Using samples obtained by bronchoscopy to decide how to treat lung infections in people with cystic fibrosis

Key messages

We found evidence from two small studies that only included children.

We do not know the best way to get samples that guide which antibiotics should be prescribed for a lung infection.

What is the background to the review?

CF is a genetic condition that causes damage to the lungs, digestive system and other organs. Breathing problems in people with CF are mainly due to repeated lung infections. Growing bugs from samples of mucus coughed up from the lower airways can allow doctors to quickly identify the bug causing the infection and start treatment early. If people can not cough up mucus, swabs are taken from the upper throat to identify the bug causing infection in the lower airways, but this may not be the most reliable method.

During a bronchoscopy, doctors examine the lower airways using a long, thin flexible tube with a light and camera at one end; they may also collect mucus. The patient needs to be sedated or put under general anaesthetic. We do not know if treatment based on samples taken during a bronchoscopy is better than treatment based on throat swabs.

This review updates a review first published in 2013.

What did we want to find out?

We wanted to find out whether to use samples obtained by bronchoscopy when deciding how to treat lung infections in people with cystic fibrosis (CF).

What did we do?

We searched for studies involving people of any age with CF. We summarised the results of the studies and rated our confidence in the evidence, based on factors such as study methods and sizes.

What did we find?

We included two studies in this review, which involved 186 children. One study recruited 170 babies under the age of six months who were diagnosed with CF and monitored them for up to five years. The second study included 30 children with CF between the ages of five and 18 years and monitored them for up to two years. In both studies, the participants were divided into two groups completely at random, with one group given antibiotics based on samples taken via bronchoscopy and the other group given antibiotics based on samples taken from the throat. In the larger study, children had a bronchoscopy when they showed features of worsening chest symptoms, whereas children in the smaller study had a bronchoscopy when they showed an increase in the Lung Clearance Index, which is a marker of early lung disease that measures how well the air moves within the lungs.

Main results

Neither study showed any difference between the groups in terms of lung function, weight, body mass index or in the score calculated by a computed tomography (CT) scan of the lungs when measured at either two or five years. The longer study also showed that there was no difference in how many children in each group had an infection with Pseudomonas aeruginosa at five years of age, or per year of follow-up, or how often a child was unwell with respiratory symptoms. In this longer study, children in the bronchoscopy group were admitted to hospital more often, although admissions were generally shorter than in the comparison group. There was no difference between the groups in the overall cost of care and health-related quality of life scores.

Side effects reported during and after bronchoscopy were not serious; the most common side effect was increased coughing (in one third of children).

There is currently not enough evidence to support the regular use of bronchoscopy to diagnose and treat lung infections in children with CF.

What are the limitations of the evidence?

Evidence was limited to two studies. While these studies were well designed, we could not combine their results because they used different methods of measurement. We have low confidence in the reliability of the evidence for most outcomes, and moderate confidence in the results for two outcomes (high-resolution computed tomography scoring and cost-of-care analysis). We cannot be sure about the results because of the small number of children included. Also, the studies only included young children, so we do not know if the results would be the same in adults with CF. Since the treatment of a first infection with Pseudomonas aeruginosa is highly successful, larger and longer studies are needed to detect small differences between the groups. Conducting such large studies is extremely difficult and more sensitive measures should be used.

How up to date is this evidence?

We last searched for evidence in November 2023.

Authors' conclusions: 

This review, limited to two well-designed randomised controlled studies, shows no evidence to support the routine use of BAL for the diagnosis and management of pulmonary infection in preschool children with CF compared to the standard practice of providing treatment based on results of oropharyngeal culture and clinical symptoms. No evidence is available for adults.

Read the full abstract...

Early diagnosis and treatment of lower respiratory tract infections is the mainstay of management of lung disease in cystic fibrosis (CF). When sputum samples are unavailable, diagnosis relies mainly on cultures from oropharyngeal specimens; however, there are concerns about whether this approach is sensitive enough to identify lower respiratory organisms.

Bronchoscopy and related procedures such as bronchoalveolar lavage (BAL) are invasive but allow the collection of lower respiratory specimens from non-sputum producers. Cultures of bronchoscopic specimens provide a higher yield of organisms compared to those from oropharyngeal specimens. Regular use of bronchoscopy and related procedures may increase the accuracy of diagnosis of lower respiratory tract infections and improve the selection of antimicrobials, which may lead to clinical benefits.

This is an update of a previous review that was first published in 2013 and was updated in 2016 and in 2018.


To evaluate the use of bronchoscopy-guided (also known as bronchoscopy-directed) antimicrobial therapy in the management of lung infection in adults and children with cystic fibrosis.

Search strategy: 

We searched the Cochrane Cystic Fibrosis Trials Register, compiled from electronic database searches and handsearching of journals and conference abstract books. We also searched three registries of ongoing studies and the reference lists of relevant articles and reviews. The date of the most recent searches was 1 November 2023.

Selection criteria: 

We included randomised controlled studies involving people of any age with CF that compared the outcomes of antimicrobial therapies guided by the results of bronchoscopy (and related procedures) versus those guided by any other type of sampling (e.g. cultures from sputum, throat swab and cough swab).

Data collection and analysis: 

Two review authors independently selected studies, assessed their risk of bias and extracted data. We contacted study investigators for further information when required. We assessed the certainty of the evidence using the GRADE criteria.

Main results: 

We included two studies in this updated review.

One study enrolled 170 infants under six months of age who had been diagnosed with CF through newborn screening. Participants were followed until they were five years old, and data were available for 157 children. The study compared outcomes for pulmonary exacerbations following treatment directed by BAL versus standard treatment based on clinical features and oropharyngeal cultures.

The second study enrolled 30 children with CF aged between five and 18 years and randomised participants to receive treatment based on microbiological results of BAL triggered by an increase in lung clearance index (LCI) of at least one unit above baseline or to receive standard treatment based on microbiological results of oropharyngeal samples collected when participants were symptomatic.

We judged both studies to have a low risk of bias across most domains, although the risk of bias for allocation concealment and selective reporting was unclear in the smaller study. In the larger study, the statistical power to detect a significant difference in the prevalence of Pseudomonas aeruginosa was low because Pseudomonas aeruginosa isolation in BAL samples at five years of age in both groups were much lower than the expected rate that was used for the power calculation. We graded the certainty of evidence for the key outcomes as low, other than for high-resolution computed tomography scoring and cost-of-care analysis, which we graded as moderate certainty.

Both studies reported similar outcomes, but meta-analysis was not possible due to different ways of measuring the outcomes and different indications for the use of BAL.

Whether antimicrobial therapy is directed by the use of BAL or standard care may make little or no difference in lung function z scores after two years (n = 29) as measured by the change from baseline in LCI and forced expiratory volume in one second (FEV1) (low-certainty evidence). At five years, the larger study found little or no difference between groups in absolute FEV1 z score or forced vital capacity (FVC) (low-certainty evidence). BAL-directed therapy probably makes little or no difference to any measure of chest scores assessed by computed tomography (CT) scan at either two or five years (different measures used in the two studies; moderate-certainty evidence). BAL-directed therapy may make little or no difference in nutritional parameters or in the number of positive isolates of P aeruginosa per participant per year, but may lead to more hospitalisations per year (1 study, 157 participants; low-certainty evidence). There is probably no difference in average cost of care per participant (either for hospitalisations or total costs) at five years between BAL-directed therapy and standard care (1 study, 157 participants; moderate-certainty evidence).

We found no difference in health-related quality of life between BAL-directed therapy and standard care at either two or five years, and the larger study found no difference in the number of isolates of Pseudomonas aeruginosa per child per year. The eradication rate following one or two courses of eradication treatment and the number of pulmonary exacerbations were comparable in the two groups. Mild adverse events, when reported, were generally well tolerated. The most common adverse event reported was transient worsening of cough after 29% of procedures. Significant clinical deterioration was documented during or within 24 hours of BAL in 4.8% of procedures.