Performance of low-cost monitors to assess household air pollution

• Published in: Environmental research Vol. 163; pp. 53 - 63
• Main Authors: Curto, A., Donaire-Gonzalez, D., Barrera-Gómez, J., Marshall, J.D., Nieuwenhuijsen, M.J., Wellenius, G.A., Tonne, C.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Inc 01-05-2018; Elsevier
• Subjects: AIR POLLUTION; AIR POLLUTION MONITORS; AIR QUALITY; Air quality sensing; CARBON MONOXIDE; COMBUSTION; DISEASE INCIDENCE; ENVIRONMENTAL SCIENCES; Household air pollution; HOUSEHOLDS; INDIA; Low cost instrument; MORTALITY; Particulate matter; PARTICULATES; SENSORS; SPAIN; VALIDATION; Air quality sensing; Carbon monoxide; Low cost instrument; Household air pollution; Particulate matter
• ISSN: 0013-9351; 1096-0953
• DOI: 10.1016/j.envres.2018.01.024

Exposure to household air pollution is a leading cause of morbidity and mortality globally. However, due to the lack of validated low-cost monitors with long-lasting batteries in indoor environments, most epidemiologic studies use self-reported data or short-term household air pollution assessments as proxies of long-term exposure. We evaluated the performance of three low-cost monitors measuring fine particulate matter (PM2.5) and carbon monoxide (CO) in a wood-combustion experiment conducted in one household of Spain for 5 days (including the co-location of 2 units of HAPEX and 3 units of TZOA-R for PM2.5 and 3 units of EL-USB-CO for CO; a total of 40 unit-days). We used Spearman correlation (ρ) and Concordance Correlation Coefficient (CCC) to assess accuracy of low-cost monitors versus equivalent research-grade devices. We also conducted a field study in India for 1 week (including HAPEX in 3 households and EL-USB-CO in 4 households; a total of 49 unit-days). Correlation and agreement at 5-min were moderate-high for one unit of HAPEX (ρ = 0.73 / CCC = 0.59), for one unit of TZOA-R (ρ = 0.89 / CCC = 0.62) and for three units of EL-USB-CO (ρ = 0.82–0.89 / CCC = 0.66–0.91) in Spain, although the failure or malfunction rate among low-cost units was high in both settings (60% of unit-days in Spain and 43% in India). Low-cost monitors tested here are not yet ready to replace more established exposure assessment methods in long-term household air pollution epidemiologic studies. More field validation is needed to assess evolving sensors and monitors with application to health studies.
•We assessed PM2.5 and CO low-cost monitors in semi-controlled and field studies.•Devices experienced high failure rate and susceptibility to environmental changes.•Low-cost monitors tested may not be ready to replace more established methods.