Studies meeting the following criteria will be selected.
Study design
Studies investigating effects of industrial environmental air pollution or exposure in outdoor occupational settings on new-onset asthma and/or COPD often use cross-sectional study designs based on prevalence [
6,
13]. The practicability of cross-sectional analyses for investigations of newly suspected causative factors and the lower demand of resources compared to incidence-based analyses might cause this trend [
14]. Evidence from prevalence-based studies might be sensitive, because results for emerging and less studied threats in an earlier stage of risk-evaluation would be included. However, we define studies based on prevalence, respectively cross-sectional studies, lacking an observation period for the registration of new-onset disease, as ineligible. Consequently, case series, and case reports are also excluded.
The most appropriate observational study designs to indicate a causal relationship between exposure and effect are incidence-based [
14]. Therefore, we will include incidence studies (e.g., cohort studies) and incidence case-control studies [
14]. Inclusion of this higher level of evidence may lead to the most specific and sound results.
Systematic reviews and meta-analyses are ineligible. However, if relevant for the objective, we will keep them for the screening of reference lists.
Exposure
Airborne exposure to chemicals released from industrial processes is eligible. Examples are petrochemicals, polycyclic aromatic hydrocarbons (PAH), formaldehyde, chlorine, ammonia, nitric oxides, isocyanates, acid anhydrides and metals (metal salts). Man-made mineral matter (e.g., cement dust) is also included. Chemical pollution caused by fertilizers or processed organic matter, such as ammonia from manure is included. However, allergens, e.g., pollen, fungi, and the common allergens in workplaces (e.g., wheat flour, latex, animal dander and enzymes) are not eligible. Naturally occurring minerals, such as silica and asbestos, are also excluded. Radiation and exposure to chemical warfare agents in bellicose conflicts are ineligible. Particular matter (PM) of unspecified constituents and gaseous emissions are explicitly included. The definition of the term “industry” used in this systematic review is: “companies and activities involved in the process of producing goods for sale, especially in a factory or special area” [
15]. In this sense, the authors explicitly include the branches agriculture, construction, and electricity generation.
Environmental exposure is eligible when it is caused by the release of airborne pollutants from industrial processes into environmental air.
There are no limitations to the method of exposure assessment for eligibility. However, data based on actual measurements are desirable and of higher quality, i.e., personal air monitoring or ambient air monitoring. Studies can also determine exposure statuses based on geographic proximity to a known source of industrial air pollution. This would be a surrogate for measurement data, which relies on the assumption that quantities of pollutants are higher close to industrial sources and lower further away. Confirmation of nearby sources of pollution by questionnaire data is possible. Environmental modeling that can use different kinds of data with various methods is accepted. Occupational exposure is eligible when airborne substances in outdoor workplaces are also found in environmental air. Consequently, available studies on health effects in the environment determine the eligibility of studies on substances in the occupational field (e.g., formaldehyde).
Presence and/or concentration of exposure in the air of workplaces have to be monitored. Alternatively, the presence of substances in a workplace must be apparent, because exposure to a certain material of interest was verified under comparable conditions, e.g., asphalt paving in road construction. Biomonitoring is not obligatory.
If the exposure is mixed, e.g., dusts or complete emission of industrial origin, but contains eligible substances, the study will be included.
Household exposure is not defined as industrial in this review and therefore excluded. Examples are fuel aerosols released by domestic cooking as well as cooking in the streets, use of domestic chemicals and household products.
Traffic air pollution in public spaces is not included in the scope of this review, despite a share that might be attributable to industrial activities (e.g., transport of materials and products).
If a study reports effects of chemicals that usually enter the human organism in a different way than via the respiratory tract and respiratory exposure is not proven and mentioned (e.g., Bisphenol A), the study will be excluded.
We refer to exposure that is present in outdoor air and that is not enclosed resp. confined to compartments, e.g., in buildings, with the term “environmental exposure”. The term “environmental air” corresponds to outdoor air in the same sense throughout this protocol.
Comparison
For the estimation of measures of effect, cohort studies may compare a group of subjects exposed to industrial air pollution with a control-group of non-exposed subjects. Sometimes, a control-group consists of subjects exposed to lower levels of the investigated pollution than the exposed group. Case-control studies also require exposure statuses assigned to all subjects for the estimation of effect measures. Usually, the statuses are exposure and no exposure among both cases and controls. However, comparisons are not required for eligibility, i.e. if a study reports incidence. Moreover, with the broad range of potentially causal substances and methods of exposure assessment, controls in studies of interest may have different characteristics. Therefore, specific characteristics of the comparisons, e.g., certain cut-off values, are inconsiderable for eligibility. However, selection and comparability of controls are part of the quality assessment.
Diagnosis of the condition of interest
We determine the following criteria and references for the diagnosis of the conditions of interest new-onset asthma and COPD.
We determine studies that rely on diagnosis of asthma according to the ERS, ATS and/or GINA guidelines [
16‐
18] to be of high quality concerning diagnosis.
Diagnosis of COPD according to the leading international guidelines of the ERS, ATS and GOLD [
19,
20] for adults and children is likewise determined as high quality.
For example, a forced expiratory volume in 1 s over forced vital capacity ratio (FEV1/FVC) below the 5th percentile, the lower limit of normal (LLN) and the equivalent to a z-score of − 1.645, is the diagnostic criterion of ERS/ATS distinguishing between a healthy and potentially pathologic lung function [
16]. A post-bronchodilator FEV1/FVC below 70% of the predicted ratio is the limit for a diagnosis of COPD according to GOLD [
16,
19].
There may be particular studies that do not explicitly mention these references. In such cases, we accept reporting of the diagnostic measures and limits that the aforementioned guidelines demand as equivalent.
The reference of a particular study to a diagnostic standard will be traced to a previous publication if indicated.
Diagnosis of asthma and COPD by a physician is the minimum diagnostic criterion for inclusion in qualitative and/or quantitative analysis. Studies using data in accordance with the aforementioned criteria provided by registries or collected with questionnaires are explicitly eligible. Registries may use appropriate case definitions for asthma and COPD, and questionnaires may ask for diagnoses by a physician. Definitions of the outcomes that are based on symptoms alone are not accepted for inclusion.
Report characteristics
The language of included reports is English. Articles in other languages than English will not be processed.
Articles need to be published in indexed (MEDLINE, EMBASE or Web of Science) scientific journals and subjected to peer review.
We exclude grey literature (conference abstracts/proceedings, theses, etc.), letters to the editor, and unpublished data. If comments or letters to the editor refer to eligible studies, they will be included as adjuncts to the particular article.