Seasonal trends and explainable patterns of meningitis: Data preprocessing on fever and neurological symptoms syndrome

Addressing and removing explainable patterns from syndromic data is required to timely detection of outbreaks. This study aims to detect and remove explainable patterns of fever and neurological symptoms syndrome as suspected meningitis which has been occurred in Hamadan province.

We analyzed data on reported cases of suspected meningitis in Hamadan province, between March 2010 and March 2013. Line, local polynomial and moving averages (MA) charts, autocorrelation and partial autocorrelation functions, mean differences and nonparametric Mann- Kendall statistics were used to identify explainable patterns in the data on suspected cases of meningitis. Fourier series and LOWESS (Locally Weight Regression and Smoothing Scoter plot) was used to remove such patterns.

Local polynomial charts, autocorrelation and partial autocorrelation functions, mean differences and nonparametric Mann- Kendall statistics indicated the presence of explainable patterns include Day-of- Week (DOW), weekend, holiday effects, seasonality and temporal trend in the syndromic data of fever and neurological symptoms. Overall, LOWESS in removing explainable patterns and Fourier series in improve of normality data have had better performance.

Results showed the presence of explainable patterns in the data of suspected cases of meningitis. Accordingly, timely and accurately detection of meningiti's outbreak required data smoothing.