Background: Urinary Tract Infections (UTIs) are very common, especially among women. UTIs are seasonal with more cases in the summer and fewer cases in the winter. We examine the trends in admissions for UTI, both seasonal and non-seasonal, for men and women in different age groups.
Methods: Using the Healthcare Costs and Utilization National Inpatient Sample for 1998-2011, we apply time series and semiparametric Generalized Additive Mixed Models (GAMM) methods to estimate the trend and seasonal intensity of UTI inpatient discharges. The time series framework includes a linear effect for year and autoregressive coefficients, one at lag 1 to capture month-to-month variance and a second at lag 12 to capture seasonality. The GAMM estimates a smooth function for the data, and we included an AR1 correlation structure. Incidence was converted to an age and sex specific Z-score to enable comparison between the age groups and sexes.
Results: Hospitalizations with UTI as the primary diagnoses increased during this period for both men and women. In 1998 there were 136,326 cases and in 2011 there were 243,702 cases. (See Figure 1.) However, the rate of increase among men is less than half that of women (9.4% vs 19.1% of a standard deviation per year by time series modeling and 9.2% vs 19.6% of a standard deviation per year according to GAMM). These trends do not vary by age.
UTI rates for both men and women exhibit seasonality. However, among women the seasonal intensity diminishes with age. We found that the estimated seasonal coefficient from the time series model among women decreased by 0.002 (p < 0.0001) correlation units per year of age. In the GAMM framework, we found the range decreased by 1.72% (p = 0.0008) of a standard deviation per year of age. In other words, UTIs among young women are highly seasonal, but UTIs among older women do not vary by season.
Conclusion: Admissions for UTI are increasing, and admissions among women are increasing at twice the rate as among men. Additionally, we find that for women, seasonal intensity decreases with age. This finding suggests that the epidemiology and risk factors for women vary by age, especially with respect to seasonality. Future work is needed to understand the cause and reduction of seasonality of UTIs across age in women.
J. Cavanaugh, None
L. Polgreen, None
P. M. Polgreen, None