View list of archived news items.

ISDS Annual Meeting, Raliegh, North Carolina, December 3-5, 2008

Dr. Steven Cohen and Julia Wenger of InForMID attended the 7thannual meeting of the International Society for Disease Surveillance. Together they delivered three presentations.

S.A. Cohen and E.N. Naumova. Accounting for acceleration of disease rates with age in biosurveillance systems: The SIMPLE Method

Abstract: Age is often one of the most important predictors of disease rates, and can be a potential confounder when estimating rates in disease surveillance.

Using the SIMPLE method (Slope-Intercept Method for Population Linear Estimation), we estimated age-specific pneumonia and influenza rates from Medicare hospitalization claims  and regressed log-transformed disease rates against age to obtain an intercept and slope parameters.

R squared values for states ranged from 0.71 to 0.98, and for divisions between 0.94 and 0.99.  The slope and intercept parameters are highly correlated to each other and to age-adjusted disease rates.

The SIMPLE method works best for diseases with high incidence and can provide useful information about disease dynamics in surveillance of the elderly population.

J.B. Wenger and E.N. Naumova. Seasonal Patterns of Respiratory Diseases: A Proxy for Influenza?

Abstract: Influenza surveillance is made difficult by inconsistent laboratory testing, deficiencies in testing techniques, and coding subjectivity in hospital records. We hypothesized that respiratory diseases may serve as a useful proxy for influenza in pediatric populations.

Hospital visits for respiratory diseases were extracted from billing records for all children admitted to Children’s Hospital of Wisconsin from 1997 to 2006 for primary diagnoses of bronchitis, pneumonia, other upper respiratory infections (URI), and influenza.
Respiratory diseases peaked during winter months. The lowest absolute intensity was for influenza while bronchitis was highest A moderate association for peak timing of infection were found between influenza and URI. For absolute intensity, moderate correlations were found between influenza and bronchitis, and influenza and URI.

Seasonal respiratory diseases represent significant challenges in both surveillance and control. AHR gives researchers the means to quantify seasonal outbreaks and compare seasonal characteristics of multiyear time series of common pediatric respiratory diseases.

J.B. Wenger and E.N. Naumova. What Happens in Vegas, Doesn’t Stay in Vegas: Traveling Waves of Pneumonia and Influenza in the US Elderly Population, 1991-2004

Abstract: To better prepare for influenza epidemics, their spatiotemporal variations need to be assessed. This study focuses on the season-by-season timing and geographic shift of influenza in the elderly population aged 65 or above.

Records on in-patient visits for influenza, defined by having an ICD-9-CM code of 487, were obtained from the Centers for Medicare and Medicaid Services for the years 1991 through 2004. The 248,889 records were aggregated by states and weeks. A flu season is defined as July 1st through June 30th. Timing of the epidemic peaks as well as absolute intensity were derived from harmonic regression.

Over the 13 seasons, the time difference between the first and the last state-wide influenza peaks was 4 weeks, with annual differences ranging from 5 to 22 weeks. In the majority of the seasons, a west-to-east (Nevada to New England) pattern was observed, although inconsistent pattern had occurred. There is an inverse relationship between peak week and absolute intensity. The West North Central area consistently experienced the most intense influenza seasons.

Seasonal variation of influenza in the US presents a challenge for preparedness. States need to be informed with other states' epidemics in order to better provide interventions for the susceptible.