Relationship of incidence of otitis media to weather, air pollutants, airborne allergens, and upper respiratory tract infection incidence

Abstract

Study objectives: We study the effects of air pollutants, weather, airborne allergens, incidence of upper respiratory tract infection (URI), and the September 11, 2001, attack on the incidence of otitis media in the 11 public hospitals and 6 clinics run by the New York City Health and Hospital Corporation (HHC). Methods: This was a retrospective study of electronic records of all patients treated in 11 HHC Hospitals and 5 clinics in New York City with a diagnosis of otitis media. We obtained airborne pollutant levels from the Environmental Protection Agency, weather information from the National Weather Service, and otitis media incidence from the HHC billing computers. With the conservative autoregressive integrated moving average (ARIMA) modeling, epiphenomena that vary reliably during the course of the year are less likely to be viewed as significant. Results: The study covered 4,322 days, beginning January 1, 1991. The fifth, 50th, and 95th percentiles for visits per day were 70, 180, and 343, respectively, with marked seasonal variations. The ARIMA response variable was the change in the number of visits for otitis media from the previous week. Autoregressive terms for days 1, 7, and 365 and moving average term for day 7 were all significant (all P P x ) effect ( P =.0290) predicts that increasing NO x levels by 10% would result in 512 additional otitis media visits per year. Temperature effects were somewhat complex. A 14-day decrease in minimum temperature was associated with increased otitis media visits ( P =.0146). The effect is dynamic. If the temperature increased by a constant 1 degree per year, our model predicts a decrease in otitis media visits of 92 visits per year. The September 11 attack was associated with a decrease in otitis media visits ( P =.0022 on September 11 and P =.0085 on September 18.) We did not find significant exacerbating effects of pharyngitis incidence, airborne allergens, or other pollutants. Conclusion: We speculate that the decrease in number of otitis media visits seen around September 11 was due to decreased mobility within New York City. Using conservative ARIMA modeling, we found a positive relationship between the number of otitis media visits and increasing NO x levels, decreasing ambient temperature, and increasing URI incidence.