Assessing Total Fungal Concentrations on Commercial Passenger Aircraft Using Mixed-Effects Modeling

Abstract

The primary objective of this study was to compare airborne fungal concentrations onboard commercial passenger aircraft at various in-flight times with concentrations measured inside and outside airport terminals. A secondary objective was to investigate the use of mixed-effects modeling of repeat measures from multiple sampling intervals and locations. Sequential triplicate culturable and total spore samples were collected on wide-body commercial passenger aircraft (n = 12) in the front and rear of coach class during six sampling intervals: boarding, midclimb, early cruise, midcruise, late cruise, and deplaning. Comparison samples were collected inside and outside airport terminals at the origin and destination cities. The MIXED procedure in SAS was used to model the mean and the covariance matrix of the natural log transformed fungal concentrations. Five covariance structures were tested to determine the appropriate models for analysis. Fixed effects considered included the sampling interval and, for samples obtained onboard the aircraft, location (front/rear of coach section), occupancy rate, and carbon dioxide concentrations. Overall, both total culturable and total spore fungal concentrations were low while the aircraft were in flight. No statistical difference was observed between measurements made in the front and rear sections of the coach cabin for either culturable or total spore concentrations. Both culturable and total spore concentrations were significantly higher outside the airport terminal compared with inside the airport terminal (p-value < 0.0001) and inside the aircraft (p-value < 0.0001). On the aircraft, the majority of total fungal exposure occurred during the boarding and deplaning processes, when the aircraft utilized ancillary ventilation and passenger activity was at its peak.

Keywords:

• aircraft
• culturable fungi
• mixed effects models
• total fungi
• total spore fungi

ACKNOWLEDGMENTS

This study was supported, in part, by an Interagency Agreement with the Federal Aviation Administration.

The authors gratefully acknowledge data collection support from Donald Booher, Brian Curwin, Kevin H. Dunn, Kevin L. Dunn, Belinda Johnson, John McKernan, and Nancy Nilsen; and the valuable field guidance from Tom Bloom, Jennifer Topmiller, Barbara Grajewski, Martha Waters, and Elizabeth Whelan. The authors appreciate the statistical guidance from Susan Woskie, Rebecca Gore, and Long Ngo. Finally, the authors acknowledge the valuable contributions of Jonathan Burstein, James Deddens, Donald Milton, Christine Rogers, Teresa Schnorr, and Jack Spengler during data analysis and manuscript preparation.

The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the views of the National Institute for Occupational Safety and Health.

Notes

^A Y_ijk refers to the natural log transformed air level from flight i, at sampling interval j, and replicate k; cov(Y_ijk,Y_y'j′k′), the covariance matrix for Y, is given by ZGZ′+ R.

^B For Structure I, the RANDOM statement specifies that G is σ_b ²I, where I is the 1× 1 identity matrix. The SUBJECT = FLIGHT option in the REPEATED statement specifies that R is a block diagonal matrix with blocks that are determined by the flights. The GROUP = INTERVAL option in the REPEATED statement specifies that the block for a particular flight is also a block diagonal matrix where the sub-blocks are determined by the sampling intervals. With the options as specified here, the TYPE = CS option in the REPEATED statement specifies that the structure for each sub-block is compound symmetry, with different parameters (i.e., σ_j ² and φ_j) for each sub-block.

^A The minimum detectable concentration of airborne culturable fungi was approximately 8.8 CFU/m³ = 1 CFU / (113.4 L × 0.001 m³/L) for all samples except deplaning for which the minimum detectable concentration was approximately 17.6 CFU/m³ = 1 CFU / (56.8 L × 0.001 m³/L) based on median air volumes of 113.4 L and 56.8 L for nondeplaning and deplaning samples, respectively.

^B The minimum detectable concentration of airborne total spore count fungi is approximately 25.2 spores/m³ = 3.81 × 1 spore / (151.1 L × 0.001 m³/L) for all samples, based on a median air volume of 151.1 L and a median multiplication factor of 3.81.

^C Range is for samples positive for growth or spores.

^A Modeled variance of the natural log transformed fungi concentrations for samples taken in the specified sampling interval.

^B Modeled GM significantly less than outside origin airport terminal GM (Tukey-Kramer adjusted p-value < 0.001).

^C Model adjusts for location in coach cabin (front, rear).

^D Modeled GM significantly less than boarding and deplaning GMs (Tukey-Kramer adjusted p-value < 0.0001).

^E Modeled GM significantly less than outside destination airport terminal GM (Tukey-Kramer adjusted p-value < 0.0001).

^A Modeled variance of the natural log transformed fungi concentrations for samples taken in the specified sampling interval.

^B Modeled GM significantly less than outside origin airport terminal GM (Tukey-Kramer adjusted p-value < 0.0001).

^C Model adjusts for location in coach cabin (front, rear).

^D Modeled GM significantly higher than cruise GMs (Tukey-Kramer adjusted p-value < 0.01).

^E Modeled GM significantly higher than midclimb GM (Tukey-Kramer adjusted p-value < 0.01).

^F Modeled GM significantly less than outside destination airport terminal GM (Tukey-Kramer adjusted p-value < 0.0001).