Fig 1.
Flow diagram of progress of clusters (shops) through phases of the randomized trial.
The diagram also shows the number of individual PID TVOC workshifts and the number of summa canisters (excluding duplicates and blanks) at each assessment in each group.
Table 1.
Baseline shop characteristics for auto and beauty shops that participated, by sector and intervention group.
Table 2.
Baseline demographics for auto and beauty shop workers who used a PID monitor to measure TVOCs, by sector and intervention group.
Fig 2.
TWA TVOC concentrations for each workshift for each shop at each assessment.
Data for each workshift at a shop have been placed on the same vertical line. Auto shop data are shown in the top three plots, and beauty shop data are shown in the bottom three plots. Data from shops in the immediate intervention group are on the left side of each plot, and data from shops in the delayed intervention group are on the right.
Table 3.
Estimated intervention effects along with their 95% CIs from the unadjusted and adjusted mixed models for auto and beauty shops.
Fig 3.
Hazard scores versus average shop TVOC concentrations at each assessment.
The average shop TVOC concentrations were calculated by taking the geometric mean of the TWA TVOCs for each workshift of that shop for that assessment. Regression lines with 95% CIs are shown for each sector at each assessment.
Fig 4.
Average shop TVOCs versus air exchange rate for beauty shops.
Each datapoint in this plot represents TVOC and air exchange rate data at one beauty shop at each assessment: shop TVOCs are the geometric average of the TWA TVOCs for each workshift of that shop for each assessment; air exchange rates are the average of the start- and end-of-day air exchange rates at each shop assessment. Regression lines with 95% CIs are shown for each group. Grey vertical dashed lines at the minimum recommended 4-6 ACH for beauty shops are shown for reference.