The clinical impact of nucleic acid amplification tests on the diagnosis and management of tuberculosis in a British hospital

Transcription

1 1 Tropial and Infetious Diseases Unit, Royal Liverpool University Hospital, Liverpool, UK; 2 Department of Medial Mirobiology, Royal Liverpool University Hospital, Liverpool, UK; 3 Tuberulosis Researh Unit, Cardiothorai Centre, Liverpool, UK Correspondene to: Dr M Taegtmeyer, Tropial and Infetious Diseases Unit, Royal Liverpool University Hospital, Presot Street, Liverpool L7 8XP, UK; M.Taegtmeyer@liverpool. a.uk Reeived 17 May 2007 Aepted 9 Otober 2007 Published Online First 5 Deember 2007 The linial impat of nulei aid amplifiation tests on the diagnosis and management of tuberulosis in a British hospital M Taegtmeyer, 1 N J Beehing, 1 J Sott, 2 K Seddon, 2 S Jamieson, 3 S B Squire, 1 H C Mwandumba, 1 A R O Miller, 1 P D O Davies, 3 C M Parry 2 ABSTRACT Bakground: Nulei aid amplifiation tests (NAAT) based on PCR provide rapid identifiation of Myobaterium tuberulosis and the detetion of rifampiin resistane. Indiations for their use in linial samples are now inluded in British tuberulosis guidelines. Methods: A retrospetive audit of patients with suspeted myobaterial infetion in a Liverpool hospital between 2002 and Doumentation of the impat of NAAT usage in aid fast baillus (AFB) mirosopy positive samples on linial pratie and the influene of a multidisiplinary group on their appropriate use, ompared with British guidelines. Results: Myobateria were seen or isolated from 282 patients and identified as M tuberulosis in 181 (64%). NAAT were indiated in 87/123 AFB positive samples and performed in 51 (59%). M tuberulosis was onfirmed or exluded by this method in 86% of tested samples within 2 weeks, ompared with 7% identified using standard methods. The appropriate use of NAAT inreased signifiantly over the study period. The NAAT result had a linial impat in 20/51 (39%) tested patients. Culture results suggest the potential for a diret linial impat in 8/36 (22%) patients in whih it was indiated but not sent and 5/36 (14%) patients for whom it was not indiated. Patients managed by the multidisiplinary group had a higher rate of HIV testing and appropriate use of NAAT. Conlusions: There were signifiant linial benefits from the use of nulei aid amplifiation tests in this low prevalene setting. Our data suggest that there would be additional benefit from their use with all AFB smear positive linial samples. Tuberulosis (TB) is an inreasing problem in the UK 1 and prompt diagnosis improves outome and failitates appropriate infetion ontrol and publi health ations. 2 Myobaterium tuberulosis (MTB) and environmental myobateria have a similar mirosopi appearane and moleular tests that provide rapid identifiation are inreasingly used in linial pratie. 3 4 British guidelines 2 5 reommend the use of nulei aid amplifiation tests (NAAT) in aid fast baillus (AFB) smear positive samples if rapid onfirmation of TB diagnosis would alter the are of the patient, or before onduting a large ontat-traing initiative. Global onerns about rising rates of multidrug resistant (MDR) and extensively drug resistant strains of MTB 6 7 mean that tehnologies allowing early detetion of resistane are inreasingly important. As rifampiin resistane is ommonly Tuberulosis due to mutations in a single gene (the rpob gene), this an also be deteted using a NAAT method. 8 9 Although the urrently available tests for rpob mutations do not detet all ases of rifampiin resistane, and will not detet isolated isoniazid resistane, the assoiation between rifampiin resistane and MDR-TB is strong, with one report showing 95% of rifampiin resistant strains to be assoiated with resistane to isoniazid. 10 British guidelines therefore also advoate use of rapid diagnosti tests for rifampiin resistane if a risk assessment suggests a patient might have MDR- TB. Detetion of rifampiin resistane by NAAT is taken to indiate MDR-TB until full sensitivity profiles beome available. Liverpool has a low prevalene but a rising inidene of TB, with 47 ases notified in 2002 and 86 in 2005, out of a population of approximately whih has hanged little in that time period. 11 The reent inreases reflet national trends 12 and oinide with Liverpool being designated as a dispersal entre for refugees in We onduted a retrospetive analysis of the linial impat of the use of NAAT in patients with suspeted TB, who had a linial sample that was AFB smear positive between 2002 and 2006, and we examined the influene of a multidisiplinary approah on the appropriate use of investigations in this low prevalene setting. METHODS The Department of Mirobiology of the Royal Liverpool University Hospital reeives samples for TB diagnosis from inpatient wards, from a hest lini that undertakes TB sreening, from the Tropial and Infetious Diseases Unit and from some general pratitioners (family physiians). A multidisiplinary group was onvened in 2003 to improve the management and oordination of are of patients with TB. This group inludes infetious disease and hest liniians as well as linial mirobiologists and TB speialist ommunity nurses. In a weekly meeting, the group reviews patients with suspeted TB and patients on treatment with ative problems, inluding deisions about sending samples for NAAT. It had been the poliy of the tropial and infetious disease unit for at least 7 years, and more reently of the multidisiplinary team, to request HIV antibody testing for all patients with suspeted or proven myobaterial infetion, unless the patient refuses onsent. Timing of HIV antibody testing related to Thorax: first published as /thx on 16 November Downloaded from on 1 September 2018 by guest. Proteted by opyright. Thorax 2008;63: doi: /thx

2 Tuberulosis Table 1 Isolate Myobaterial isolates and usage of nulei aid amplifiation tests the time of availability of myobaterial tests was not reorded for the purposes of this study. British guidelines have reommended HIV testing of patients with TB sine the year Sputum and bronhial lavage samples obtained from patients with suspeted myobaterial infetion were routinely treated with dithiothreitol (Sputasol, Oxoid, Basingstoke, UK), entrifuged and the deposit used to make a smear. For samples from sterile sites, suh as fluids or tissues, a diret smear was made. All smears were then examined by auramine phenol stain using fluoresene mirosopy. Non-sterile samples were deontaminated with 4% sodium hydroxide and onentrated by entrifugation. The deontaminated non-sterile samples and all sterile samples were inoulated into MBBaT bottles ontaining liquid media for myobaterial ulture and inubated in an automated system for 6 10 weeks (Organon Teknika, Dublin, Ireland). Culture positive samples were submitted to the Regional Centre of Myobateriology, Newastle, UK for identifiation and sensitivity testing. MDR isolates of MTB were defined as those resistant to rifampiin and isoniazid. 6 If the sample was AFB smear positive sputum, a fresh undeontaminated sample was requested and sent to the United Kingdom Health Protetion Ageny Myobaterial Referene Unit who performed NAAT using the INNO-LiPA Rif.TB assay (Immunogenetis, Zwijndreht, Belgium). 14 Samples were sent if rapid onfirmation of MTB would AFB smear positive NAAT indiated Sent Not sent NAAT not indiated AFB smear negative M tuberulosis omplex (MDR isolate)* 180 (8) 98 (4) 47 (3) 32 (1) 19 (0) 82 (4) M avium omplex M gordonae M kansasii Other{ Culture negative *Multidrug resistant isolate (resistant to rifampiin and isoniazid). {M xenopi (6), M malmoense (5), M helonae (4), M absessus (3), M marinum (1), M lentiflavum (1), M szulgai (1), environmental Myobateria not speiated (4). AFB, aid fast baillus; MDR, multidrug resistant; NAAT, nulei aid amplifiation tests. Figure 1 Use of nulei aid amplifiation tests (NAAT) between 2002 and 2006 in 87 aid fast baillus smear positive samples for whih their use was indiated by British guidelines. 2 Perentages indiate the proportion of samples with NAAT performed eah year. influene linial management (by exlusion of non-tb myobateria in HIV positive or otherwise immunoompromised patients); if there were onerns about possible drug resistane, inluding MDR-TB (previous TB treatment, ontat with MDR-TB, a patient from a TB endemi area); or if a large ontat-traing exerise would be needed if the isolate was onfirmed as MTB. 2 Data from all TB ases were routinely olleted by the linial staff on a simple proforma and annual files stored. A retrospetive ase note audit and review of laboratory reords was onduted to supplement information from this ontemporaneously olleted data. All patients with AFB smear positive linial samples submitted between January 2002 and Deember 2006 were identified and inluded. Data were olleted on the timing of smear and ulture results, the use of NAAT and their impat on linial deision making. The appropriateness of referral for NAAT and the proportion of patients with TB sreened for HIV infetion were determined from the linial data and stratified aording to whether the patient was managed by the multidisiplinary group or by other physiians. Data were entered and analysed in Exel, using simple desriptive statistis. Proportions were expressed as perentages with 95% onfidene intervals (CI) where appropriate and ompared using the x 2 test or Fisher s exat test. Continuous variables were ompared using the t test. For all analyses, statistial signifiane was inferred if the p value was below The study was registered with the hospital s linial information audit ommittee and all personal data were stored and analysed in aordane with national guidelines. 15 RESULTS During the study period, samples were submitted from 6985 patients. Myobateria were seen on smears or isolated from the ultures of 651 samples taken from 282 (4%) of the investigated patients. A total of 123 (44%) of the 282 patients produed a sample that was AFB smear positive on at least one oasion. Ninety-three (76%) of these 123 samples were sputum, 13 (11%) were bronhoalveolar lavage samples, 7 (6%) were ervial lymph node aspirates and the remainder were from other sites. TB was onfirmed by ulture in 180 patients (8/180 MDR), by NAAT alone in one patient and other myobateria were isolated from the remaining 100 patients (table 1). Two patients were santy smear positive on single samples but ulture negative. One was a lymph node aspirate from a patient previously treated for TB and was NAAT positive Thorax: first published as /thx on 16 November Downloaded from on 1 September 2018 by guest. Proteted by opyright. 318 Thorax 2008;63: doi: /thx

3 Tuberulosis Table 2 Nulei aid amplifiation test results from aid fast baillus smear positive samples, for M tuberulosis. The other was from a bronhoalveolar lavage in a patient with a linial and histologial diagnosis of bronhial arinoma. Using ulture as the gold standard, the smear sensitivity for M tuberulosis was 98/180 (54% (95% CI 47%, 62%)) and speifiity was 77/102 (75% (95% CI 66%, 83%)). Use of NAAT NAAT were indiated for 87/123 (71%) patients and were performed in 51 (59%). In the remaining 36, NAAT was not performed beause it was thought unneessary or not onsidered (n = 19) or an additional undeontaminated sample was not available (n = 17). NAAT was arried out on 44/66 (67% (95% CI 55%, 77%)) AFB positive sputum samples in whih they were indiated ompared with 7/21 (33% (95% CI 16%, 55%)) non-sputum samples (p,0.001). The appropriate use of NAAT inreased over the study period (fig 1). Between 2002 and 2004, NAAT were indiated in 42 AFB smear positive samples and performed in 16 (38% (95%CI 24%, 53%)). This inreased in 2005 and 2006 to 35/45 (78% (95% CI 64%, 88%)) samples (p,0.001). The ommonest indiation for requesting NAAT was that the patient originated from a TB endemi ountry where there was NAAT indiates MTB (n (%)) Culture onfirms MTB (n (%)) NAAT indiates rifampiin resistane (n (%)) Indiation for NAAT present Sent (82) 47 (92) 2 (4) 3 (6) Myobaterial (77) 10 (77) 0 0 identifiation* Rifampiin (82) 37 (97) 2 (5) 3 (8) resistane* Not sent 36 N/A 32 (84) N/A 1 (3) Myobaterial 4 2 (50) 0 identifiation* Rifampiin (94) 1 (3) resistane* Indiation for NAAT not present 36 N/A 19 (53) N/A (34) 98 (80) 2 (2) 4 (3) *The prinipal reason for sending sample for NAAT. AFB, aid fast baillus; MDR, multidrug resistant; MTB, Myobaterium tuberulosis; N/A, not appliable; NAAT, nulei aid amplifiation tests. Figure 2 Delay between reeipt of aid fast baillus (AFB) smear positive samples and identifiation of isolate by nulei aid amplifiation tests (NAAT) (in 51 sent for NAAT), identifiation of the isolate by ulture (in all 123 samples) and detetion of rifampiin sensitivity (in all 98 samples that grew MTB), Culture indiates rifampiin resistane (n (%)) a onern about the possibility of drug resistant infetion. A total of 27 (31%) of the 87 patients in whom NAAT were indiated were HIV positive. However, HIV status was not determined in 20/87 (23% (95% CI 15%, 33%)) patients in whom NAAT were otherwise indiated and 27/36 (75% (95% CI 59%, 87%)) patients in whom NAAT were not indiated. Identifiation and sensitivity results Final myobaterial identifiation and sensitivity results with a breakdown of NAAT usage are summarised in table 1 and the results of NAAT ompared with routine ulture in the AFB smear positive samples are shown in table 2. NAAT deteted MTB in 41/47 samples in whih MTB was subsequently isolated (sensitivity 87% (95% CI 75%, 95%)). In five patients NAAT was negative (in three only santy AFB were seen on the smear) but MTB was eventually isolated (inluding one MDR strain), and in the sixth, inhibitors rendered the NAAT result indeterminate. In one further sample, a lymph node aspirate from a patient previously treated for TB, MTB was deteted by NAAT but not isolated by ulture. The speifiity of the NAAT results was 3/4 (75% (95% CI 24%, 99%)). MTB was signifiantly more likely to be isolated from a sample in whih NAAT was indiated by the British guidelines ompared with those in whih it was not indiated (79/87 (91% (95% CI 83%, 96%)) vs 19/36 (53% (95% CI 37%, 69%)); p,0.001). NAAT deteted a mutation onsistent with rifampiin resistane in two patients, both onfirmed by subsequent ulture. In one patient with a rifampiin resistant isolate NAAT were negative for MTB. In one further patient with a rifampiin resistant isolate, NAAT were indiated but not sent. The time to identifiation of MTB and detetion of rifampiin resistane was signifiantly redued by the use of NAAT. The mean (SD) time to identifiation of MTB and detetion of rifampiin resistane in patients with NAAT performed was 9.8 (5.9) days ompared with 26.0 (10.9) days to identifiation (p,0.001) and 40.1 (12.9) days to sensitivity test results (p,0.001) using standard methods. For all of the AFB smear positive samples, a total of 86% of samples were identified by NAAT within 2 weeks ompared with 7% of samples identified using standard methods (fig 2). Clinial impat of the NAAT result The use of NAAT had a linial impat in 20/51 (39% (95% CI 27%, 53%)) patients for whom it was performed. Three patients who had started TB treatment were able to stop within 2 weeks Thorax: first published as /thx on 16 November Downloaded from on 1 September 2018 by guest. Proteted by opyright. Thorax 2008;63: doi: /thx

4 Tuberulosis and in one of these ases a major prison ontat-traing exerise was avoided; two of four patients with MDR-TB were identified promptly by NAAT; in three patients, detetion of MTB onfirmed the need for a hospital ontat-traing exerise; in five previously treated patients, MDR-TB was exluded; and in seven patients for whom there was unertainty about the diagnosis, TB was onfirmed and appropriate treatment ontinued. In addition, MDR-TB was exluded in 26 patients who originated in TB endemi ountries. If NAAT had been sent from the other 36 patients in whih they were indiated, the subsequent ulture results suggest it ould have had an impat in 8 (22% (95% CI 11%, 38%)) by deteting one further ase of MDR-TB, exluding MDR-TB in two patients previously treated, onfirming TB in two patients and exluding it in three for whom there was diagnosti unertainty. There ould have been further impat by exluding MDR-TB in 26 patients in whom there were risk fators. If NAAT had also been sent for the 36 patients with AFB positive samples, in whom it was not indiated by the British guidelines, the subsequent ulture results suggest that they ould have had a linial influene by stopping unneessary treatment in 5/35 (14% (95% CI 5%, 29%)) patients who did not have TB and a potential linial impat by providing a rapid onfirmation of TB in a further 19 patients and exluding the diagnosis in 12. Impat of the multidisiplinary group The multidisiplinary group was onvened at the start of During the study period, 132/282 (47%) patients with a myobaterium seen or isolated were managed by the multidisiplinary group. NAAT were indiated and performed in 45/ 66 (68% (95% CI 56%, 79%)) AFB positive samples managed by the multidisiplinary group and 6/21 (29% (95% CI 12%, 50%)) managed by other physiians (p = 0.001). Of the 181 patients diagnosed with TB by ulture or NAAT during the study period, 37 (20%) were HIV positive, 78 (43%) were HIV negative and in 66 (36%) the HIV status was unknown. A HIV test was performed for 95/111 (86% (95% CI 78%, 91%)) of TB ulture positive patients managed by the multidisiplinary group but only for 20/70 (29% (95% CI 19%, 40%)) patients managed by the other physiians (p,0.001) DISCUSSION In our region, a relatively high proportion of environmental myobaterial isolates an onfuse the linial piture and ompliate deisions about empirial treatment and ontattraing. In this low prevalene setting for MTB, the sensitivity and speifiity of the NAAT were onsistent with other studies and the NAAT had a diret impat on the linial management of one-third of AFB smear positive patients in whom it was used. This inluded patients for whom treatment was hanged and situations in whih ontat-traing exerises were either ommened or stopped. In addition, the NAAT result was of value in onfirming that the orret therapy had been instituted in other patients. Usage of NAAT signifiantly dereased the time to identifiation of the myobateria and detetion or exlusion of rifampiin resistane, allowing prompt alterations in management when needed. This study shows that the urrent British guidelines for the use of NAAT in smear positive samples provide a high positive preditive value for deteting MTB and MDR strains. Our data also suggest that extending the indiations for performing NAAT on AFB positive samples to inlude all smear positive samples in low prevalene settings would have additional linial benefit. In partiular, in an area where more than onethird of the myobateria isolated are not M tuberulosis, 11 in the appropriate linial irumstanes rapid exlusion of TB allows patients with AFB positive samples to stop antituberulous treatment. The results of this study ontrast with those from a London teahing hospital where the impat of NAAT was onsidered to be minimal. 16 Firstly, this previous study did not look at the benefits of NAAT for the detetion of rifampiin resistane. Furthermore, in that setting, an additional two of 19 patients were started on treatment when the NAAT result beame available. AFB negative samples were also examined by NAAT but there was no evidene that liniians were using negative NAAT results to stop treatment. It is possible that the proportion of MTB in AFB smear positive patients is higher in London ompared with our setting. If the prior probability that a patient with an AFB positive sample atually has TB is low, as in our patients, the value of NAAT is inreased, as our data demonstrate. A ommon reason for not sending a sample for NAAT was the need for an undeontaminated sample, requested to redue the risk of potential ontamination from the referring laboratory. On some oasions, however, it was diffiult to obtain a further sample; even repeat sputum samples posed logisti diffiulties at times. Furthermore, the proess of determining whether a patient fulfills the riteria to send a sample an introdue additional delay mitigated in our study by the multidisiplinary group. Our data imply that the usage of NAAT would be improved and simplified by sending all AFB positive samples and also by allowing a deontaminated sample to be sent if an undeontaminated sample is not quikly and easily obtainable. One indiation for sending a sample for NAAT is knowledge of the patient s HIV status. 2 5 This is also ritial for the appropriate management of patients. However, HIV status was not determined for 23% of the AFB positive patients in whom NAAT were indiated or for 75% of the AFB positive patients in whom NAAT were not indiated. Overall, HIV tests were not done for one-third of patients with onfirmed TB. This is an important gap in the assessment of patients with TB and may reflet a lak of onfidene among some physiians in dealing with HIV related issues. The introdution of a multidisiplinary group in our hospital led to a signifiant inrease in HIV testing among patients who were AFB smear positive or with onfirmed TB. Formal osting of the inreasing use of NAAT has not been undertaken as part of this study but existing ost effetiveness studies from the UK and overseas indiate that the use of NAAT is overall more ost effetive when averted drug treatment, inpatient hospital stays and further investigations are aounted for The three ases where NAAT results differed from the final identifiation or sensitivity and the need for a full sensitivity profile imply that standard diagnosti methods still need to be performed. CONCLUSIONS There have been signifiant linial benefits from the use of NAAT in this low prevalene setting. The urrent British guidelines for the use of NAAT have a high positive preditive value when implemented orretly and our data indiate that there would be additional linial benefit from NAAT being applied to all smear positive linial samples. A multidisiplinary Thorax: first published as /thx on 16 November Downloaded from on 1 September 2018 by guest. Proteted by opyright. 320 Thorax 2008;63: doi: /thx

5 Tuberulosis What is already known about this topi What this study adds Nulei aid amplifiation tests allow rapid identifiation of Myobaterium tuberulosis, and detetion of rifampiin resistane, in AFB positive linial samples. British guidelines inlude indiations for when to use NAAT in linial pratie. The appropriateness and linial impat of these guidelines is not known. A NAAT result had a signifiant linial impat in 39% of the patients with AFB positive samples in whih the test was indiated and sent. The use of NAAT in all AFB positive samples would have additional linial benefits. A signifiant number of patients with tuberulosis are still not tested for HIV. Weekly review of new myobaterial diagnoses by a multidisiplinary group may lead to an inrease in the appropriate use of NAAT and an improved uptake of HIV testing. team led to inreased appropriate use of NAAT and sreening of patients for HIV infetion. Aknowledgements: The staff of the Royal Liverpool University Hospital Department of Mirobiology, and the Regional Centre of Myobateriology, Newastle and the United Kingdom Health Protetion Ageny National Myobaterial Referene Unit. We are grateful for the omments on the manusript by Dr Timothy Brown, Health Protetion Ageny National Myobaterial Referene Unit and Dr Rihard Cooke, University Hospital Aintree, Liverpool, UK, Competing interests: None. Ethis approval: The study is a registered hospital audit and ethial approval was not required. REFERENCES 1. Health Protetion Ageny. Tuberulosis: latest epidemiologial data (aessed 6 February 2008). 2. National Collaborating Centre for Chroni Conditions. Tuberulosis: linial diagnosis and management of tuberulosis and measures for its prevention and ontrol. London: Royal College of Physiians, Caws M, Drobniewski FA. Moleular tehniques in the diagnosis of Myobaterium tuberulosis and the detetion of drug resistane. Ann N Y Aad Si 2001;953: Drobniewski FA, Hoffner S, Rush-Gerdes S, et al. Reommended standards for modern tuberulosis laboratory servies in Europe. Eur Respir J 2006;28: Control and prevention of tuberulosis in the United Kingdom: ode of pratie Joint Tuberulosis Committee of the British Thorai Soiety. Thorax 2000;55: Aziz MA, Wright A, Laszlo A, et al. Epidemiology of antituberulosis drug resistane (the Global Projet on Anti-tuberulosis Drug Resistane Surveillane): an updated analysis. Lanet 2006;368: Gandhi NR, Moll A, Sturm AW, et al. Extensively drug-resistant tuberulosis as a ause of death in patients o-infeted with tuberulosis and HIV in a rural area of South Afria Lanet 2006;368: Telenti A, Imboden P, Marhesi F, et al. Detetion of rifampiin-resistane mutations in Myobaterium tuberulosis. Lanet 1993;341: Goyal M, Shaw RJ, Banerjee DK, et al. Rapid detetion of multidrug-resistant tuberulosis. Eur Respir J 1997;10: Drobniewski FA. Diagnosing multidrug resistant tuberulosis in Britain. Clinial suspiion should drive rapid diagnosis. BMJ 1998;317: Corless JA, Stokton PA, Davies PD. Myobaterial ulture results of smear-positive patients with suspeted pulmonary tuberulosis in Liverpool. Eur Respir J 2000;16: Antoine D, Maguire H, Story A. Epidemiology and response to the growing problem of tuberulosis in London. Euro Surveill 2006;11: Home Offie. Immigration and Asylum At. London: The Stationery Offie, (aessed 6 February 2008). 14. Sam IC, Drobniewski F, More P, et al. Myobaterium tuberulosis and rifampin resistane, United Kingdom. Emerg Infet Dis 2006;12: Department of Health. The Caldiott Committee Report on the Review of Patient- Identifiable Information. London: Department of Health, Conaty SJ, Claxton AP, Enoh DA, et al. The interpretation of nulei aid amplifiation tests for tuberulosis: do rapid tests hange treatment deisions? J Infet 2005;50: NHS Health Tehnology Assessment Programme. The linial and osteffetiveness of diagnosti tests for the detetion of myobaterial infetion, (aessed 6 February 2008). 18. Drobniewski FA, Watterson SA, Wilson SM, et al. A linial, mirobiologial and eonomi analysis of a national servie for the rapid moleular diagnosis of tuberulosis and rifampiin resistane in Myobaterium tuberulosis. J Med Mirobiol 2000;49: van Cleeff M, Kivihya-Ndugga L, Githui W, et al. Cost-effetiveness of polymerase hain reation versus Ziehl Neelsen smear mirosopy for diagnosis of tuberulosis in Kenya. Int J Tuber Lung Dis 2005;9: Thorax: first published as /thx on 16 November Downloaded from on 1 September 2018 by guest. Proteted by opyright. Thorax 2008;63: doi: /thx