Early Warning and Prediction of Scarlet Fever in China Using the Baidu Search Index and Autoregressive Integrated Moving Average With Explanatory Variable (ARIMAX) Model: Time Series Analysis (Preprint)

• DOI: 10.2196/preprints.49400
• Publication Date: 2023-10-30T15:01:07Z
• Authors: Tingyan Luo; Jie Zhou; Jing Yang; Yulan Xie; Yiru Wei; Huanzhuo Mai; Dongjia Lu; Yuecong Yang; Ping Cui; Li Ye; Hao Liang; Jiegang Huang
• Keywords: Scarlet fever; Rank correlation

<sec> <title>BACKGROUND</title> Internet-derived data and the autoregressive integrated moving average (ARIMA) ARIMA with explanatory variable (ARIMAX) models are extensively used for infectious disease surveillance. However, effectiveness of Baidu search index (BSI) in predicting incidence scarlet fever remains uncertain. </sec> <title>OBJECTIVE</title> Our objective was to investigate whether a low-cost BSI monitoring system could potentially function as valuable complement traditional surveillance China. <title>METHODS</title> ARIMAX were developed predict China using from National Health Commission People’s Republic between January 2011 August 2022. The procedures included establishing keyword database, selection filtering through Spearman rank correlation cross-correlation analyses, construction comprehensive (CSI), modeling training sets, testing comparing prediction performances. <title>RESULTS</title> monthly 4462.17 (SD 3011.75) cases, annual exhibited an upward trend until 2019. database contained 52 keywords, but only 6 highly relevant ones selected modeling. A high observed reported cases CSI (r&lt;sub&gt;s&lt;/sub&gt;=0.881). We ARIMA(4,0,0)(0,1,2)&lt;sub&gt;(12)&lt;/sub&gt; model, + (Lag=0) ARIMAX(1,0,2)(2,0,0)&lt;sub&gt;(12)&lt;/sub&gt; combined BSI. 3 had good fit passed residuals Ljung-Box test. ARIMA(4,0,0)(0,1,2)&lt;sub&gt;(12)&lt;/sub&gt;, (Lag=0), demonstrated favorable predictive capabilities, mean absolute errors 1692.16 (95% CI 584.88-2799.44), 1067.89 402.02-1733.76), 639.75 188.12-1091.38), respectively; root squared 2036.92 929.64-3144.20), 1224.92 559.04-1890.79), 830.80 379.17-1282.43), percentage 4.33% 0.54%-8.13%), 3.36% –0.24% 6.96%), 2.16% –0.69% 5.00%), respectively. outperformed better performances smaller values. <title>CONCLUSIONS</title> This study that can be early warning fever, serving supplement systems.