined to the blood stream; and (3) culture of bacteria that are not primarily responsible for the pneumonia. In lung aspirates, the frequency of skin or laboratory contamination should be similar to that of blood cultures, which is normally 1 to 3%. In the context of a radiographic pneumonia, differentiating whether an aspirate culture is derived from lung tissue or merely from the blood of a punctured vessel9 may seem an unnecessary distinction. However, several studies have documented different organisms in blood and lung simultaneously, and bacteremia may be a secondary consequence of severe illness rather than the primary cause of the pneumonia.23,31,66,82

Secondary infection may also be established in the lung tissue itself, eg, S aureus pneumonia may be superimposed on an episode of influenza.8384 The viral disease may compromise the mechanical defenses of the larynx and the clearing action of the mucociliary escalator, allowing inhalation, stasis, and growth of organisms normally confined to the pharynx. In the same manner, one bacterial infection may facilitate infection with another, and the finding of dual infection in lung aspirates may occur in as many as 25 to 40% of children with pneumonia.24,66 Case fatality rates in patients with dual infections are higher.12 Although two pathogens may lead to a worse prognosis than one, it is more likely that dual infection is simply a marker of advanced disease; as the patient approaches death, more colonizing bacteria may gain a foothold in the normally sterile lung tissue. Secondary infecting pathogens are not “false-positives” because they all contribute to the condition of the patient; however, if the objective is to define the pathogen initiating the pneumonia, as it would be in evaluating the potential impact of a pathogen-specific vaccine, then cultures of secondary organisms are misleading.

There is no superior “gold standard” by which the sensitivity of lung aspiration can be assessed. Therefore, the appropriate measure of the value of the procedure is its diagnostic yield, or the percentage of positive diagnoses. A summary of etiologic series using lung aspirates is presented for children, in Table 1 , and for adults, in Table 2 . The mean diagnostic yield for culture of a single respiratory pathogen is 41% in children and 47% in adults. For any culture, single or multiple, the mean yield is 48 to 49% in both groups. A summary yield must be interpreted with caution as there is significant variation in the estimates of individual studies within both children and adults (χ2, p < 0.0005), which probably reflects the different age structures and geographic locations of the populations studied and the differential access to antibiotics.

Sensitivity of Lung Aspirate vs Blood Culture

A true estimate of the culture sensitivity of lung aspirates can be obtained for bacteremic pneumonia by comparing lung aspirate results with those of simultaneous blood cultures. Equally, the sensitivity of blood cultures in pneumonia can be estimated from the results of lung aspirate cultures. Thirteen studies provide the data to make these comparisons (Table 3 ). The combined estimate of sensitivity for lung aspiration is 0.74, and for blood cultures, 0.37. Although there is wide variation among the estimates of blood culture sensitivity, from 0.05 to 0.67, there is much better agreement on the estimates of lung aspirate sensitivity. In six studies, it is possible to calculate the additional yield of lung aspirates over and above that of blood cultures.5,23,80,82,100,102 Of 479 patients sampled, blood cultures identified an etiology in 82 patients, lung aspirates identified an etiology in 199 patients, and the combination of the two techniques identified an etiology in 229 patients. The addition of lung aspirates to blood cultures as an investigative tool increased the diagnostic yield from 17 to 48%.

Sensitivity to Different Pathogens

Some organisms that are common causes of pneumonia are only rarely bacteremic; a typical example is H influenzae.110 Among Kenyan adults with pneumonia, H influenzae was obtained from 10 of 259 lung aspirates but from only 1 of 518 blood cultures.,82 Among 170 adults from PNG with pneumonia, 15 diagnoses of H influenzae were made; 4 patients had positive blood cultures and 14 patients had positive lung aspirate cultures.,111 In both studies, failure to perform lung aspirates would have significantly underestimated the contribution of H influenzae.

Lung aspirate studies may also reveal additional or discordant isolates when compared with results of blood cultures.5,23,31,66,82 Among 100 children in the Gambia, different species of pathogen were isolated in blood and lung aspirate cultures from four patients; for a child with S pneumoniae in blood cultures and S aureus in the lung aspirate, the latter investigation may have provided life-saving information.,23

Sensitivity to Multiple Etiologies

When two organisms are implicated in pneumonia, the possibility of a false-positive test should be considered seriously. When both organisms are isolated from lung aspirate cultures, however, the evidence implicating dual etiology is relatively strong. H influenzae and S pneumoniae were first co-isolated in lung aspirates 80 years ago31,58 and have been found repeatedly throughout the century.12,21,2324,37,44,66,87,103 Dual etiology appears to be more common among children. For example, H influenzae and S pneumoniae were isolated in 18 of 83 lung aspirates (22%) in children from the Papua New Guinea (PNG) Highlands,66 but in only 2 of 90 aspirates (2%) from adults in coastal PNG.21 Other pairs of pathogens have been cultured together,31,79,100 and three and four organisms have been found in the same lung aspirate,24,66 although higher orders of co-isolation frequently include one or more probable skin contaminants. It is notable how frequently Branhamella catarrhalis appears as a co-culture in lung juice, yet how infrequently it is isolated alone,,24,66 suggesting that its role in pneumonia is rarely that of the initiating cause.

Dual infection can occur within one species. Two different serotypes of pneumococci were isolated from 7 of 1,255 aspirations performed by Bullowa45 on adults with pneumonia. Two serotypes have also been isolated from postmortem lung cultures in a United States adult and from an antemortem lung aspirate in a Gambian child.23,112 In the PNG study, two different populations of H influenzae were isolated from lung aspirates in eight children, with each culture comprising one encapsulated strain and one unencapsulated.,12,66 As the encapsulated strain was also more likely to be isolated from blood, the authors argue that lung aspirates may reveal a pattern of serotypes that is more representative of those causing pneumonia and is more relevant for the selection of antigens in polyvalent vaccines.66 Among adults with pneumonia in Kenya, serotypes 1, 4, and 14 accounted for 11 of 17 pneumococci (65%) found in lung aspirates alone, but for only 6 of 25 isolates (24%) found in blood alone (authors’ unpublished observations).

Diagnostic Yield in TB

TB was first diagnosed by lung tap in 1923,52 and mycobacteria have been isolated from lung juice in numerous subsequent aspirate studies.62,8182,94,101,103,113116 Schuster et al81 were the first to propose aspiration as a specific diagnostic technique for pulmonary TB and argued their case by reporting aspirates from 10 Chilean infants with an appropriate clinical syndrome; in eight, M tuberculosis was cultured on Lowenstein-Jensen medium, and in seven, acid-fast bacilli were seen on lung juice smears, providing a rapid diagnosis. From the eight infants who received TB diagnoses, culture of gastric washings proved negative in six. In a similar study from Udaipur, India, 16 of 30 infants with a compatible clinical illness had positive lung juice cultures, five of which were positive on lung aspirate smears.,113

From a series of 1,255 lung aspirates in Sweden, Dahlgren and Ekstrom114performed a retrospective audit of 212 cases in which a diagnosis of TB had been suspected. The patients were investigated using fine-needle aspiration under fluoroscopic control with Ziehl-Neelsen stains, Lowenstein-Jensen cultures, and guinea pig inoculation. Based on 5-year follow-up figures, TB was missed by lung aspirate in only 26 of 197 true-positive patients, and false-positive diagnoses were encountered in 15 patients. This gives a sensitivity of 87% and positive predictive value of 92%. The case definition in all these studies is subjective and ill-defined, so it is difficult to relate these very successful results to an unselected population of patients with chronic respiratory illness. Nonetheless, in patients with a strong clinical suspicion of TB and negative gastric lavage and sputum smears, lung aspiration may lead to a diagnosis, particularly if the needle is radiologically guided to the center of the diseased area.115 Furthermore, in etiologic studies of acute pneumonia, particularly in developing countries, M tuberculosis has been found either alone or in combination with other respiratory pathogens in 2 to 9% of lung aspirate cultures, indicating that all aspirate samples should be cultured for Mycobacteria.,62,82,101

Diagnostic Yield in Bronchopneumonia

The patchy nature of bronchopneumonia suggests that it might lend itself less readily to lung aspirate diagnosis.4243 Only three studies have directly contrasted the diagnostic yield in comparable populations with lobar pneumonia and bronchopneumonia. In Lyon’s study,31 the yield from 20 children with lobar pneumonia and 18 with bronchopneumonia was 50% in each group. Glynn and Digby52 studied 31 adults and 13 children and obtained positive cultures in 4 of 6 bronchopneumonia patients (67%) and in 18 of 38 lobar pneumonia patients (47%). In Kenyan adults, respiratory pathogens were cultured in 7 of 28 patients with bronchopneumonia (25%) and in 57 of 231 patients with lobar pneumonia (25% [authors’ unpublished observations]). In the latter two studies, there was an excess of S aureus in cultures from bronchopneumonia patients. The largest experience of bronchopneumonia comes from a population of 505 Chilean infants in whom the diagnostic yield was a very adequate 45%.,79 There is no evidence to support the idea that lung aspirate cultures are less sensitive in patients with bronchopneumonia.

One of the commonest causes of a negative result in a lung aspiration study is failure to perform the test! Physicians have an understandable reluctance to puncture the thorax, particularly in children,31,117 and patients have an understandable anxiety. Diak-paromre and Obi100 limited their punctures to the posterior thorax “in order to reduce anxiety.” Lyon31 wrote in 1922, “At best, it is a somewhat painful procedure, and often excites in the child patient a terror which for several days to come reawakens with every appearance of the physician.” Shortening the needle from 15 cm to 3 cm will reduce much of this anxiety, but there are other obstacles to puncture. In many African cultures, consent to invasive procedures can only be given by the father of a pediatric patient, but it is usually the mother who takes the child to the hospital. This may have been a factor in a study of 90 Gambian infants with acute respiratory infection, in which lung aspirates were obtained in only 2 patients.118 A second limitation is the timing of presentation. Lung puncture studies usually require radiographic evidence of consolidation, but patients presenting early in the disease may not yet have developed radiologic changes. A good example is given by Douglas and Devitt,106 who performed lung aspiration in a patient with clinical signs of bronchial breathing before the radiograph was formally reviewed. The aspirate grew S pneumoniae, and the radiograph was reported as normal.

Occlusion of the needle by a skin plug45 will yield a false-negative result, but this can be avoided by using a stylet or by injecting a small volume of sterile broth.119The consolidation may be missed if it is small or if the operator is inexperienced. The changes on the radiograph may represent malignancy, pulmonary edema, pulmonary eosinophilia, or other noninfectious etiology.120 Because the number of organisms cultured from patients with pneumococcal pneumonia increases as the disease progresses and falls as a patient recovers naturally, punctures performed too early or too late may fail to detect the infection.13,22,121

In seven studies with appropriate data, prior use of antibiotics reduced the diagnostic yield in all but one, with a mean reduction of 32% (Table 4 ). Antibiotics given after hospital admission have also been associated with a 57% reduction in diagnostic yield, although patients selected for aspiration after failure of initial therapy are not representative of newly presenting patients.122 Because of the large antibiotic effect, some authors have specifically excluded patients with a prior history of antibiotics.106 While this will lead to higher diagnostic yields, it excludes a group who differ on a number of potentially important variables, including severity of illness, and may produce an unrepresentative description of pneumonia etiology. If the presence of antibiotic is measured in urine, these patients do not need to be excluded and compensation can be made for the calculated antibiotic effect; furthermore, the history of antibiotic use correlates poorly with biological measures of antibiotic use.66 A separate approach to the problem of antibiotics is to use nonculture-based methods. Polymerase chain reaction for penicillin-binding protein genes PBP 2x and PBP 2b have been positive in lung aspirate specimens from 75% (9/12) and 89% (16/18) respectively, of adult patients with pneumonia whose blood or lung aspirate cultures were positive.11,123 Antigen detection in lung aspirate specimens using pneumococcal omniserum has been reported in several studies,11,104,109,123125 with sensitivities ranging from 12 to 92%.

Delay in culturing lung aspirate specimens can lead to significant loss of sensitivity.77 Positive results also will go undetected, particularly in populations with high HIV prevalence, if the aspirate specimen is not examined for cryptococcal antigen or stained for P carinii.,14,126 Legionella, Mycoplasma, and Chlamydia have rarely been sought in lung aspirate specimens.16,18 Respiratory viruses have been isolated by culture from only 7 of 176 aspirates (4%) in four studies of children,66,79,104,127 but the yield rises to 28 to 46% when indirect immunofluoresence is also employed.62,65 Ultimately, the sample volume is too small for comprehensive testing, and the microbiologic assays must be tailored to detect the likely pathogens in a given population.

Lung aspirate cultures may also be defined as “negative” if the organism isolated does not fit with our current concept of pneumonia etiology: an example is given by the isolation of poliovirus type 1 from the lungs of an Indian child.127 We have isolated Staphylococcus epidermidis from a Kenyan adult in lung aspirate cultures and two separate blood cultures, but dismissed the organism as a contaminant; against such prejudice, it would be difficult to detect a case of S epidermidis pneumonia should one ever arise.

Death

Rapidly popularized after its introduction, lung puncture was widely practiced at the turn of the 20th century but quickly became associated with fatal complications. Horder’s Lancet report29 of the technique in 1909 did not mention serious side effects, but contemporary medical publications referred to at least 11 deaths after thoracic puncture.128129 In 1905, an editorial in the British Journal of Childhood Diseases opined, “exploratory puncture of the chest is looked on by the younger members of the profession as a harmless proceeding, and on a par with the use of the stethoscope for the examination of children’s lungs, indeed there are some who give precedence to the needle.”130

We have been able to find 17 reports of death in patients who have undergone lung puncture (Table 5 ). There is no denominator for the 10 reports published between 1898 and 1920. Most of these cases had empyema or unresolved pneumonia; investigation of acute pneumonia was the indication in only one case. Three patients died of pulmonary hemorrhage, one of air embolism, and another of pneumothorax. The cause of death is unclear in the remainder; most patients died suddenly, and two deaths were preceded by neurologic symptoms suggestive of air embolism.

In the context of published series, seven deaths are reported from a denominator of 6,001 lung puncture procedures, giving a mortality rate of approximately 1 in 850 (Table 6 ). In three of these reports (patients 12, 15, and 17; Table 5), the investigators did not consider that the lung puncture was the cause of death, but merely report that the death occurred within a short time after the procedure. In most hospitals, the mortality of unselected pneumonia patients in the first hour after presentation is unlikely to be < 1 in 1,000, which suggests that moribund patients are normally avoided in lung puncture series. Interpretation of etiologic series using lung aspirates should take account of this bias toward less virulent microorganisms.

The complications in four fatalities directly attributable to lung puncture include one case each of air embolism and delayed recognition of pneumothorax; both are avoidable by an appropriate protocol.9,64 In a third case, a pneumothorax developed in a Filipino child, causing respiratory distress; a chest tube was put in place, but the child died of hemorrhage when the tube became dislodged.98 In the final case, lung aspiration caused a small pneumothorax and almost certainly contributed to death of a PNG child.66 However, the authors66 argue that, for any individual in the study, the potential benefits of lung aspiration still outweighed the potential risks. Optimization of antibiotic therapy allowed by accurate knowledge of etiology and susceptibility patterns can reduce mortality. In the PNG series, the case fatality rate among children in the study was 11% (9/83), compared with 17% (52/297) in a local control cohort without lung aspiration.66 The cohorts were not directly comparable and the difference is not statistically significant, but the argument is rational. Where the effect of lung aspiration on therapy has been monitored, it has influenced a change in 6 to 23% of patients,21,23,80 often quickly identifying unexpected pathogens, such as S aureus and M tuberculosis, that are associated with high mortality.,23

Pneumothorax

Pneumothorax is the major nonfatal complication of lung aspiration. From a summary of > 3,000 reported procedures, pneumothorax occurred in 3.3%, or 1 in 30 (Table 6). Among the studies summarized, some included routine postprocedure radiographs while others did not; many authors do not report whether radiographs were evaluated routinely. A small laminar pneumothorax is clearly a minor problem; pneumothorax requiring the insertion of a chest drainage tube is much more serious. Chest drainage was required following only 0.5%, or 1 in 200 procedures, among the series summarized in Table 6. By comparison, chest tube insertion is required in approximately 12%138 of aspiration lung biopsies, indicating the higher risk associated with a larger needle.

The risk of significant pneumothorax can be reduced by puncturing the lung only once40 and by avoiding inexperienced operators.25,40,93 Chronic obstructive airways disease and Pneumocystis carinii pneumonia (PCP) are particular risks for pneumothorax after thoracentesis,,14,139 and patients with less respiratory reserve are more likely to require drainage should a small pneumothorax develop.

In patchy bronchopneumonia, the consolidation may not abut against the pleura, and the risk of pneumothorax is theoretically higher than in lobar pneumonia. What little evidence exists tends to contradict this idea. In a series of 505 Chilean infants with bronchopneumonia, pneumothorax requiring chest drainage occurred in only 2 patients (0.4%)79; among 30 Indian children with suspected TB followed up with fluoroscopic examination, none developed pneumothoraces.113

Other Complications

Empyema, air embolism, and pulmonary hemorrhage were greatly feared by the pioneers of lung aspiration, but they are mentioned only rarely in the series collected in Table 6. There is not a single report of empyema thoracis, and numerous investigators in the preantibiotic era argued that the risk of empyema after pneumonia was not increased by the introduction of lung aspiration.15,40,85 Bullowa9 attributes five adverse events among > 2,500 aspirates to air embolism, but otherwise the complication is confined to patients undergoing aspiration lung biopsy.,25,140In closed lung aspiration, with the syringe firmly attached to the needle, the risk of introducing air to a vessel is minimal, but it increases significantly if a technique is used in which a stylet must be removed after a large needle is introduced.141

Although minor hemoptysis is relatively common, prolonged or serious pulmonary hemorrhage after lung puncture was reported after only 2 of> 3,000 procedures, and neither case was fatal. In studies that have directly compared fine-needle aspiration with larger-gauge biopsy cutting needles, the risk of pulmonary hemorrhage is considerably higher following lung biopsy.142143 In fact, in recent decades, major hemorrhage appears to be confined to patients undergoing trephine biopsy with large-bore needles.144In the biopsy of pulmonary nodules, aerated lung along the needle track does not provide any compression in the event of a vessel puncture.145

Pleuritic pain is rarely reported but occurs with a consistent frequency of approximately 2% in studies that do report it.80,101,113 Among Kenyan adults, 6 of 259 patients complained of pain requiring anal-gesia and one developed protracted hiccups (au-thors’ unpublished observations). Surgical emphysema62,96,113,146 and hematoma21 are infrequent local complications.

In addition to community-acquired pneumonia, lung aspiration has been used to investigate lung cancer, pulmonary nodules, diffuse lung disease, granulomata, lung abscess, cavitating lung disease, and chronic and unresponsive pneumonias.145 These aspirates are usually conducted under fluoroscopic or ultrasound control, using longer and wider-gauge needles. There is considerable heterogeneity in the conditions investigated, although the principal use of the technique is to diagnose primary lung neoplasms. Table 7 summarizes 19 investigations in which, by contrast, the objective of the study was to define an infective etiology in patients who did not have simple, acute, community-acquired pneumonia.

In these complex patients, conventional bacterial pathogens are encountered less frequently than P carinii, Nocardia spp, Mycobacteria, Legionella, Cytomegalovirus, Cryptococcus neoformans, Aspergillus, and other fungi. The diagnostic yield is impressive, with 60% of all investigations lead-ing to an etiologic diagnosis; however, the case definition in most studies is ill-defined, several investigators report only a successful subset of a larger experience,17,19,126,151 and many studies involve two or more punctures of the same patient.14,17,25,6869,119,142,147,150

What is clear is that the frequency of serious, nonfatal complications in complex patients is considerably higher than in simple pneumonia patients, with 18% of all aspirates leading to pneumothorax and 8% leading to insertion of a chest drain. Pulmonary hemorrhage is also more common, with serious events occurring in about 1% of patients. Air embolism was reported in only 1 of 1,018 patients. Wider reviews of> 8,000 lung aspiration procedures in patients with abnormal radiologic lung lesions have documented similar complication rates with only one fatality, no cases of air embolus, and only three cases of major hemorrhage; pneumothorax requiring chest drainage occurred in 8%.144,152153 Occasional case reports have drawn attention to fatal hemorrhage154 or air embolism.140

In patients with HIV, bedside lung aspiration with an ultrathin needle has succeeded in identifying an infective pulmonary pathogen in 29 of 47 procedures undertaken in Spanish adults150; half of these had PCP. Induced sputum specimens were positive in only 3 of 15 patients with PCP, suggesting that lung aspiration was targeted at patients whose early induced sputum specimens were known to be negative. Wallace et al147 diagnosed PCP using lung aspirates in 10 of 16 patients with AIDS or suspected AIDS, but 7 of these patients developed pneumothoraces and 3 required drainage. The risk of pneumothorax is increased in patients with previous or concurrent PCP or with a history of nebulized pentamidine use,155156 which constrains the use of lung aspiration in HIV-infected patients. In sub-Saharan Africa, where PCP is an uncommon cause of pneumonia in HIV, lung aspiration has been successful in the diagnosis of bacterial pneumonia.82

Guidance for the safe conduct of lung aspiration in 1904 advised premedication with brandy and the following emergency action: “Should the heart cease beating or the breathing become suspended during or after the puncture immediate resort by the medical attendant to artificial respiration and the injection hypodermically of ether and strychnine.”133 Although the procedure has remained unchanged in over 100 years, the preparation for the procedure and resuscitation techniques have evolved consid-erably.

The equipment required to support the procedure consists of a supply of oxygen, a chest drain insertion kit,157 and smaller cannulae for the rapid relief of smaller pneumothoraces. The success of the procedure is critically dependent on localization of disease; chest radiographs in two planes,82 CT scans,17,148 or ultrasonography19,126,151 can all enhance localization based on clinical signs. They may also reveal unexpected contraindications such as bullae or Echinococcus cysts.131 The prothrombin time, partial thromboplastin time, and platelet count should be checked when possible, and lung aspiration should be avoided in patients with significant derangements, eg, a prolongation of the prothrombin time by > 50% or a thrombocytopenia of < 70,000/mL.

A knowledge of the surface anatomy of the thorax is essential, especially in the child, to avoid striking the heart and great vessels; the illustrations in the cadaver study of Abdel-Khalik et al24 may be a useful aid. Targeting only peripheral consolidation, and therefore using a short needle (< 5 cm), will avoid puncture of a major vessel. Similarly, use of a finer needle (21- or even 25-gauge) will decrease the likelihood of pneumothorax without compromising the search for infectious etiologic agents.16 The aspirate system should be properly sealed throughout the procedure and use of a stylet should be avoided, if possible, to eliminate the possibility of air embolism during transfer to the syringe. Multiple stabs within the lung parenchyma may increase the risk of hemorrhage and have not been demonstrated to increase the diagnostic yield.21 Avoidance of normal lung will reduce the risk of pneumothorax, and lung puncture should never be done bilaterally.9 Patients who are too hypoxemic to suspend respiration for 1 to 2 s (Gherman and Simon119) and those who are unable to tolerate a pneumothorax briefly should not be investigated. In adults, COPD and lung bullae are relative contraindications, depending on severity and location; similarly, puncture should be avoided in children with hyperinflation. Pulmonary hypertension is also a contraindication,158 but the pressure in the pulmonary circulation is rarely known in the acute investigation of pneumonia.

The vital signs of the patient should be monitored closely for 24 to 48 h and a chest radiograph taken 1 h after the procedure. Of 160 pneumothoraces that developed in patients undergoing aspiration biopsy, 89% were evident on the radiograph performed immediately after the procedure, and 9% were first seen 1 h later. Pneumothoraces that appeared after the 1-h radiograph were never sufficiently important to require chest drainage.138 Patients requiring mechanical ventilation have been avoided by most investigators, although successful aspirations have been performed with suspension of assisted ventilation in eight patients.18

Lung aspiration has an undoubted role in etiologic studies of pneumonia. It provides considerable additional information over blood cultures on the species and serotypes causing disease, and their antibiotic susceptibilities, with high specificity. It is usually acceptable to the patient and carries a minimal risk of serious adverse effects, and the individual undergoing aspiration stands to benefit personally from the microbiologic information obtained. It has been so successful in some tropical environments that it is now commonly used as an adjunct to routine clinical management.62

Outside scientific studies lung aspiration may be considered on an individual basis for patients who have not responded to initial therapy, who may have nosocomial superinfection, who are immunocompromised, or in whom TB is suspected but has not been confirmed by examination of the sputum or gastric lavage.91 The more complex the patient’s clinical course, the more valuable the information obtained directly from the lung; at the same time, the more likely the procedure is to be complicated by coagulopathy, thrombocytopenia, pleural infection, or mechanical ventilation. There are no studies directly comparing lung puncture with blind therapy or other diagnostic procedures in these cases, but it seems likely, extrapolating from epidemiologic studies of pneumonia, that early use of the procedure in appropriately selected patients may improve patient outcomes. If lung aspiration is reserved as a diagnostic measure of last resort for the ICU patient with fever, diffuse lung disease, mechanical ventilation, multiorgan failure, and nosocomial flora in sputum, it will not provide timely diagnostic information and may well contribute to death. However, in the presence of dense peripheral consolidation, percutaneous transthoracic lung puncture can provide life-saving etiologic information with minimal risk in experienced hands; it can be performed quickly at the bedside, it requires no sedation, and it is likely to be more acceptable to the patient and less traumatic than the major alternative, bronchoscopic lavage, performed a few days later.

Abbreviations: PCP = Pneumocystis carinii pneumonia; PNG = Papua New Guinea; TB = tuberculosis

Supported by the Wellcome Trust of Great Britain through a Wellcome Trust research training fellowship in clinical epidemiology, No. 035375 (Dr. Scott).

Table Graphic Jump Location
Table 1. A Summary of Published Lung Aspirate Studies of Community-Acquired Pneumonia in Children Indicating the Numbers of Cultures Positive for Streptococcus pneumoniae, H influenzae, and S aureus and the Number and Percentage of Aspirates That Yielded a Single Bacterial Culture and Were Negative on Culture*
* 

Results of Gram’s stains, antigen detection, or polymerase chain reaction are not included in the table. Denominators have been adjusted where observations are missing.

Table Graphic Jump Location
Table 2. A Summary of Published Lung Aspirate Studies of Community-Acquired Pneumonia in Adults Indicating the Numbers of Cultures Positive for S pneumoniae, H influenzae, and S aureus and the Number and Percentage of Aspirates That Yielded a Single Bacterial Culture and Were Negative on Culture*
* 

Results of Gram’s stains, antigen detection, or polymerase chain reaction are not included in the table. Denominators have been adjusted where observations are missing.

Table Graphic Jump Location
Table 3. The Sensitivity of Lung Aspirate Cultures for Bacteremic Pneumonia and of Blood Cultures for Pneumonia with Positive Lung Juice Cultures
* 

Calculations are based on isolates of S pneumoniae and H influenzae only.

 

Calculations were restricted to patients who had no evidence of prior use of antibiotics on admission.

Table Graphic Jump Location
Table 4. Proportion of Lung Aspirates Yielding Positive Bacterial Cultures in Groups of Patients Who Have or Have Not Received Antibiotics Prior to Lung Puncture, With Risk Ratios and 95% Confidence Limits for a Positive Culture Given Antibiotics
Table Graphic Jump Location
Table 5. Clinical and Postmortem Details From 17 Patients Who Died Following Lung Puncture
* 

As cited in Porter Parkinson.134

Table Graphic Jump Location
Table 6. Number of Major Complications Following Lung Aspirate Procedures Reported in 42 Series With Summary Risks Per Procedure
Table Graphic Jump Location
Table 7. Diagnostic Yield and Number of Major Complications Following Lung Aspiration in 19 Studies of Patients With Complex Respiratory Disease Together With an Infective Pneumonitis
* 

The denominator is composed of the number of disease episodes, which is greater than the number of patients studied but less than the total number of aspirates performed.

 

The number of aspiration procedures, not the number of patients, is used to calculate the risks associated with lung aspiration (risk per procedure).

We are very grateful to Dr. Nicola Strnad for translation from German.

. British Thoracic Society Research Committee and the Public Health Laboratory Service (1987) Community-acquired pneumonia in adults in British hospitals in 1982–1983: a survey of aetiology—mortality, prognostic factors and outcome.Q J Med62,195-220. [PubMed]
 
Marrie, TJ, Durant, H, Yates, L Community-acquired pneumonia requiring hospitalization: 5-year prospective study.Rev Infect Dis1989;2,586-599
 
Fang, GD, Fine, M, Orloff, J, et al New and emerging etiologies for community-acquired pneumonia with implications for therapy: a prospective multicenter study of 359 cases.Medicine (Baltimore)1990;69,307-316. [PubMed]
 
Scott JAG, Hall AJ, Leinonen M. Validation of immune-complex enzyme immunoassays for the diagnosis of pneumococcal pneumonia in adults in Kenya. Clin Diagn Lab Immunol 2000 (in press).
 
Davidson, M, Tempest, B, Palmer, D Bacteriologic diagnosis of acute pneumonia: comparison of sputum, transtracheal aspirates and lung aspirates.JAMA1976;235,158-163. [PubMed] [CrossRef]
 
Lieberman, D, Schlaeffer, F, Boldur, I, et al Multiple pathogens in adult patients admitted with community-acquired pneumonia: a one-year prospective study of 346 consecutive patients.Thorax1996;51,179-184. [PubMed]
 
Dick, R, Heard, RB, Hinson, KFW, et al Aspiration needle biopsy of thoracic lesions: an assessment of 227 biopsies.Br J Dis Chest1974;68,86-94. [PubMed]
 
Gellis, SS Lung puncture and aspiration. Am J Dis Child. 1971;;122 ,.:277
 
Bullowa, JGM The management of the pneumonias: for physicians and medical students.1937,89-108 Oxford University Press. New York, NY:
 
Dorca, J, Manresa, F, Esteban, L, et al Efficacy, safety, and therapeutic relevance of transthoracic aspiration with ultrathin needle in nonventilated nosocomial pneumonia.Am J Respir Crit Care Med1995;151,1491-1496. [PubMed]
 
Garcia, A, Roson, B, Perez, JL, et al Usefulness of PCR and antigen latex agglutination test with samples obtained by transthoracic needle aspiration for diagnosis of pneumococcal pneumonia.J Clin Microbiol1999;37,709-714. [PubMed]
 
Gratten, M, Montgomery, J The bacteriology of acute pneumonia and meningitis in children in Papua New Guinea: assumptions, facts and technical strategies.P N G Med J1991;34,185-198. [PubMed]
 
Rosenow, EC A bacteriological and cellular study of the lung exudate during life in lobar pneumonia.J Infect Dis1911;8,500-503
 
Chaudhary, S, Hughes, WT, Feldman, S, et al Percutaneous transthoracic needle aspiration of the lung.Am J Dis Child1977;131,902-907. [PubMed]
 
Kereszturi, C, Hauptman, D Serum treatment of pneumonia in children.J Pediatr1934;4,331-334
 
Zavala, DC, Schoell, JE Ultrathin needle aspiration of the lung in infections and malignant diseases.Am Rev Respir Dis1981;1123,125-131
 
Conces, DJ, Clark, SA, Tarver, RD, et al Transthoracic needle biopsy: value in the diagnosis of pulmonary infections.AJR Am J Roentgenol1989;152,31-34. [PubMed]
 
Torres, A, Jimenez, P, Puig de la Bellacasa, J, et al Diagnostic value of non-fluoroscopic percutaneous lung needle aspiration in patients with pneumonia.Chest1990;98,840-844. [PubMed]
 
Yang, PC, Luh, KT, Chang, DB, et al Ultrasonographic evaluation of consolidation.Am Rev Respir Dis1992;146,757-762. [PubMed]
 
Manresa, F, Dorca, J Needle aspiration techniques in the diagnosis of pneumonia.Thorax1991;46,601-603. [PubMed]
 
Barnes, DJ, Narqi, S, Igo, JD The role of percutaneous lung aspiration in the bacteriological diagnosis of pneumonia in adults.Aust N Z J Med1988;18,754-757. [PubMed]
 
Thomas, HM, Parker, F Results of antemortem lung punctures in lobar pneumonia: their bearing on the mechanism of crisis.Arch Intern Med1920;26,125-132
 
Falade, AG, Mulholland, EK, Adegbola, RA, et al Bacterial isolates from blood and lung aspirate cultures in Gambian children with lobar pneumonia.Ann Trop Paediatr1997;17,315-319. [PubMed]
 
Abdel-Khalik, AK, Askar, AM, Ali, M The causative organisms of bronchopneumonia in infants in Egypt.Arch Dis Child1938;13,333-342. [PubMed]
 
Woolf, CR Applications of aspiration lung biopsy with a review of the literature.Dis Chest1954;25,286-301
 
Leyden, VII. Verhandlungen des Vereins für innere Medizin: über infectiöse Pneumonie.Dtsch Med Wochenschr1883;9,52-54
 
Günther, VII. Verhandlungen des Vereins für innere Medizin: über infectiöse Pneumonie.Dtsch Med Wochenschr1883;9,52-54
 
Talamon, Coccus de la pneumonie.Bull Soc Anat Paris1883;63,475-481
 
Horder, TJ Lung puncture: a new application of clinical pathology.Lancet1909;2,1345-1346
 
Puncture of the lung as a diagnostic procedure [editorial]. Lancet 1909; 2:1366–1367.
 
Lyon, AB Bacteriologic studies of one hundred and sixty-five cases of pneumonia and post-pneumonic empyema in infants and children.Am J Dis Child1922;23,72-87
 
Tchistovitch, N Contribution à l’étude de la pathogénie de la crise dans la pneumonie fibrineuse.Ann Inst Pasteur1904;18,304-322
 
Lister, FS An experimental study of prophylactic inoculation against pneumococcal infection in rabbit and man.Publ S Afr Inst Med Res1916;1,231-287
 
Woolf, CR Applications of aspiration lung biopsy with a review of the literature. Dis Chest. 1954;;25 ,.:286
 
Abrahams, A, Hallows, NF, Eyre, JWH, et al Purulent bronchitis: its influenzal and pneumococcal bacteriology.Lancet1917;2,377-382
 
Stewart, D Lung puncture in lobar pneumonia.Lancet1930;2,520-521
 
Ellison, JB Pneumonia in measles.Arch Dis Child1931;6,37-52. [PubMed]
 
Locke, EA, Thomas, HM, O’Hara, D Acute respiratory infections at the Boston City Hospital.Boston Med Surg J1920;183,480-490
 
Cruickshank, R The treatment of lobar pneumonia by Felton serum: the serological typing of pneumonia.Glasgow Med J1931;115,59-63
 
Sappington, SW, Favorite, GO Lung puncture in lobar pneumonia.Am J Med Sci1936;191,225-234
 
Alston, JM, Stewart, D Incidence of the serological types of the pneumococcus in pneumonia.Br Med J1930;2,860-862. [PubMed]
 
Stankiewicz, R, Przesmycki, F, Kolago, J, et al La flore bacterienne des pneumopathies (pneumonie et bronchopneumonie) de l’enfant, en considérant les différents types de pneumocoques.Rev Fr Pediatr1932;8,324-341
 
Blacklock, JWS, Guthrie, KJ Pneumococcal infections in infancy and childhood.J Pathol Bacteriol1933;36,349-368
 
Shibley, GS, Rogers, ES Occurrence of rough pneumococci in lungs of patients with lobar pneumonia.Proc Soc Exp Biol Med1933;21,6-9
 
Bullowa, JGM The reliability of sputum typing and its relation to serum therapy.JAMA1935;105,1512-1518
 
Bullowa, JGM, Greenbaum, E Pneumococcic pneumonia in infants and in children.Am J Dis Child1937;53,22-31
 
Garnham, PCC, de Smidt, FPG Preliminary note on pneumonia in Nairobi with particular reference to type specific incidence of pneumococcus.East Afr Med J1931;8,150-163
 
Gohar, MA, Abboud, MA The causative organisms of pneumonia in children in Egypt.J Lab Clin Med1933;19,751-754
 
Netter, A L’epidémie d’influenza de 1918.Bull Acad Med1918;80,275-286
 
Chickering, H, Park, J Staphylococcus aureuspneumonia.JAMA1919;72,617-626
 
Fletcher, W Meningococcus broncho-pneumonia in influenza.Lancet1919;1,104-105
 
Glynn EE, Digby L. Bacteriologic and clinical observations on pneumonia and empyemata with special reference to the pneumococcus and to serum treatment. London, UK: HMO Stationery Office, Great Britain Medical Research Council Special Report Series No. 79, 1923; 105–112.
 
Rice, RP, Loda, F A roentgenographic analysis of respiratory syncytial virus pneumonia in infants.Radiology1966;87,1021-1027. [PubMed]
 
Johnson, HD, Johnson, WW Pneumocystis cariniipneumonia in children with cancer.JAMA1970;214,1067-1073. [PubMed]
 
Stites, DP, Glezen, WP Pulmonary nocardiosis in childhood: a case report.Am J Dis Child1967;114,101-105. [PubMed]
 
Wyllie, WG, Sheldon, W, Bodian, M, et al Idiopathic pulmonary hemosiderosis (essential brown induration of the lungs).Q J Med1948;13,25-48
 
Eyre, J The bacteriology of bronchopneumonia: a statistical analysis.J Pathol Bacteriol1910;14,160-163
 
Clough, MC A study of the incidence of the types of pneumococci isolated from acute pneumonia and other infections, and an analysis of the cases classified by types in regard to mortality complications, associated diseases, bacteremia, and leucocytosis.Johns Hopkins Hosp Bull1917;28,306-311
 
Cecil, RL, Baldwin, HS, Larsen, NP Lobar pneumonia, a clinical and bacteriologic study of two thousand typed cases.Arch Intern Med1927;40,253-280
 
Liston, WG Note on the bacteriology of broncho-pneumonia in children.Arch Dis Child1929;4,283-290. [PubMed]
 
Axton, JH, Giles, HM Lung puncture aspirate and blood culture taken immediately after death in malnourished children.S Afr Med J1972;46,563-565. [PubMed]
 
Adegbola, RA, Falade, AG, Sam, BE, et al The etiology of pneumonia in malnourished and well-nourished Gambian children.Pediatr Infect Dis J1994;13,975-982. [PubMed]
 
Fagbule, DO Bacterial pathogens in malnourished children with pneumonia.Trop Geogr Med1993;45,294-296. [PubMed]
 
Wesley, AG, Sutton, JB, Widrich, AJ The etiology of pneumonia associated with measles in Bantu children.S Afr Med J1971;45,1402-1404. [PubMed]
 
Avila, M, Salomon, H, Carballal, G, et al Isolation and identification of viral agents in Argentinian children with acute lower respiratory tract infection.Rev Infect Dis1990;12(Suppl 8),S974-S981. [PubMed]
 
Shann, F, Gratten, M, Germer, S, et al Etiology of pneumonia in children in Goroka hospital, Papua New Guinea.Lancet1984;2,537-541. [PubMed]
 
Scott, JAG, Hall, AJ, Hannington, A, et al Serotype distribution and prevalence of resistance to benzylpenicillin in three representative populations ofStreptococcus pneumoniaeisolates from the coast of Kenya.Clin Infect Dis1998;27,1442-1450. [PubMed]
 
Castellino, RA, Blank, N Etiologic diagnosis of focal pulmonary infection in immunocompromised patients by fluoroscopically guided percutaneous needle aspiration.Radiology1979;132,563-567. [PubMed]
 
Palmer, DL, Davidson, M, Lusk, R Needle aspiration of the lung in complex pneumonias.Chest1980;78,16-21. [PubMed]
 
Riley, ID, Tarr, PI, Andrews, M, et al Immunisation with a polyvalent pneumococcal vaccine: reduction of adult respiratory mortality in a New Guinea Highlands community.Lancet1977;1,1338-1341. [PubMed]
 
Mulholland, K, Hilton, S, Adegbola, R, et al Randomised trial ofHemophilus influenzaetype-b tetanus protein conjugate vaccine for prevention of pneumonia and meningitis in Gambian infants.Lancet1997;349,1191-1197. [PubMed]
 
Sinha, DP, Hughes, JR Lung tap in lower respiratory tract infection in children.Indian Pediatr1966;3,385-392. [PubMed]
 
Lees, AW, McNaught, W Bacteriology of lower-respiratory tract secretions, sputum, and upper-respiratory-tract secretions in “normals” and chronic bronchitis.Lancet1959;2,1112-1115. [PubMed]
 
Laurenzi, GA, Potter, TR, Kass, EH Bacteriologic flora of the lower respiratory tract.N Engl J Med1961;265,1273-1278. [PubMed]
 
Kalinske, RW, Parker, RH, Brandt, D, et al Diagnostic usefulness and safety of transtracheal aspiration.N Engl J Med1967;276,604-608. [PubMed]
 
Berman, SZ, Mathison, DA, Stevens, DD, et al Transtracheal aspiration studies in asthmatic patients in relapse with “infective” asthma and in subjects without respiratory disease.J Allergy Clin Immunol1975;56,206-214. [PubMed]
 
Moser, KM, Maurer, J, Jassy, L, et al Sensitivity, specificity and risk of diagnostic procedures in a canine model ofStreptococcus pneumoniaepneumonia.Am Rev Respir Dis1982;125,436-442. [PubMed]
 
Pecora, DV, Yegian, D Bacteriology of the lower respiratory tract in health and chronic diseases.N Engl J Med1958;258,71-74. [PubMed]
 
Mimica, I, Donoso, E, Howard, J, et al Lung puncture in the etiological diagnosis of pneumonia.Am J Dis Child1971;122,278-282. [PubMed]
 
Zalacain R, Llorente JL, Gaztelurrutia L, et al. La punción transtorácica aspirativa con aguja ultrafina en las neumonías de alto riesgo adquiridas en la comunidad. Med Clin (Barc) 1993; 100:567–570.
 
Schuster, CA, Duffau, TG, Nichols, RE, et al Lung aspirate puncture as a diagnostic aid in pulmonary tuberculosis in childhood.Pediatrics1968;42,647-650. [PubMed]
 
Scott JAG, Hall AJ, Muyodi C, et al. The etiology, outcome and mortality risk factors for acute pneumonia in adults on the coast of Kenya. Lancet 1999 (in press).
 
Evans, AD, Evans, M Staphylococcal infection of the lower respiratory tract in adults with influenza.Lancet1956;1,771-774
 
Schwarzmann, SW, Adler, JL, Sullivan, RJ, et al Bacterial pneumonia during the Hong Kong influenza epidemic of 1968–1969: experience in a city county hospital.Arch Intern Med1971;127,1037-1041. [PubMed]
 
Alexander, HE, Craig, HR, Shirley, RG, et al Validity of etiological diagnosis of pneumonia in children by rapid typing from nasopharyngeal mucus.J Pediatr1941;18,31-35
 
Disney, ME, Wolff, J, Wood, BSB Staphylococcal pneumonias in infants.Lancet1956;1,767-771
 
Hughes, JR, Bose, SK, Kloene, W, et al Acute lower respiratory tract infections in Calcutta children: an etiologic study.Indian Pediatr1966;3,201-211. [PubMed]
 
Schuster, A, Pino, M, Neira, M, et al La puncion biopsia pulmonar como metodo diagnostico de las neumopatias de la infancia.Pediatria (Santiago)1966;9,9-12
 
Klein, JO Diagnostic lung puncture in the pneumonias of infants and children.Pediatrics1969;44,486-492. [PubMed]
 
Cruickshank, JG, Gelfand, M, Sanders, D Lung puncture in the diagnosis of acute pneumonia in children and adults.Cent Afr J Med1971;17,247-249. [PubMed]
 
Garcia de Olarte, D, Trujillo, H, Uribe, A, et al Lung puncture-aspiration as a bacteriologic diagnostic procedure in acute pneumonias of infants and children.Clin Pediatr (Phila)1971;10,346-350. [PubMed]
 
Alix, MC, Chiu, VC Lung tap: a new diagnostic procedure.Philipp J Pediatr1973;15,22-23
 
Prinsloo, JG, Cicoria, A Is longpunksie-aspirasie ’n onskuldige prosedure?S Afr Med J1974;48,597-598. [PubMed]
 
Rozov, T, Sakane, PT, Novaes, JRC, et al Contribuição ao estudo da etiolgia das pneumopatias da infância, por meio da aspiração pulmonar transcutânea.Pediatria Pratica (Sao Paulo)1974;45,221-332
 
Escobar, JA, Dover, AS, Duenas, A, et al Etiology of respiratory tract infections in children in Cali, Colombia.Pediatrics1976;57,123-130. [PubMed]
 
Silverman, M, Stratton, D, Diallo, A, et al Diagnosis of acute bacterial pneumonia in Nigerian children.Arch Dis Child1977;52,925-931. [PubMed]
 
Shakunthala, SK, Mallikarjuna Rao, G, Urmila, S Diagnostic lung puncture aspiration in acute pneumonia of newborn.Indian Pediatr1978;15,39-44. [PubMed]
 
Cunanan, GD, Yambao, JJ, Caces, RF, et al Diagnostic lung aspiration in severe and complicated childhood pneumonia.Philippine J Pediatr1979;28,169-178
 
Ferreira, OS, Carvalho, MRdC Estudo comparativo entre os resultados das culturas do material colhido pela “swab” faringeo e pela puncao pulmonar, em 58 crincas portadoras de pneumonia aguda.J Pediatr (Rio J)1980;48,211-215
 
Diakparomre, MA, Obi, JO Aetiological diagnosis of pneumonia in children by lung puncture.Nigerian J Paediatr1981;8,61-64
 
Kalra, SK, Sasidharan, T, Vatwani, V, et al Lung puncture: a diagnostic aid in childhood pneumonia.Indian Pediatr1981;18,727-730. [PubMed]
 
Wall, RA, Corrah, PT, Mabey, DCW, et al The etiology of lobar pneumonia in the Gambia.Bull World Health Organ1986;64,553-558. [PubMed]
 
Ikeogu, MO Acute pneumonia in Zimbabwe: bacterial isolates by lung aspiration.Arch Dis Child1988;63,1266-1267. [PubMed]
 
Forgie, IM, O’Neill, KP, Lloyd-Evans, N, et al Etiology of acute lower respiratory tract infections in Gambian children: II. Acute lower respiratory tract infection in children ages one to nine years presenting at the hospital.Pediatr Infect Dis J1991;10,42-47. [PubMed]
 
Nuzum, JW, Pulat, I, Stangl, FG, et al Pandemic Influenza and pneumonia in a large civil hospital.JAMA1918;71,1562-1565
 
Douglas, RM, Devitt, L Pneumonia in New Guinea: I. Bacteriological findings in 632 adults with particular reference toHemophilus influenzae.Med J Aust1973;1,42-49
 
Macfarlane, JT, Adegboye, DS, Warrell, MJ Mycoplasma pneumoniaeand aetiology of lobar pneumonia in northern Nigeria.Thorax1979;34,713-719. [PubMed]
 
Macfarlane, JT, Finch, RG, Ward, MJ, et al Hospital study of adult community-acquired pneumonia.Lancet1982;2,255-258. [PubMed]
 
Bella, F, Tort, J, Morera, M-A, et al Value of bacterial antigen detection in the diagnostic yield of transthoracic needle aspiration in severe community acquired pneumonia.Thorax1993;48,1227-1229. [PubMed]
 
Wallace, RJ, Musher, DM, Martin, RR Hemophilus influenzaepneumonia in adults.Am J Med1978;64,87-93. [PubMed]
 
Barnes, DJ, Naraqi, S, Igo, JD Hemophilus influenzaepneumonia in Melanesian adults: report of 15 cases.Thorax1987;42,889-891. [PubMed]
 
Page, MI, Lunn, JS Pneumococcal serotypes associated with acute pneumonia.Am J Epidemiol1973;98,255-261. [PubMed]
 
Bhandari, B, Singh, SV, Sharma, VK Lung puncture: aid to bacteriological diagnosis of pulmonary tuberculosis in children.Indian Pediatr1971;8,716-719. [PubMed]
 
Dahlgren, SE, Ekstrom, P Aspiration cytology in the diagnosis of pulmonary tuberculosis.Scand J Respir Dis1972;53,196-201. [PubMed]
 
Rothenberg, J, Poellein, S, Penkava, RR Diagnosing tuberculosis by fine-needle aspiration biopsy.Am Fam Physician1982;25,119-120
 
Tewari, AD, Sen, R, Mittal, KK, et al Lung puncture aspiration in the diagnosis of acute pneumonias.Indian Pediatr1991;28,647-652. [PubMed]
 
Wollestein, M, Benson, AW Types of pneumococcus found in the pneumonias of infants and young children.Am J Dis Child1916;12,254-265
 
Forgie, IM, O’Neill, KP, Lloyd-Evans, N, et al Etiology of acute lower respiratory tract infections in Gambian children: I. Acute lower respiratory tract infections in infants presenting at hospital.Pediatr Infect Dis J1991;10,33-39. [PubMed]
 
Gherman, CR, Simon, HJ Pneumonia complicating severe underlying disease: a current appraisal of transthoracic lung puncture.Dis Chest1965;48,297-304. [PubMed]
 
Tobin, MJ Diagnosis of pneumonia: techniques and problems.Clin Chest Med1987;8,513-527. [PubMed]
 
San Pedro, GS, Campbell, CD, Tryka, AF Factors affecting the sensitivity of needle aspiration in a rodent model of Gram-negative bacillary pneumonia.Chest1989;95,198S-199S
 
Zalacain, R, Llorente, JL, Gaztelurrutia, L, et al Influence of three factors on the diagnostic effectiveness of transthoracic needle aspiration in pneumonia.Chest1995;107,96-100. [PubMed]
 
Ruiz-Gonzalez, A, Nogues, A, Falguera, M, et al Rapid detection of pneumococcal antigen in lung aspirates: comparison with culture and PCR technique.Respir Med1997;91,201-206. [PubMed]
 
Misra, S, Bhakoo, ON, Ayyagiri, A, et al Clinical and bacteriological profile of neonatal pneumonia.Indian J Med Res1991;93,366-370. [PubMed]
 
Aderaye, G Community acquired pneumonia in adults in Addis Abeba: etiologic agents, clinical and radiographic presentation.Ethiop Med J1994;32,115-123. [PubMed]
 
Liaw, Y-S, Yang, P-C, Yu, C-J, et al Direct determination of cryptococcal antigen in transthoracic needle aspirate for diagnosis of pulmonary cryptococcosis.J Clin Microbiol1995;33,1588-1591. [PubMed]
 
Hughes, JR, Sinha, DP, Cooper, MR, et al Lung tap in childhood: bacteria, viruses, and mycoplasmas in acute lower respiratory tract infections.Pediatrics1969;44,477-485. [PubMed]
 
Russell, AE Lung puncture. Lancet. 1909;;2 ,.:1539
 
Armstrong H. A possible danger in exploratory puncture of the chest. Liverpool Medico-Chirugical J 1906:110–114.
 
Exploratory puncture of the chest: abuse of the needle—a warning [editorial]. Br J Child Dis 1905; 2:466–473.
 
Russell, AE On death occurring during or after exploratory puncture of the lung.St Thomas’s Hosp Rep1898;28,465-484