Accelerating tobacco regulatory science in the United States: The center for rapid surveillance of tobacco

Abstract

The U.S. tobacco marketplace has undergone rapid and profound change over the past decade, marked by the emergence of new product categories, brands, and marketing strategies that challenge traditional public health surveillance. Delays in identifying and responding to early signals—most notably during the rapid rise of JUUL—highlighted critical gaps in the timeliness, flexibility, and scope of existing surveillance approaches. In response, the National Cancer Institute and the U.S. Food and Drug Administration funded the Center for Rapid Surveillance of Tobacco (CRST) to support near real-time monitoring of tobacco marketing, the tobacco marketplace (i.e., availability and sales), and behaviors. This paper provides a high-level overview of CRST's conceptual framework, which is grounded in public health surveillance principles and organized around signal generation, refinement, and evaluation. CRST leverages and triangulates diverse data sources, including marketing data, point-of-sale audits, retail sales data, non-traditional data streams, and behavioral surveys of youth and adults, to detect emerging public health concerns and inform regulatory action. We also highlight examples from this special issue that illustrate the application of rapid surveillance methods and demonstrate how multi-source triangulation can strengthen the validity and actionability of emerging signals in a rapidly evolving tobacco landscape.

Highlights

• •Introduces the Center for Rapid Surveillance of Tobacco.
• •Describes a rapid, multi-source surveillance framework for monitoring tobacco use.
• •Shows how triangulating data strengthens signal detection and analysis.
• •Emphasizes timely surveillance for tobacco regulation and public health.

1. Introduction

The 2009 Family Smoking Prevention and Tobacco Control Act (TCA) granted the U.S. Food and Drug Administration (FDA) authority to regulate tobacco products to protect public health. As written, the Act should have led to a contraction of the tobacco marketplace by implementing rigorous approval processes through which new products could enter the market, such as the Premarket Tobacco Product Application (PMTA) pathway. However, FDA's initial authority was limited to cigarettes, smokeless tobacco, and roll-your-own tobacco. By the time FDA finalized the deeming rule in 2016—extending authority to all tobacco products—the marketplace had changed substantially. New product categories (e.g., e-cigarettes, nicotine pouches), new brands, and an abundance of flavors (e.g., Pumpkin Spice Cigarillo) had proliferated. By 2016, nicotine pouches had already entered U.S. test markets (Delnevo et al., 2020), and e-cigarettes had transformed the retail environment as they entered traditional brick-and-mortar stores and rapidly evolved across multiple device generations.

These marketplace changes were accompanied by major shifts in tobacco use behaviors, including substantial declines in cigarette smoking among youth and adults and rising use of non-cigarette products and multiple tobacco products. Most notably, the National Youth Tobacco Survey (NYTS) documented a sharp increase in past 30-day e-cigarette use among high school students, from 11.7% in 2017 to 20.8% in 2018 (Cullen et al., 2018). Although the Centers for Disease Control and Prevention (CDC) and FDA could not empirically identify a cause, they speculated that the increase may have been driven by the growing popularity of the JUUL™ e-cigarette brand (Cullen et al., 2018). To the casual observer, the spike appeared sudden; yet early signals had been present well before 2018. As early as March 2017, JUUL™’s popularity was reported in college and high school newspapers and soon thereafter in national media. By mid-2017, JUUL™ emerged as the second best-selling e-cigarette brand in Nielsen Market Scanner data, and by late 2017 it became the market leader. Despite these indicators, questions about JUUL™ were not incorporated into major national surveys such as the NYTS until 2019.

In hindsight, signals related to JUUL™ emerged early, but the public health, scientific, and regulatory communities failed to systematically recognize, evaluate, and respond in a timely manner. Notably, calls for improved surveillance of non-cigarette and emerging tobacco products—and for rapid response systems capable of identifying early marketplace signals—had been made more than a decade prior (Delnevo and Bauer, 2009). The unexpected rise in youth e-cigarette use created renewed urgency, prompting the National Cancer Institute (NCI) to convene an interagency meeting in October 2019 titled Rapid Surveillance in a Rapidly Changing Tobacco Landscape. As the non-cigarette tobacco category expanded, public health faced both challenges (e.g., poly-tobacco use) and opportunities (e.g., harm reduction). Indeed, while increased product diversity in the noncombustible marketplace may facilitate switching away from combustibles, novel nicotine products may also encourage experimentation with nicotine, an addictive substance, among youth and non-users.

In August 2021, NCI and FDA issued NOT-OD-21-175, a notice of intent to fund a Center for Rapid Surveillance of Tobacco (CRST). Designed to address gaps in traditional surveillance, CRST provides near real-time monitoring and analysis of tobacco marketing, product availability, and use patterns to inform public health and regulatory decision-making. In May 2022, the Rutgers Center of Excellence in Rapid Surveillance of Tobacco was funded by NCI and FDA to execute CRST. This manuscript provides a high-level overview of CRST and its conceptual framework and highlights papers in this special issue that are part of CRST and/or leverage rapid surveillance data sources and methods.

2. Methods

CRST is a collaborative initiative led by Rutgers University, funded by NCI and the FDA Center for Tobacco Products. Surveillance and monitoring activities occur at both a national level as well as in six sentinel states (NJ, VT, NC, KY, OH, and CA) and is guided by an interdisciplinary team of lead investigators, co-investigators, expert consultants and a steering committee. The aims of CRST are focused on the rapid surveillance of 1) tobacco marketing to generate signals of interest; 2) the marketplace to generate and refine signals of interest; and 3) changes in tobacco product use behaviors to generate, refine, and evaluate signals of interest. These aims occur in tandem to implement an optimal rapid surveillance of tobacco to enhance FDA's regulation of tobacco products.

A core function of CRST is to identify and interpret early signals that may indicate emerging issues in the tobacco environment. Signals refer to data or insights that could indicate significant patterns of concern, yet may fall outside the scope of traditional public health surveillance (Balajee et al., 2021). Signals act as a “canary in the coal mine” or early warnings, prompting further investigation to determine whether they represent an actual public health problem. In the context of CRST, signals can be unexpected increases in product sales or use, including rapid uptick in sales or youth adoption of a particular brand or device. Signals may also emerge from regulatory concerns, including the rapid sales growth of unauthorized products like GeekBar™ (Ganz et al., 2025a; Jensen et al., 2025), or from new marketing strategies that target vulnerable populations or seek to circumvent policies.

Importantly, a signal is preliminary and inherently uncertain, requiring triangulation across multiple data sources before it should be considered actionable. Single-source reports or isolated events are rarely sufficient on their own; instead, signals are evaluated based on the seriousness of the issue, the quality of the data, and whether multiple independent sources converge on the same finding. As such, CRST engages in robust triangulation that integrates marketing surveillance, convenience store sales data, point-of-sale audits, national surveys, and expert field observations. This facilitates the ability to distinguish genuine signals from background “noise” in a complex and constantly evolving tobacco marketplace.

CRST draws on FDA's sentinel initiative (Ball et al., 2016; Platt et al., 2018), and public health surveillance principles (German et al., 2001), for our conceptual framework which employs three signal steps and attends to priority public health surveillance attributes. Signal generation focuses on identifying potential signals. Signal refinement evaluates a signal's magnitude and significance. Lastly, signal evaluation requires formal epidemiological analysis with more rigorous data systems. As outlined in Ganz et al., different data sources are mapped to each signal phase with different priority surveillance attributes (e.g., standardization, accuracy, representativeness, flexibility, and timeliness) at each signal phase (Ganz et al., 2025a). In brief, flexibility and timeliness are emphasized for signal generation and refinement and accuracy and representativeness are emphasized for signal refinement and evaluation. In the context of rapid surveillance there are tradeoffs—systems strong in some attributes are often weak in others and vice versa. Indeed, the large national surveys, like NYTS, are strong with respect to standardization and representativeness, making them ideal for signal evaluation; but these systems are challenged with respect to flexibility and timeliness, limiting their ability to pivot quickly in response to an emerging threat, as demonstrated by the JUUL™ example. On the other hand, our Omnibus survey (see (Bover Manderski et al., 2024). – which does not use a probability sample, is less representative, but highly flexible and timely, making it ideal for signal refinement. Via triangulation, the strengths of one system compensate for the weaknesses of another, increasing confidence and bolstering validity of findings across a range of data sources. This approach is exemplified in Pacek et al., where youth tobacco use patterns were triangulated across a convenience sample online survey and two different probability online surveys (Pacek et al., 2025a). Importantly, they noted consistent patterns of youth tobacco use and brand preference – despite different sampling approaches and costs.

Early signals are rarely first identified in traditional survey data. Using a variety of data sources, CRST monitors and describes changes in tobacco product marketing across both traditional and non-traditional channels to generate early signals of interest. As described by Ganz et al., these channels include consumer-facing print and online advertisements via Media Radar and Trinkets & Trash; trade publications for tobacco retailers (e.g., Convenience Store News); direct-to-consumer mail and email marketing via Mintel; and branded tobacco websites and social media accounts (Ganz et al., 2025a). These monitoring activities are invaluable for identifying new tobacco products, brands, packaging, flavors, and marketing strategies used to promote tobacco products. In addition, CRST routinely monitors tobacco marketing at the point of sale, where newly emerging brands are commonly observed—particularly in vape shops (Hrywna et al., 2025a).

Importantly, changes in tobacco marketing do not always translate into consumer uptake. Product availability at the point of sale, however, along with subsequent sales patterns, can serve as early indicators of behavioral change and help refine signals of interest. As described by Hrywna et al., CRST assesses tobacco product availability, including brand presence for emerging product categories (e.g., e-cigarettes and nicotine pouches), across tobacco retailers in six sentinel states, including specialty retailers such as vape shops and cigar bars (Hrywna et al., 2025a) These point-of-sale audits can also be leveraged to examine geographic variation in product availability. For example, in this special issue, Feizy et al. demonstrated meaningful differences by rurality in Kentucky, with e-cigarettes and vapes more commonly available in urban areas and little cigars and nicotine pouches more prevalent in rural communities (Feizy et al., 2025).

While point-of-sale audits document product, brand, and flavor availability, CRST's analysis of Nielsen market scanner data — retail sales information collected from stores across the United States — is critical for monitoring and describing changes in sale volume for both existing and emerging tobacco products at both national and regional levels (Delnevo et al., 2021; Ganz et al., 2025b; Hrywna et al., 2025b). These analyses pay particular attention to products and brands of regulatory interest to the FDA, such as those that have recently received marketing authorization order, as well as products targeted for enforcement actions, such as Geek Bar (Ganz et al., 2025b). In addition, scanner data allow for examination of specific product characteristics—such as flavor, unit count, and price—that may be driving sales. Although recent tobacco control efforts have focused heavily on e-cigarettes and nicotine pouches, Robichaud et al. demonstrates the continued importance of surveillance of cigarettes, showing that despite overall declines in cigarette sales, the deep-discount segment has grown and prices have decreased (Robichaud et al., 2026). These findings underscore that rapid surveillance remains necessary even for long-standing tobacco products. Survey data can complement these data by indicating who uses these products. Importantly, retail market scanner data has limitations; while they capture sales in the most common tobacco retail channel (i.e., convenience stores), they do not fully represent all retail environments, such as online sales. Indeed, as highlighted in this special issue, Ma et al. employed web-scraping methods to monitor online e-cigarette retailers, documenting the widespread use of concept flavor labeling as a strategy to circumvent flavor regulations and illustrating the value of alternative data sources for identifying emerging marketplace signals (Ma et al., 2025).

CRST also employs innovative, non-traditional data collection strategies to assess the products youth may be using. Jensen et al. describe a system that allows schools and community partners to submit tobacco products confiscated from young people via an easy-to-use, smartphone-optimized online survey (Jensen et al., 2025). This approach generates near real-time, product-level evidence of youth tobacco use, including both dominant and emerging brands and flavors, often well before such trends are detectable in population-based surveys. The data reported to date also closely mirror patterns of youth use where e-cigarettes remain the predominant product of choice (Pacek et al., 2025a; Villanti et al., 2025).

A guiding principle for FDA tobacco regulation is determining whether a product is “appropriate for the protection of the public health,” which ultimately depends on both product risk profiles and patterns of use. Accordingly, monitoring tobacco use behaviors, evaluating signals of interest, and understanding the epidemiology of tobacco use are central to CRST's mission. This includes assessing how products influence initiation, cessation, and dual or polytobacco use. To do so, CRST triangulates data across multiple surveys of youth, young adults, and adults, drawing on both original data collection and secondary data analyses of large national surveys, including but not limited to the NYTS (Chen-Sankey et al., 2022; Delnevo and Villanti, 2023), Population Assessment on Tobacco and Health (Jensen et al., 2023), National Survey on Drug Use and Health,(Pacek et al., 2025b) and the Tobacco Use Supplement to the Current Population Survey (Delnevo et al., 2025).

Analyses from the Truth Initiative reported in the special issue highlights rapid increases in youth polytobacco use over a short period (Do et al., 2025), and illustrates how quickly use patterns can shift and the value of high-frequency data collection for detecting inflection points that annual surveillance may miss. Indeed, in recognizing the limitations of annual surveys, CRST leverages the Rutgers Omnibus Survey (Bover Manderski et al., 2024), a quarterly online survey of 18–45 year olds from Amazon Mechanical Turk. As highlighted in the special issue by Villanti et al., CRST also conducts and maintains a longitudinal cohort of youth and young adults, which demonstrates how flexible, quarterly data collection can capture brand uptake, susceptibility, and new product trial rapidly (Villanti et al., 2025). For both the Omnibus survey and the Cohort study, the list of “usual brands” used is regularly updated to reflect the most recent marketing or marketplace signals.

CRST also places strong emphasis on rigorous and flexible measurement of tobacco use behaviors, recognizing that accurate surveillance must capture not only ever use or trial, and current use, but also patterns of cessation, product switching, dual use, and polytobacco use as the marketplace evolves. In this issue, Pacek et al., examine use behaviors in the National Survey on Drug Use and Health (Pacek et al., 2025b) which show that nearly one-quarter of adult tobacco users engage in dual or polytobacco use, with particularly high prevalence among young adults—highlighting the complexity of tobacco use patterns that single-product surveillance may overlook. At the same time, CRST prioritizes careful survey measurement for emerging products, as noted in this special issue, our research highlights consumer confusion between smokeless tobacco and nicotine pouches and survey question order effects can meaningfully distort prevalence estimates and potentially obscure or falsely amplify emerging signals (Bover Manderski et al., 2025). Together, these findings underscore the importance of precise measurement and triangulation across data sources to ensure valid interpretation of evolving tobacco use behaviors. Of note, for emerging products such as e-cigarettes and nicotine pouches, public health relevance depends on who is using the product. Indeed, as we have noted previously, temporal accuracy is therefore critical (Bover Manderski et al., 2017). Accordingly, best practices for analyzing adult populations should consider bifurcating those who quit smoking or other products based on whether cessation occurred before or after the period of widespread availability of the emerging product of interest (Delnevo et al., 2025).

3. Results

Rapid surveillance must also prioritize dissemination of findings to support timely public health and regulatory response, and CRST does so through multiple complementary activities. First, emerging findings are reported monthly to federal partners and the CRST steering committee. Second, CRST produces a weekly Tobacco Industry Marketing (TIM) alert email newsletter highlighting notable developments in tobacco product marketing and promotion (Ganz et al., 2025a). Third, CRST maintains a publicly accessible resource hub designed to share emerging signals, data products, and tools with researchers, regulators, practitioners, and other stakeholders (https://tobaccocrst.org). Resources available through the website include a searchable image library of tobacco and nicotine products confiscated from youth (Jensen et al., 2025), as well as archived TIM alerts, data briefs from CRST's various surveillance activities, and fact sheets that provide comprehensive signal analyses leveraging multiple CRST data sources. Collectively, these dissemination efforts are intended to shorten the time between signal identification and action and operate in parallel with, and in addition to, traditional scientific dissemination activities, including peer-reviewed manuscripts and presentations at scientific conferences.

A prominent example of CRST's dissemination approach is the Geek Bar fact sheet, which provides a clear illustration of how early signals are communicated, refined, and updated over time. Geek Bar was reported early through CRST's monthly briefings to federal partners after it was first identified as an emerging issue in the youth confiscated tobacco product surveillance system (Jensen et al., 2025), where the brand appeared with increasing frequency. This initial signal was subsequently triangulated using additional CRST data streams, including retail sales data and tobacco marketing surveillance (Ganz et al., 2025a). As Geek Bar™ continued to gain market share and survey data confirmed patterns first observed in the confiscated product data—including its emergence as the top brand reported by youth across multiple surveys (Pacek et al., 2025a; Villanti et al., 2025). The fact sheet, first released in July 2024, was revised twice to reflect the evolving evidence base. CRST's dissemination efforts related to Geek Bar™ were leveraged by the FDA, which issued warning letters in December 2024 to more than 100 retailers for selling the unauthorized product (FDA, 2026).

4. Conclusion

Tobacco use in the United States continues to evolve in response to rapid changes in product design, marketing strategies, and regulatory environments, underscoring the growing need for timely and adaptive surveillance systems. The importance of rapid surveillance extends beyond the U.S., as illustrated by a paper in this special issue which found via post-policy surveillance that New Zealand's excise tax reduction on heated tobacco products revealed limited awareness and modest uptake despite a substantial price intervention—highlighting the value of early evaluation to assess policy impact (White et al., 2025).

The CRST model demonstrates how triangulating multiple data sources strengthens the validity and interpretability of emerging public health signals. In a dynamic and increasingly complex marketplace, rapid signal detection is critical for regulators, where delays may result in missed opportunities for intervention. By balancing speed with rigor and sustaining flexible, multi-source surveillance mechanisms, CRST provides critical evidence to support timely public health and regulatory decision-making.

Declaration of generative AI and AI-assisted technologies in the manuscript preparation process

During the preparation of this work, the authors used ChatGPT, a large language model developed by OpenAI, to assist with language editing and clarity. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.

CRediT authorship contribution statement

Cristine D. Delnevo: Writing – review & editing, Writing – original draft, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Ollie Ganz: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Patrick Barnwell: Writing – review & editing, Validation, Supervision, Formal analysis. Heidi Seabrooks-Smith: Writing – review & editing, Supervision, Project administration, Conceptualization. Kathyrn Edwards: Writing – review & editing, Visualization, Resources, Conceptualization.

Fundings

This work was supported by a grant U01CA278695 from the National Cancer Institute (NCI) of the National Institutes of Health and the U.S. Food and Drug Administration (FDA) Center for Tobacco Products. This content is solely the responsibility of the authors and does not necessarily represent the official views of the NCI or FDA.The authors receive funding via grants from the National Institutes of Health (CDC, OG, MH, ACV, MTBM, JJ, MBS, MM) and grants or contracts from state health departments (MH, MTBM, MBS). CDD serves as the chair of the U.S. Food and Drug Administration (FDA) Tobacco Product Scientific Advisory Board. The authors have no conflicts of interest to disclose.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Cristine Delnevo reports financial support was provided by National Institutes of Health National Cancer Institute. Cristine Delnevo reports financial support was provided by US Food and Drug Administration. Cristine Delnevo reports a relationship with US Food and Drug Administration that includes: board membership and funding grants. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.