Geographic variability of interstitial lung disease diagnoses and impact of air pollution on disease outcomes.

Background And Objective: Anecdotal reports suggest interstitial lung disease (ILD) phenotypes vary regionally. We aimed to assess geographic variability of ILD diagnoses and impact of air quality on disease outcomes across the state of New South Wales, Australia.

Methods: Consecutive patients referred to an ILD multidisciplinary meeting, receiving referrals from across NSW (Feb 2014-Feb 2017), were included. Comparative frequencies of ILD diagnoses between regions were compared using Fisher's exact tests. Satellite-based land use regression models were used to estimate mean annual air pollution exposure for patients' home address at diagnosis. Associations between air pollution exposure and mortality and disease progression were assessed using multivariable Cox proportional hazard models.

Results: 437 ILD patients [mean age 67 ± 13 years; 41 % female, 48 % ever-smokers] were included. Air pollution at the year of diagnosis was 6.6 ± 2.0 μg/m for PM2.5, and 8.1 ± 4.2 ppb for NO. Exposure to NO was significantly higher in major cities than in regional areas (p < 0.001), while exposure to PM2.5 did not differ significantly (p = 0.373). In regional and remote areas, relative frequency of hypersensitivity pneumonitis (HP) was 1.75 times higher (p = 0.078) than in major cities. Among 329 ILD patients with ≥6 months follow-up data, disease outcomes were associated with specific ILD diagnosis and baseline FVC, but not associated with air pollution exposure.

Conclusion: We found a trend towards higher relative frequency of HP and lower relative frequency of non-IPF IIP in regional and remote areas compared to major cities. There was no association between mean air pollution exposure at diagnosis and disease outcomes.