AJTCCM VOL. 29 NO. 4 2023 165
Background. Bronchiectasis, once rarely encountered, appears to be increasing in prevalence in South Africa (SA) and globally. ere is a
lack of published data on non-cystic brosis (CF) bronchiectasis, specically in low- to middle-income countries, despite the high rates of
risk factors such as HIV, pulmonary tuberculosis, and other infections.
Objectives. Given this lack of data, to review the characteristics of adult patients with non-CF bronchiectasis at a tertiary academic hospital
in Johannesburg, SA.
Methods. is was a single-centre, retrospective record review, including all cases of non-CF bronchiectasis that were in the records of the
adult pulmonology clinic at Charlotte Maxeke Johannesburg Academic Hospital as of April 2017.
Results. ere were 197 patients, with a slight predominance of males, and the patients were generally young. e HIV rate was higher than
the national average (34.8% v. 13.7%), and the HIV-positive patients had a high TB prevalence (86.9%). Pseudomonas spp. were cultured from
sputum in 15.3% of cases. Fewer than half of the cohort had the diagnosis of bronchiectasis conrmed by high-resolution chest tomography.
Airway obstruction (forced expiratory volume in 1 second/forced vital capacity ratio <70%) was observed in 47.0% of patients. Treatment
with a short-acting beta-2-agonist was prescribed in 62.9%, a long-acting beta-2-agonist in 43.6% and inhaled corticosteroids in 51.3%.
Antibiotic therapy during exacerbations was used in 44.2%, mainly amoxycillin-clavulanate (66.7%).
Conclusion. While single centre and retrospective, this study adds to the data on non-CF bronchiectasis in sub-Saharan Africa and should
encourage further research to increase our understanding of adult non-CF bronchiectasis in SA.
Keywords. Adult, non-cystic brosis bronchiectasis, sub-Saharan Africa.
Afr J Thoracic Crit Care Med 2023;29(4):e1017. https://doi.org/10.7196/AJTCCM.2023.v29i4.1017
Bronchiectasis, rst reported in 1819 and once known as an orphan
disease, is becoming increasingly prevalent globally.[1-3] It is a clinical
syndrome characterised by both cough and sputum production
along with abnormal thickening and dilation of the bronchial wall
visible on lung imaging. e pathophysiology has been described
by Coles ‘vicious cycle’ hypothesis, where chronic neutrophil
inammation promotes damage to the mucosa, submucosa and
muscular components of the bronchial wall, eventually leading
to airway dilation.[4] is chronic inammation, associated with
hypersecretion of mucus, impairment of mucociliary clearance
mechanisms and subsequent bacterial colonisation, continually
perpetuates the cycle of airway damage.[3,5] High-resolution chest
tomography (HRCT) is considered the gold standard for imaging
diagnosis of bronchiectasis.[6] While it is possible to diagnose
bronchiectasis using a chest radiograph (CXR) alone, it is estimated
that ~10% of cases will be missed.[7]
Most published data on bronchiectasis come from high-income
countries, where patient demographics show that mainly older females
are aected, with ~40% having idiopathic disease. Other documented
causes include infection, immunodeciency, ciliary abnormalities and
allergic bronchopulmonary aspergillosis (ABPA).[3,8] e global burden
of bronchiectasis has been documented to be increasing, particularly
in the past two decades, as indicated in publications from Asia (China,
Singapore, Korea, India), Europe (UK, Spain, Germany, Israel) and the
USA.[2,3,9-11] Of all countries with published large-scale studies, only
India can be classed as a low- to middle-income country (LMIC).
Non-cystic brosis bronchiectasis: A single-centre retrospective
study in Johannesburg, South Africa
G Titus,1 MB ChB, MMed (Int), FCP (SA); S Hassanali,2 BSc, MBBS, MMed (Int);
C Feldman,1 MB BCh, DSc, DMed (honoris causa), PhD, FRCP, FCP (SA)
1 Division of Pulmonology, Department of Internal Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa
2 Section of Pulmonology Medicine, Department of Medicine, Aga Khan University Hospital, Nairobi, Kenya
Corresponding author: G Titus (drgjtit[email protected])
Study synopsis
What the study adds. is study adds to published data detailing the clinical characteristics of adult non-cystic brosis (CF) bronchiectasis
in low- and middle-income countries (LMICs).
Implications of the ndings. As a retrospective descriptive study, the ndings summarise the characteristics of adults with non-CF
bronchiectasis in a cohort from Johannesburg, South Africa. e ndings suggest that the characteristics of bronchiectasis in this region
appear to be similar in several ways to those in other LMICs, but quite dierent from those in the developed world.
166 AJTCCM VOL. 29 NO. 4 2023
ere is therefore a paucity of adult data in LMICs, despite a high
prevalence of risk factors for bronchiectasis, including frequent early
childhood infections and high rates of pulmonary tuberculosis (TB)
and HIV. A systematic review by Gao etal.,[9] which investigated the
aetiology of bronchiectasis in adults worldwide, identied a single
study from Africa, and the majority of cases in this African cohort
were either post-infective, most commonly TB, or idiopathic. In
2014, a branch of the European Respiratory Society partnered with
the Respiratory Research Network of India in order to establish a
bronchiectasis registry in India. is collaboration established what
is to our knowledge the first adult bronchiectasis registry in any
LMIC. e Indian ndings were described and compared with data
from a substudy of the European registry (the European Multicentre
Bronchiectasis Audit and Research Collaboration (EMBARC)
database), which included data from Europe and Israel.[3,11] e data
from India suggest a very dierent disease prole from that seen in
high-income countries. In particular, the Indian study identied a high
proportion of young male patients with bronchiectasis, oen caused
by previous TB.[3] A subsequent study of patients with post-TB lung
disease in Malawi showed that 44% of the patients had bronchiectasis
in one or more lobes, and 9.4% had one or more destroyed lobes
conrmed on HRCT imaging.[12] Currently, there appears to be a
single bronchiectasis registry in South Africa (SA) in paediatrics, and
there have apparently been no such initiatives in adults.[13]
Based on the above data, we wished to review the characteristics of
adult patients with non-cystic brosis (CF) bronchiectasis seen at a
tertiary academic hospital in Johannesburg, SA.
is was a single-centre, retrospective record review, including all
cases of non-CF bronchiectasis that were in the records of the adult
pulmonology clinic at Charlotte Maxeke Johannesburg Academic
Hospital as of April 2017. Inclusion criteria were adult patients ³18
years of age, with a diagnosis of non-CF-related bronchiectasis made at
the discretion of the treating physician. e study was approved by the
Human Research Ethics Committee (Medical) of the University of the
Witwatersrand (ref. no. M220771). A waiver of consent was approved
because the study was retrospective in nature with no direct patient
Patient records in the pulmonology clinic are duplicated, with one
le held in the patient record system at the hospital and a separate
copy of the le being held securely in the pulmonology department.
e latter les were retrieved, and various patient details were entered
into an electronic database, including demographic data, history of
smoking, occupation, possible risk factor exposure, exposure to or
previous diagnosis of TB, HIV status, putative cause according to the
treating physician, clinical presentation, investigations performed for
diagnostic purposes, radiographic features, results of lung function
testing, microbiology and treatment. e database was deidentied
and was handled only by the research team.
Lung function was assessed according to the European Community
of Coal and Steel reference equations. Categorical variables were
represented as frequencies and percentages. Continuous variables were
represented as means and standard deviations (SDs) when normally
distributed, or as medians and interquartile ranges (IQRs) when not
normally distributed. Statistical analysis consisted of summary and
descriptive statistics, as well as comparative analysis of categorical
data using the χ2 test, and was performed using Python 3.10 (Python
Soware Foundation, USA).
Overall, 197 patients (n=101 (51.2%) male) were enrolled in the study,
with a median (IQR) age of 49 (38 - 60) years. Fiy-two patients (26.4%)
reported that they were either current (n=12; 6.1%) or ex-smokers
(n=37; 18.8%); 3 were identied as smokers, but it was not specied in
the le whether they were current or ex-smokers. e remaining 145
patients either reported never smoking or had no recorded smoking
status. e date of rst bronchiectasis diagnosis was recorded in 196
patients. At the time the study ended, April 2017, the median duration
of bronchiectasis among the patients was 48.0 (26.9 - 88.5) months.
Body mass index (BMI) was recorded in 167 patients, with a median
value of 20.9 (18.5 - 24.8).
Reported occupations among the cohort (n=61) were diverse,
ranging from unemployment to managerial, and from oce work to
mining and pharmaceutical drug manufacturing, totalling 35 dierent
occupations. In the total cohort, there were 9 patients (4.6%) who
reported occupational exposure as a cause of their bronchiectasis.
e occupations considered by the patients to be associated with that
exposure were mining (n=2 patients), plumbing (n=2), cement work,
pharmaceutical drug manufacturing, chemical manufacturing, spray
painting and mechanic (n=1 each).
Table1 shows the frequency of the various radiological methods
used for conrmation of the diagnosis of bronchiectasis. While HRCT
and a CXR are well-known methods for diagnosing bronchiectasis,
bronchography is an older modality, not currently used, performed by
instilling contrast material via a bronchoscope.[14] Of note, only 86patients
(43.7%) had HRCT imaging done to conrm the diagnosis. Overall, 164
patients (83.2%) had a CXR; of these, 88 (44.6%) had only a CXR, 1 (0.5%)
had a CXR and a bronchogram, and the remaining 75 (38.1%) had both a
CXR and HRCT. Eleven patients (5.6%) had HRCT alone, and in 22 cases
(11.1%) there was no record of what imaging was done.
In the 86 HRCT records, the main distribution of disease was
recorded for 49 patients (57.0%). Twenty patients (40.8%) had upper-
lobe involvement, 9 (18.4%) had lower-lobe involvement, and 20
(40.8%) had both upper- and lower-lobe involvement. Specically for
the study population that had a previous diagnosis of TB, 30 records
existed. Twelve patients (40.0%) had upper-lobe involvement, 5 (16.7%)
had lower-lobe involvement, and the remaining 13 (43.3%) had both.
Table2 shows the frequency of reported aetiologies, based on
reporting in the le by the attending physicians. irty-eight patients
did not have a documented aetiology, while others had multiple
potential aetiologies listed. ere was a total of 187 reported aetiologies
for the 159 patients. e most frequent aetiology documented was TB
(n=144; 77.0%), followed by various other causes (n=21; 11.2%) and
then recurrent infection (n=10; 5.3%).
The 21 aetiologies that fell under ‘other causes’ included asthma,
second-hand smoke exposure, aspiration, cancer, epilepsy, essential
thrombocytosis, human papillomavirus infection, gastro-oesophageal
reflux, splenomegaly, kyphosis, porphyria, post-renal transplant,
mycetoma and sarcoidosis. Some of these reported aetiologies may
represent associated comorbid conditions rather than actual causes of
the bronchiectasis.
AJTCCM VOL. 29 NO. 4 2023 167
Of the total cohort, 144 patients (73.1%) had a previous diagnosis of
TB. ere was no consistent record in the patient les as to how the
diagnosis of TB had been made. For more than half of these patients
(n=81; 56.3%), the le indicated that they had had only one episode
of TB. However, recurrent episodes of TB were reported for some
patients: two episodes for 28 patients (19.4%), three for 4 (2.8%), four
for 4 (2.8%), and seven for 1 (0.7%). Twenty-six patients did not have
a record of the number of their TB episodes.
HIV status was documented for 175 patients, with 61 (34.9%)
being HIV positive. Of the HIV-positive patients, 53 (86.9%) had
TB documented as the aetiology of bronchiectasis. Conversely, these
53patients accounted for only 36.8% of the total TB cohort.
For the total cohort, a breakdown of patients into specific
classications based on lung function results is shown in Fig.1.
Of the 144 patients with a previous diagnosis of TB, it was possible to
categorise the ndings on lung function studies for only 127 (88.1%),
as some values were unrecorded. Of these 127 patients, 44 (34.6%) were
categorised as having a possible restrictive defect, 40 (31.5%) as having
obstruction with a low forced vital capacity (FVC), 20 (15.7%) as having
obstruction alone, and 23 (18.1%) as having normal lung function.[15]
e mean (SD) reversibility of the lung function (% change in forced
expiratory volume in 1 second (FEV1)) for the total cohort was 7.84%
We also assessed clinical ndings. e presence or absence of crackles
in the chest was documented in the records of 172 patients, in 101
(58.7%) of whom crackles were noted and in 71 (41.3%) of whom they
were not. e presence or absence of clubbing was documented in the
records of 157 patients, with clubbing present in 79 patients (50.3%)
and absent in 78 (49.7%). In the TB cohort, crackles were present in
71 patients (49.3%) and clubbing was present in 66 (45.8%). Patients
with a previous diagnosis of TB therefore accounted for 70.3% of the
documented crackles and 83.5% of the documented clubbing found
in the whole cohort.
Of the total cohort, 98 patients had at least one sputum
microbiology sample recorded in their les, taken during a period
of exacerbation. Of these 98 patients, 54 (55.1%) had no organism
cultured. e micro-organism cultured most oen was Pseudomonas
aeruginosa (n=15 patients; 15.3%), followed by mixed ora (n=11;
11.2%). Other organisms cultured were Mycobacterium tuberculosis
(n=4; 4.1%), Haemophilus inuenzae (n=3; 3.3%), Staphylococcus
aureus (n=3; 3.3%), Escherichia coli (n=2; 2.0%), Mycobacterium
avium-intracellulare (n=2; 2.0%), fungi (n=2; 2.0%), Moraxella
catarrhalis (n=1; 1.0%), Klebsiella pneumoniae (n=1; 1.0%) and human
parainuenza virus (n=1; 1.0%).
Exacerbations were recorded in 128 patients (64.9%), 26 (13.2%)
did not experience exacerbations, and the remaining 43 had no record
of presence or absence of exacerbations.
Of the total cohort, 124 patients (62.9%) were treated with a
short-acting beta-2-agonist and 86 (43.6%) with a long-acting beta-
2-agonist. Inhaled corticosteroid therapy was used in 101 patients
(51.3%). ese numbers add up to more than 197, as many patients
were on multiple forms of treatment.
Eighty-seven patients (44.2%) received antibiotic therapy during
exacerbations, mainly amoxycillin-clavulanate (n=58/87; 66.7%).
Patients with lung function data,
Total cohort,
n=88 (53%)
<70% (airway obstruction),
n=78 (47%)
≥80% (normal),
n=27 (16%)
<80% (possible restriction),
n=61 (37%)
35 - 49% (severe),
n=24 (14%)
60 - 69% (moderate),
n=13 (8%)
<35% (very severe),
n=18 (11%)
>70% (mild),
n=12 (7%)
50 - 59% (moderately severe),
n=11 (7%)
Fig.1. Lung function results and classication. (FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity; FEV1% = % change in FEV1.)
168 AJTCCM VOL. 29 NO. 4 2023
Use of amoxycillin-clavulanate and macrolide therapy during
exacerbations was recorded in 19 patients (9.6%). Use or non-use
of a cephalosporin along with a macrolide during exacerbations was
recorded in 36 of the total cohort, of whom 32 (88.9%) were noted
to have received such therapy. Prophylactic macrolide therapy was
recorded to have been used in 21 patients (10.7% of the total). No
information on macrolide use was recorded in 161 patient les.
Long-term oxygen therapy was used for 19 patients (9.6%), and
12 (6.1%) underwent surgery. Of these surgical procedures, 7 were
pneumonectomies, while in the remaining 5 the type of surgery was
not specied.
Use or non-use of postural drainage was recorded in 170 patients, of
whom 162 (95.3%) were noted to perform postural drainage.
e major ndings in this study of 197 patients diagnosed with and
being followed up for bronchiectasis at a tertiary academic hospital
in Johannesburg were that there was a slight predominance of males,
the patients were generally young, the HIV rate was higher than the
national population average, and by far the most common reported
cause was TB. Additionally, fewer than half of the patients had HRCT
imaging, which is considered the gold standard for diagnosis of
e slight predominance of males in the present study is similar
to that in the cohort from India, but markedly dierent from the
European study (Table3) and from an additional EMBARC study
from Europe, in which most patients were female.[3,11]
In terms of age, the present cohort was somewhat younger that
the cohort from India, and much younger than the European cohort.
However, the IQRs were very similar between the present cohort and
the Indian cohort, which suggests the two cohorts were generally
in a similar range, while the IQR in the European study was higher
(57-74), clearly suggesting a significantly higher age group for
patients in the European study.[3,11] e reason for the dierences in
age is uncertain, but may well relate to dierences in the aetiology of
the bronchiectasis in the dierent regions.
e BMI was recorded in 167 patients in the present study, with a
median value of 20.9. is gure was very similar to that of the Indian
cohort and lower than the European cohort.[3,11] While bronchiectasis
itself may be associated with a low BMI, which acts as a prognostic
factor,[1] the lower BMIs in the SA and Indian cohorts compared with
the European cohort may also be associated with the fact that TB is the
most common cause of bronchiectasis in the former regions.
e rates of smokers and ex-smokers in the present cohort and the
cohort from India were similar, while both were signicantly lower
than the reported rate of smokers in the European cohort.[3,11]
It should be noted that there is not yet clinical evidence of a direct
causal link between occupational exposure and bronchiectasis,
rather just associations. An advanced PubMed literature search for
articles linking or evaluating bronchiectasis and occupation returned
only two articles, one of which was a case study.[16,17] Bronchiectasis
therefore does not currently appear to be recognised as a work-
related disease.
A major nding in the present study was that fewer than half of
the patient cohort (n=86; 43.7%) appear to have had HRCT imaging
done to conrm the diagnosis of bronchiectasis. is is probably
because of resource constraints, which are also documented in other
countries. HRCT is considered expensive in Nigeria, largely owing
to poverty and lack of health insurance.[18] Additionally, relatively
few medical centres in SA are equipped with computed tomography
scanners. Ittherefore appears likely that there may be nancial
reasons for the low rate of HRCT imaging. One patient had had a
bronchogram as part of their imaging work-up, but this modality of
imaging has long been discontinued worldwide.
e literature supports the nding of predominantly upper-lung
involvement in post-TB bronchiectasis,[19,20] and although there
was signicant upper-lobe involvement in our cohort, there were
also a fair number of patients with lower-lobe and also mixed-lobe
Post-TB was the most commonly reported cause of bronchiectasis
in both the SA and Indian cohorts (71.1% and 35.5%, respectively),
and while aetiology was not reported at all in the European cohort
(Table3), TB was much less common in other European studies
(4.9%).(3,11) e high rate of post-TB aetiology documented in the
SA cohort is not surprising given the prevalence of TB in our setting.
Post-TB bronchiectasis has been documented as a distinct entity
with higher rates of haemoptysis and non-tuberculous microbacteria
isolation, more frequent exacerbations, and higher disease severity
scores.[10] Post-infective aetiology, but not TB-related, was reported
more frequently in the Indian cohort and in other European studies
(21.2%) compared with the lower rate in the SA cohort.[3,11] ABPA
was similarly reported at a higher rate in the Indian cohort (2.8%
in European studies).[3,11] Idiopathic disease was also reported at a
higher rate in the Indian cohort and in European studies (38.1%) than
in SA.[3,11] Other diverse causes were reported more frequently in SA
than in India or Europe (1.9%).[3,11] Chronic obstructive pulmonary
disease was reported in the Indian cohort and in European studies
(8.1%), but not at all in the SA cohort.[3,11] Conversely, autoimmune
disease and immunodeficiency were reported in the SA and
European studies (2.5% and 4.1%, respectively), but not at all in
Table1. Radiological methods used to document
bronchiectasis (N=197)
n (%)
HRCT 86 (43.7)
CXR only 88 (44.6)
CXR + bronchogram 1 (0.5)
Not recorded 22 (11.1)
HRCT = high-resolution computed tomography; CXR = chest radiograph.
Table2. Reported causes of non-cystic brosis
bronchiectasis* (N=159)
n (%)
Tuberculosis 144 (77.0)
Other causes 21 (11.2)
Recurrent infection 10 (5.3)
Allergic bronchopulmonary aspergillosis 6 (3.2)
Rheumatoid arthritis 3 (1.6)
Hypogammaglobulinaemia 3 (1.6)
*Sometimes multiple.
AJTCCM VOL. 29 NO. 4 2023 169
the Indian cohort.[3,11] Asthma as a cause was reported in the SA
and Indian cohorts at similar levels, but was reported to occur more
frequently in Europe (6.9%).[3,11]
With reference to the post-TB aetiology detailed above, it is
interesting to note the context of TB nationally in both SA and
India. The latest World Health Organization country data (for
2021) documented an estimated TB prevalence in SA of 513 (95%
condence interval (CI) 348 - 709) patients per 100 000 population.
For India, the prevalence was 210 (95% CI 178 - 244) per 100 000
population.[21] is ratio of 513 to 210 for national TB prevalences
(2.4:1) is very similar to the ratio of post-TB aetiology in the two
cohorts (2.0:1).
FEV1 % predicted values were dierent among the three cohorts.
e SA cohort had the lowest FEV1 % predicted (52.0) compared
with the cohorts from India (61.4) and Europe (73.8). When
comparing IQRs, it should be noted that the IQR for India completely
encompasses the IQR for the SA cohort.[3,11]
e HIV rate in the present cohort (34.8%) was higher than the national
population average (13.7%).[22] Additionally, the HIV cohort had a high
prevalence of documented TB (86.9%). Further statistical analysis
evaluating HIV status and TB aetiology indicated that they were linked
rather than independent variables (χ2 statistic = 8.01; two-sided p-value
= 0.009).
Microbiological data were available for all three cohorts (Table3). e
prevalence of P. aeruginosa was relatively similar between the cohorts.
e prevalence of M. catarrhalis was also similar in the SA and Indian
cohorts, but slightly higher in the European cohort. However, the other
organisms cultured diered among the cohorts. e SA population
had a higher prevalence of H. inuenzae than the Indian cohort, while
the highest prevalence was in the European cohort. Conversely, the SA
population had no reported cases of Enterobacteriaceae, whereas the
Indian and European cohorts had similar prevalences.[3,11]
e spirometry pattern in the SA cohort diered from the Indian
cohort, with a higher percentage of obstructive ndings with low FVC
Table3. Summary of comparative cohorts*
South African study
(the present study) Indian study European study
Age (years), median (IQR) 49 (38 - 60) 56 (41 - 66) 67 (57 - 74)
Male, % 51.2 56.9 38.9
BMI, median (IQR) 20.9 (18.5 - 24.8) 21.5 (18.5 - 24.5) 24.8 (21.8 - 28.1)
Current/ex-smokers, % 26.4 28.2 38.1
Functional status
FEV1 (% predicted) median (IQR) 52.0 (42.0 - 71.5) 61.4 (41.9 - 80.5) 73.8 (54.0 - 92.1)
Spirometry pattern, %
Obstruction 47.6 34.8 -
Low FVC 36.3 26.7 -
Normal 16.1 38.5 -
Microbiology, %
Pseudomonas aeruginosa 15.3 13.7 15.0
Haemophilus inuenzae 3.3% 0.5 21.9
Staphylococcus aureus 3.3 2.3 6.0
Moraxella catarrhalis 1.0 1.0 5.9
Enterobacteriaceae - 9.8 6.1
Treatment, %
Long-term macrolide 10.6 6.2 NR
Aetiology, %
Post-TB 71.1 35.5 NR
Idiopathic 0.5 21.4 NR
Post-infective (non-TB) 6.6 22.4 NR
ABPA 2.0 8.9 NR
COPD - 5.3 NR
Asthma 3.0 2.5 NR
Other 7.0 4.1 NR
Autoimmune 1.5 - NR
Immunodeciency 1.0 - NR
IQR = interquartile range; BMI = body mass index; FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity; TB = tuberculosis; ABPA = allergic bronchopulmonary aspergillosis;
COPD = chronic obstructive pulmonary disease; NR = not reported in the above European cohort – the data reported for aetiology of bronchiectasis in Europe, in the discussion section, are from the
European Bronchiectasis Registry (EMBARC study).[11]
*is table (Table2 in Dhar etal.[3]) is reproduced with permission from Lancet Global Health, under a Creative Commons License Deed (Attribution-NonCommercial-NoDerivatives 4.0 international
(CC BY-NC-ND 4.0)) (available from https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode). e table has been modied to include some further data from the Indian cohort, and data from the
present South African study have been added for comparison.
Including cases from Israel.
170 AJTCCM VOL. 29 NO. 4 2023
and a much lower percentage of spirometry classied as normal.[3,11]
e SA population had a higher prevalence of long-term macrolide
use for exacerbations than the Indian cohort.[3,11] e most common
antibiotic used for exacerbations in the SA setting was amoxycillin-
clavulanate. Others were amoxycillin-clavulanate plus a macrolide and
a cephalosporin plus a macrolide.
Table3 presents a summary of the above ndings in the present
cohort compared with those from India and Europe/Israel.
The present study does have some potential limitations. One
major limitation is that some patients appear to have been diagnosed
as having bronchiectasis without all the currently recommended
investigations being done. In this respect, some clinicians appear
to have made the diagnosis based on a CXR alone, without the
use of HRCT. While HRCT is currently the gold standard for the
diagnosis of bronchiectasis worldwide, it is oen possible to diagnose
bronchiectasis on CXR, although ~10% of cases will be missed without
HRCT.[7] It is known that most LMICs experience resource limitations,
so there is a nancial barrier to extensive investigations.[16,18] A second
limitation is that this was a single-centre study, so not all the ndings
can be extrapolated to other institutions and/or regions in SA. irdly,
this was a retrospective record review, and details for some of the
information were missing in the patient les. For example, there were
no records in the database indicating how the previous diagnoses of
TB, or some of the other diagnoses, had been made. Nevertheless, the
strength of this study is that it is the rst investigation of this sort of
bronchiectasis in adults in SA, and would be useful as a base on which
to plan future prospective studies in this region, and for improving the
overall management of bronchiectasis.
While we acknowledge that this was a retrospective single-centre
study, it is the rst of its kind on bronchiectasis in adults in SA. e
findings suggest that the characteristics of bronchiectasis in this
region appear to be similar in several ways to those in other LMICs,
but are quite dierent from those in the developed world. e study
could function as a template for planning a future prospective study
in this region, with the goal of improving the overall management of
Declaration. None.
Acknowledgements. None.
Author contributions. Data collection was performed by SH, and all
authors signicantly contributed to the analysis and the writing of the
Conicts of interest.None.
1. O’Donnell AE. Bronchiectasis – a clinical review. N Engl J Med 2022;387(6):533-545.
2. Martinez-Garcia MA, Polverino E, Aksamit T. Bronchiectasis and chronic airway
disease: It is not just about asthma and COPD. Chest 2018;154(4):737-739. https://
3. Dhar R, Singh S, Talwar D, etal. Bronchiectasis in India: Results from the European
Multicentre Bronchiectasis Audit and Research Collaboration (EMBARC)
and Respiratory Research Network of India Registry. Lancet Glob Health
2019;7(9):e1269-e1279. https://doi.org/10.1016/S2214-109X(19)30327-4
4. Cole PJ. Inammation: A two-edged sword – the model of bronchiectasis. Eur J Respir
Dis Suppl 1986;147:6-15.
5. McDonnell MJ, Ward C, Lordan JL, Rutherford RM. Non-cystic brosis bronchiectasis.
QJM 2013;106(8):709-715. https://doi.org/10.1093/qjmed/hct109
6. Feldman C. Bronchiectasis: New approaches to diagnosis and management. Clin
Chest Med 2011;32(3):535-546. https://doi.org/10.1016/j.ccm.2011.05.002
7. Smith MP. Diagnosis and management of bronchiectasis. CMAJ 2017;189(24):E828-E835.
8. Shoemark A, Ozerovitch L, Wilson R. Aetiology in adult patients with bronchiectasis.
Respir Med 2007;101(6):1163-11670. https://doi.org/10.1016/j.rmed.2006.11.008
9. Gao YH, Guan WJ, Liu SX, etal. Aetiology of bronchiectasis in adults: A systematic
literature review. Respirology 2016;21(8):1376-1383. https://doi.org/10.1111/resp.12832
10. Fong I, Low TB, Yii A. Characterisation of the post-tuberculous phenotype
of bronchiectasis: A real-world observational study. Chron Respir Dis
2022;19:14799731221098714. https://doi.org/10.1177/14799731221098714
11. Chalmers JD, Polverino E, Crichton ML, etal. Bronchiectasis in Europe: Data on
disease characteristics from the European Bronchiectasis registry (EMBARC). Lancet
Respir Med 2023;11(7):637-649. https://doi.org/10.1016/S2213-2600(23)00093-0
12. Meghji J, Lesosky M, Joekes E, etal. Patient outcomes associated with post-tuberculosis
lung damage in Malawi: A prospective cohort study. orax 2020;75(3):269-278.
13. Verwey C, Gray DM, Dangor Z, etal. Bronchiectasis in African children: Challenges
and barriers to care. Front Pediatr 2022;10:954608. https://doi.org/10.3389/
14. George RB. Chest Medicine: Essentials of Pulmonary and Critical Care Medicine. 5th
ed. Philadelphia: Lippincott Williams & Wilkins, 2005:xvii.
15. Koegelenberg CF, Swart F, Irusen EM. Guideline for oce spirometry in adults, 2012.
S Afr Med J 2012;103(1):52-62. https://doi.org/10.7196/SAMJ.6197
16. Bhatta N, Dhakal SS, Rizal S, Kralingen KW, Niessen L. Clinical spectrum of patients
presenting with bronchiectasis in Nepal: Evidence of linkage between tuberculosis,
tobacco smoking and toxic exposure to biomass smoke. Kathmandu Univ Med J
(KUMJ) 2008;6(2):195-203.
17. Charpin D, Jullian H, Garbe L, Cau P, Fuentes P, Vervolet D. [Severe bronchial stenosis
with upstream bronchiectasis in an arc welder: Causal relation or epiphenomenon?]
Rev Mal Respir 1997;14(2):137-139.
18. Adetiloye A, Erhabor G, Awopeju O, Adewole O, Onini E, Adewuya O. Challenges of
diagnosing and managing bronchiectasis in resource-limited settings: A case study.
Pan Afr Med J 2019;32:82. https://doi.org/10.11604/pamj.2019.32.82.18167
19. Choi H, Lee H, Ra SW, etal. Clinical characteristics of patients with post-tuberculosis
bronchiectasis: Findings from the KMBARC Registry. J Clin Med 2021;10(19):4542.
20. Martinez-Garcia MA, Guan WJ, De-la-Rosa D, etal. Post-TB bronchiectasis: From
pathogenesis to rehabilitation. Int J Tuberc Lung Dis 2023;27(3):175-181. https://doi.
21. Bagcchi S. WHO’s Global Tuberculosis Report 2022. Lancet Microbe 2023;4(1):e20.
22. Statistics South Africa. Mid-year population estimates 2021. 19 July 2021. https://
www.statssa.gov.za/publications/P0302/P03022021.pdf (accessed 1 June 2023).
Submiited 24 April 2023. Accepted 17 September 2023. Published 27 November 2023.