Quantifying Health Policy Uncertainty in China Using Newspapers: Text Mining Study.

Area of Science:

• Public Health
• Health Policy Analysis
• Text Mining and Data Science

Background:

• China has implemented numerous health measures since the 2003 SARS outbreak and the 2019 COVID-19 pandemic.
• Existing studies lack a comprehensive index to quantify health policy uncertainty, encompassing volatility, risk, and policy characteristics in China.

Purpose of the Study:

• To develop and construct a Health Policy Uncertainty (HPU) index for China using text mining techniques.
• To quantify the volume of health policy reports and news articles to reflect policy uncertainty from 2003 to 2022.

Main Methods:

• Utilized text mining on articles from 11 influential Chinese newspapers (Wisenews database) from 2003-2022.
• Constructed the monthly China HPU index based on keyword frequency, with robustness verified by correlation analysis (r > 0.74).
• Performed comparative analysis of China and US HPU indexes, including during SARS, COVID-19, and the Affordable Care Act period.

Main Results:

• The China HPU index demonstrated significant fluctuations, correlating with key health events.
• The HPU index peaked at 502.0 during early COVID-19 (May 2020) and reached a high of 613.8 in December 2022 following the 'New Ten Rules' policy.
• Substantial differences were observed in HPU index fluctuations between China and the US across major health crises and policy periods.

Conclusions:

• Changes in critical health policies, major events, and widespread infectious diseases significantly impact China's health policy uncertainty.
• Divergent political systems and media environments contribute to distinct HPU trends between China and the US, though COVID-19 caused volatility in both.
• This study presents the first systematic text mining approach to analyzing Health Policy Uncertainty in China.