Trends in Patient Recruitment Advertising From 2012-2019

Credit: photon_photo | stock.adobe.com

Clinical trial advertisements are a significant way through which prospective patients learn about and act upon clinical trial participation opportunities. During the recruitment phase of a clinical trial, advertising and promotion—implemented at the local level by investigative sites and at the regional or national level by research sponsors—make up a substantial share of the total patient recruitment budget.¹

Despite the substantial advertising dollars spent annually—some estimates are as high as $1.5 billion—patient recruitment remains one of the most difficult challenges that clinical research professionals encounter. Only approximately 60% of all activated investigative sites, on average, are able to enroll at least one patient into a clinical trial and nearly one-third of all randomized patients drop out of their clinical trial prematurely.

As a result, Phase II and III clinical trial durations are 36% longer, on average, than they were 10 years ago.² With the introduction of mobile technology, digital and social media, and instant messaging, capturing the attention of patients has become even more difficult and competitive. Subject matter experts in prescription drug advertising have noted the rapid growth in spending, particularly in direct-to-patient advertising³ to customize and target promotional messages.

The FDA Office of Prescription Drug Promotion has also noted that the content of these targeted promotional messages has changed over time as communications have been shortened and simplified, often overstating benefits while understating and downplaying prescription drug risks.⁴ To our knowledge, assessment of longitudinal changes in the content of clinical trial advertisements has not been conducted. There is currently no archive of clinical trial ads, indicating a real gap in one’s ability to analyze how clinical trial advertising has evolved and reached its target audience over time.

This article represents an initial effort to identify trends in the written and graphical content of clinical trial advertisements. The purpose of this research is to spark conversation about how clinical trial advertising is evolving, and how to best leverage this advertising to support not only patient recruitment but also public and patient education and engagement.

Methods

We focused on comparing visual advertisements from two distinct time periods. Visual advertisements are a prominent recruitment tactic² and include flyers, print ads, and social media posts.

Our visual advertisement search was mainly conducted by collecting publicly available ads presently displayed on websites from advertising agencies focused on patient recruitment. For older ads, we used a public internet archive and gathered all ads implemented between 2000 and 2012.

We dated the advertisements by the year that they first appeared on the website. For additional ads to round out our limited sample, we solicited samples from senior executives of advertising agencies that have been operating and developing patient recruitment advertisements on behalf of pharmaceutical companies for more than 20 years.

We coded the ads for two main observation categories: subject matter and graphical content. To control for varying lengths in advertisements, we coded only the first page of multiple-page visual advertisements, tracking just the information communicated to people at first contact.

Coding was conducted by two researchers using the same codebook, written prior to any analysis. For interrater reliability, both researchers conducted a preliminary analysis of 10 shared ads, then compared and discussed the coding to align on any reviewer discrepancies. The researchers then divided and coded the remaining advertisements.

Because the goal of this research is to capture longitudinal trends in how clinical trial patient recruitment is changing over time, we organized our ads into two temporal categories: recent ads released between 2018 and 2023, and older ads released between 2004 and 2012. We conducted statistical tests for significance of differences in the means between the two advertisement groups.

We consulted with some subject matter experts, all of whom are senior level executives within advertising agencies focused on patient recruitment. We will include their knowledge to frame our empirical results.

Data Characteristics

Table 1: Descriptive Table of Old (2004-2012) and New (2018-2023) Ads

Trends in Copy Content of Ads

To capture the subject matter content of the ads, we included measurements for word count and sentence count. We also coded the tone of the title and flyer content and categorized the sentences on each advertisement into different content types.

We included categories to encapsulate information about the investigational product, participant eligibility, participant benefits, general clinical research, general disease symptoms, contacting the research site, and sentences intending to incite action from the reader.

In comparing the older ads with the newer ads, a number of interesting trends were observed. The average word count for patient recruitment ads decreased by almost a third (28%, p = 0.118), from an average of 81 words per ad to 58 words per ad.

Graph 1: Word Count of Old vs. New Ads

A comparison of content type and purpose of sentences also implied a change in the types of sentences used between older and newer ads. The proportion of sentences devoted to urging the reader to gather more information or providing the study site’s URL and contact information significantly increased in newer ads by around 674% (p = 0.029).

There was also an upward trend in newer ads of sentences geared towards prompting the reader into action, which included sentences about the greater purpose of participating in the trial or overcoming their disease (p = 0.268).

Graph 2: Old vs. New Proportions of Sentences that are Action/Prompting Type

In comparing sentences that provide background about the clinical trial, we categorized sentences into those about general clinical research (e.g., sentences about the clinical research process, its purpose, and its goals); those about the targeted condition (e.g., sentences addressing the patient’s experience of living with the disease); and those about the investigative site (e.g., sentences about the site’s overall mission and expertise).

While the proportion of sentences dedicated to general clinical research and the investigative site was not high in our sample of older ads (0.2% and 6.5% of sentences, respectively), these sentence types were not present at all in the newer ads. The proportion of sentences addressing the target condition also decreased, from 10.6% of sentences in the older ads to 0.7% in the newer ads (p = 0.061).

Graph 3: Old vs. New Proportions of Sentences that are Background Information Type

To capture the sentences that cover administrative topics, we categorized sentences about the procedures that the participant would undergo upon signing-up to participate and sentences about participant eligibility. There was a downward trend in the proportion of sentences about participant procedures (decreased by 57%, p = 0.120). In contrast, there was an increased proportion of sentences about patient eligibility (increased by 48%, p = 0.249).

Graph 4: Old vs. New Proportions of Sentences that are Administrative Information Type

The sentence types of the titles also changed. In the older ad sample, the vast majority of the ads had titles that were statements (90.9%). Whereas the majority of the newer ads also had titles that were statements, the percentage was significantly smaller (62.5%; p = 0.011). The newer ads featured more titles in the form of questions (28.1%) than the older ads (9.1%; p = 0.068).

Graph 5: Title Types

Overall information from the advertisements was also measured as a dummy variable indicator that tracked whether the ad provided information about the procedures or possible benefits the reader could expect by participating. The percentage of visual ads that provided information about what participants could expect, without requiring further contact or a second page, decreased between older (50%) and newer ads (28.1%; p = 0.106).

Graph 6: Whether Participants Know What to Expect from the Ad

Trends in Graphical Content of Ads

There was no significant change in the amount of unique graphics used in the older and newer ads, with both using a mean of around 2.5 graphics per ad. Graphics included logos, images, cartoons, icons, and patterns.

The majority of newer and older ads also had graphics that took up a larger proportion of the ad than did the copy content. However, despite the consistency in older ads and newer ads using a variety of graphics, newer ads (71.9%) used heavier graphic content than copy content more often than did the older ads (59.1%; p = 0.337).

Graph 7: Graphics Count

Table 8: Ads that Are More Graphics-Heavy than Text-Heavy

Summary and Conclusions

Comparing older ads and newer ads revealed several interesting similarities and differences, as well as emerging trends. Overall, the reading grade level remained around grade 8, which is slightly higher than the National Institutes of Health (NIH) recommendation of grade 6-7 for health materials.⁵

Interestingly, while older and newer ads both used around the same number of unique graphics (around 2.5 each), the percentage of ads that were more graphic-heavy than text-heavy increased by around 22% from older to newer ads, indicating that newer ads appear more graphic-heavy and suggesting a focus for newer ads to be visually attention-grabbing.

Notably, the content compositions of the ads showed interesting shifts. The proportion of sentences that encourage participant engagement, in the form of sentences that induce action or encourage readers to contact the study site, increased in newer ads compared to older ads.

In contrast, the proportion of sentences conveying educational content about clinical trials, the investigative site, and the disease of interest decreased in newer ads compared to older ads. This shift in focus from education-focused content to action-inducing content may have a myriad of implications.

By focusing on enticing individuals to participate instead of educating them on what participation entails, we may be losing less clinical research literate potential participants. The average word count also decreased by around 28.2% from older to newer ads. If this trend continues, and the average word count on ads continues to decrease, what does this mean for future clinical trial patient recruitment advertisements?

In the short term, less words and more action-prompting may induce more clicks to the study website and generate more reader interest in the study. However, in the long term, with ads becoming more provocative and action-based, patient trust may suffer as an unintended consequence, as the language reinforces the view that researchers just want patients to enter their trials without caring about the patients themselves.

Regarding administrative information, the proportion of sentences about patient eligibility is higher in newer ads, while the proportion of sentences about patient benefits are lower. This raises questions about whether patients now are less aware of the possible benefits from the study but feel ineligible for the study based on the stated eligibility criteria.

Recent shifts in social media to favor short-form videos have lowered our attention spans as we become used to spending less time on each piece of media in favor of consuming more quantities of media. While newer ads are following the trends in social media short form content, caution is indicated.

With newer ads containing fewer words and relying on action-oriented language, target recipients have less of an opportunity to form an opinion before contacting the investigative site. Additional burden may be pushed to site personnel to set the patient’s expectations before they decide to participate because the ad itself lacks such information.

Overall, the comparison between ads were directionally expected; however, the sample size is small. Because neither a publicly available nor IRB-managed archive of clinical trial advertisements exist, further efforts are necessary to create a database of clinical trial ads for deeper analysis and to further support the trends found in this initial study.

Acknowledgements

The authors wish to thank Gabrielle Froum (Tufts University), Bonnie Brescia (BBKWorldwide), Robert Loll (Praxis) and Brian Schaechter (Artcraft Health) for their invaluable input and assistance.

About the Authors

Victoria Zhang and Kenneth Getz, Tufts Center for the Study of Drug Development, Tufts University School of Medicine.

References

1. Tufts Center for the Study of Drug Development (CSDD).(2020, February).New Global Recruitment Performance Benchmarks Yield Mixed Results.Impact Report: Analysis and INsight into Critical Drug Development Issues, 22(1).

2. Lamberti, M. J., Smith, Z., Henry, R., Howe, D., Goodwin, M., Williams, A., & Getz, K.(2021).Benchmarking Patient Recruitment and Retention Practices.Therapeutic Innovation & Regulatory Science, 55(1), 19–32. https://doi.org/10.1007/s43441-020-00186-4

3, Schwartz L, Woloshin S.Medical Marketing in the United States, 1997-2016.JAMA (2019); 321(1): 80 – 96.

https://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/CDER/ucm090142.htm

4. Wright K.Marketing and Advertising in Drug Development.FDA Office of Prescription Drug Promotion.Public presentation, 2018: https://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/CDER/ucm090142.htm. Accessed July 29, 2024

5. Hutchinson, N., Baird, G.L., & Garg, M.(2016).Examining the Reading Level of Internet Medical Information for Common Internal Medicine Diagnoses.The American Journal of Medicine, 129(6), 637–639. https://doi.org/10.1016/j.amjmed.2016.01.008