Table 1.

Inclusion and exclusion criteria [12].

Criterion	Inclusion and exclusion criteria
Population	We included surveys on general populations of adults (typically ≥18 years but used different thresholds according to age of majority which may vary in different countries). Surveys that included both adults and children were included but those that focused entirely on children were excluded. We excluded surveys on populations with known CRDs or respiratory diseases symptoms (for example: attendees at a respiratory clinic).
Screening procedure	We included surveys that determined the prevalence of asthma, COPD or other CRD using questionnaires, clinical examination, spirometry and/or other tests. We also included the prevalence of chronic respiratory symptoms and phenotypes.
Disease definitions	We included surveys that used definitions of CRD from globally recognised guidelines: asthma [14], COPD [15] or other CRD [7]. We defined ‘chronic’ respiratory symptoms as symptoms (such as cough, wheezing and shortness of breath) that have persisted for more than three months, or recurred in ‘attacks’. We did not include surveys on acute respiratory conditions such as pneumonia or active TB, therapeutic interventions, pharmaco-economics/cost analyses of medication or specific treatments, quality of disease management, assessment of inhaler technique, comparison between drug regimens, and health economic analyses (though we included prevalence studies that included assessments of socio-economic burden (eg, CRD-related time off work).
Burden of disease	We included population-level surveys of symptom burden, use of health care resources or societal burden (eg, absenteeism, loss of earnings).
Phenotypes	We included surveys that detected phenotypes of asthma, COPD or the overlap between these conditions.
Setting	We focused our review on low- or middle-income countries (LMICs) classified by the Organisation for Economic Cooperation and Development at the time of the survey. We included surveys in high-income countries only if the survey was also conducted in LMICs, eg, the BOLD study [16].
Study design	We included population or community surveys that aimed to determine the prevalence of one or more CRDs. The survey procedures included questionnaires, clinical examination, lung function tests (spirometry) or other tests (skin prick tests). We excluded randomised controlled trials, case-control studies and systematic reviews.