What is the association between air ventilation and pathogen transmission?

In a recent study posted to the medRxiv* preprint server, researchers assessed the association between indoor air respiratory pathogens and natural ventilation, carbon dioxide (CO2) levels, and air filtration.

Currently, little is known about the effects of indoor climate, human activity, ventilation, and air filtration on the detection and concentration of respiratory pathogens. The capacity to quantify bioaerosols in indoor air to monitor respiratory infections and transmission risk is hampered by this lack of knowledge and hence requires extensive research.

About the study

In the present study, researchers pinpoint the host, pathogen, behavioral, and environmental variables that are associated with a higher respiratory pathogen bioaerosol burden in indoor ambient air.

Between October 2021 and April 2022, the team collected 341 ambient air samples at 21 sampling locations. By treating the quantitative polymerase chain reaction (qPCR) results of each pathogen in a sample as a single observation, the team examined the independent impacts of a variety of variables on airborne pathogen detection and concentration. The generated models had within-sample correlation adjustments and included pathogen type as a covariate.

For each pathogen, the team ran a logistic regression model (LRM) with backward elimination using the retained independent variables from these models, which employed qPCR Ct values as the numerical results. For 24 days, beginning on February 7, air samples were collected simultaneously in the three nursery groups. Location 2 had three Blue PURE 221 filters installed, while location 1 had no air filtration, resulting in a total theoretical clean air delivery rate of 1770 m3/hour and 10.7 air changes per hour (ACH). Three Philips 3000i filters were placed at location 3, resulting in a total theoretical clean air delivery rate of 999 m3/hour and 6.1 air changes per hour.

Results

The study results showed that the most prevalent infections discovered when comparing the positivity rates of all samples were Streptococcus pneumoniae, human enterovirus, human bocavirus, human adenovirus, and human cytomegalovirus, in that order. The age range of three to six years had the largest percentage of samples that tested positive for at least one pathogen, followed by zero to three years, 25 to 65 years, 12 to 18 years, 18 to 25 years, six to 12 years, and more than 65 years.

In the nursery environment depicted, human bocavirus, human enterovirus, human cytomegalovirus, and Streptococcus pneumoniae were nearly always positive. Over the winter, the team noticed a long rise in human adenovirus and Pneumocystis jirovecii infections. Other viruses, such as enterovirus D68, influenza A virus, human parainfluenza virus 3, respiratory syncytial virus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), Human CoV 229E, human CoV HKU-1, human CoV OC43, enterovirus D68 exhibited shorter peaks. Variations in positive results for SARS-CoV-2, enterovirus D68, and influenza A virus matched findings from clinical samples in University Hospitals Leuven, located close to the sampling site.

Before grouping, 14 pathogens had at least 10 positive results, and 12 pathogens had at least 10 positive results. First, a LRM binary outcome for any respiratory pathogen positivity was used. Pathogen detection was shown to be inversely correlated with vocalization. Every 100 ppm increase in CO2 concentration raised the likelihood of finding a respiratory infection by 8.8%. As natural ventilation increased gradually, odds reduced by 11%. The generalized estimating equations (GEE) models and mixed logistic regression model (MLRM) that account for within-sample correlation both had equal significance levels and effect sizes.

Due to the smaller sample numbers, the power of these models was reduced. However, they still showed a significant correlation between mean CO2 and the detection of human enterovirus, other CoVs, Pneumocystis jiroveci, and Streptococcus pneumoniae. On the other hand, the team discovered a negative correlation with the discovery of human bocavirus. Furthermore, natural ventilation negatively correlated with the presence of respiratory syncytial virus and Pneumocystis jiroveci.

The qPCR Ct value declined by 0.13 for each 100 ppm rise in CO2. In contrast to the earlier analyses, natural ventilation was not substantially linked to concentration. However, there was a 0.57 increase in average Ct when air filtering was present, which was substantially linked with pathogen concentration. Also, performing an MLRM and when removing imputed values, significance levels and effect sizes were comparable.

A higher concentration of human adenovirus, human cytomegalovirus, human bocavirus, and Streptococcus pneumoniae remained favorably correlated with mean CO2. The respiratory syncytial virus was related to a lower concentration. Human bocavirus, human cytomegalovirus, various coronaviruses, and Streptococcus pneumoniae were all found in reduced amounts in the air after air filtration.

Conclusion

The study findings showed that the number of detected pathogens and their corresponding concentrations varied dramatically depending on the pathogen, month, and age group. Notably, low natural ventilation and high CO2 were independent risk factors for detection. Their concentration was independently correlated with both CO2 concentration and air filtration. The researchers supported the usage of air filtration and ventilation in reducing pathogen transmission.

*Important notice

medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.