Skip to main content
Image of woman looking at her computer monitor.

The COVID pandemic has demonstrated the importance of surveillance systems to recognize and resolve emerging threats and unmet public health needs

Early warning systems, typically taking the form of syndromic surveillance, are vital to quickly detect and respond to pandemics and other public health emergencies. Data from emergency departments and other healthcare facilities form the backbone of syndromic surveillance, including the National Syndromic Surveillance Program (NSSP) operated by the Centers for Disease Control and Prevention (CDC). In addition, other data such as social media posts provide complementary insights and can further strengthen early-warning systems. For instance, social media posts and internet search queries have repeatedly helped public health professionals get in front of influenza-like illness and foodborne illness outbreaks. However, until recently, citizen inquiries to public health hotlines have not been utilized to inform public health early warning systems. 

Lessons learned and shared at the 2021 Syndromic Surveillance Symposium

Recently, I had the opportunity to present insights from Maximus’analysis of inquiries to one of our state-level COVID-19 vaccine hotlines at this fall’s Syndromic Surveillance Symposium, organized by the Council of State and Territorial Epidemiologists (CSTE) in collaboration with NSSP. Our study approach drew on Maximus’ extensive experience supporting our government clients in their pandemic response efforts, including operating numerous public health hotlines related to COVID-19 vaccinations and other public health inquiries at the federal, state, and local levels.

Artificial intelligence (AI) and natural language processing (NLP)offer tremendous insights into emerging health threats

Using artificial intelligence (AI) and natural language processing (NLP) paired with traditional epidemiological and statistical approaches, we were able to detect meaningful public health events related to COVID-19 vaccinations as statistically significant spikes in call volume. Further analyzing the transcripts allowed us to understand the drivers behind these call-volume spikes, including people’s early challenges and unmet needs in navigating vaccine appointment scheduling and access to their personal digital vaccination records. Finally, we were able to trace changes in caller concerns and sentiment over time, including specific safety concerns that emerged associated with reports of adverse events, including the temporary pause recommended during the spring by CDC and the Food and Drug Administration (FDA) in the roll-out of one of the COVID-19 vaccines. Presenting at NSSP also provided an opportunity to discuss some of the challenges associated with analyzing call data, including baseline variability in call volume and call transcription errors – as well as proven strategies to overcome these limitations.

Insights from citizen inquiries can help fortify surveillance systems and improve the delivery of public health programs

As this case study shows, tracking inquiries to public health hotlines can provide valuable information for public health early warning systems. Because the data can be available and tracked in nearly real-time, spikes in call volume can quickly alert public health officials to potential emerging threats and breakdowns in vital public health programs. In addition, once a significant spike in call volume is detected, the data can provide vital clues for addressing the issue. Moreover, tracking call data over time can help public health experts better understand the diverse populations they serve. Agencies can track and monitor key concerns, informational needs, and barriers to accessing public health programs, whether shared across the board or unique to specific underserved populations. This information can prove vital for ensuring equitable access to public health programs and overcoming major public health challenges such as vaccine hesitancy.

This analysis was a collaboration between Maximus Population Health Data Analytics(PHD)and Maximus Performance Analytics Organization (PAX)

About the authors and contributors: PAX: Kishan Joshi is a Data Scientist, David Salvador is Manager of Data Science, and Charles Mendoza is Senior Director responsible for Data Science initiatives. Eric Stewart, Vice President, works towards the development of advance analytics capabilities and customer experience improvement for our government partners. PHD: Joel Hartsell is Director of Data Science and Analytics, and Karin Hoelzer, Senior Director, leads Maximus Population Health Data Analytics (PHD) to provide government clients with the data systems, analytical tools, and technical expertise needed to promote population health.