Comparison of Targeted Web-based Advertising vs Traditional Methods for Recruiting Older Adults into Clinical Trials

Abstract

Background:

Recruiting older adults into clinical trials can be particularly challenging. Our objective was to determine if targeted web-based advertising is an effective recruitment strategy.

Methods:

We compared the recruitment rates of traditional and targeted web-based methods for three representative clinical trials involving older adults. All studies utilized traditional recruitment methods initially, but shifted towards primarily targeted web-based advertising after experiencing slow recruitment rates.

Results:

We found that web-based advertising reached more individuals compared to traditional methods. Compared to traditional methods, web-based methods also had at least twice the rate of expressed interest, completion of telephone and in-person screening, eligibility, and enrollment. Additionally, the proportion of individuals excluded after the telephone screening did not differ according to whether targeted web-based advertising (STAMINA: 51%; Berries and Steps: 62%; ISTIM: 20%) or traditional methods traditional methods (STAMINA: 48%; Berries and Steps: 69%; ISTIM: 23%) were used within each study. Those recruited using web-based advertisements tended to be younger compared to traditional methods, but were similar in racial distribution and education.

Conclusion:

Targeted web-based advertisements may be more effective in recruiting older adults for clinical trials at a faster rate than traditional recruitment methods, but need further evaluation of compatible study designs, potential population bias, and cost effectiveness.

Introduction:

Recruitment of older adults for clinical trials remains one of the biggest challenges due to their likely frailty, comorbidities, disabilities, and fear of participating in research.¹ These challenges often result in delays,² which are not only costly but may lead to trial discontinuation.³ Inclusion of older adults is critical given people are living longer but are not necessarily healthier, and results from clinical trials are needed to provide information to better guide healthcare decision-making.

Many trials rely on traditional recruitment methods, which include flyer postings at relevant community sites, physician referrals, and post card mailings. However, these methods tend to be slow, expensive, and ineffective for some populations.⁴ Utilizing internet advertisements that are delivered to target populations on various websites/social media platforms (henceforth called “targeted web-based advertisements”) may be beneficial to help bolster recruitment, even in older adults. In recent years, the number of adults using computers online, including those over the age of 65, has been growing steadily.⁵ This not only provides an opportune platform to advertise trials but also to target specific populations that may be more difficult to reach using traditional methods. Additionally, online advertisements offer the benefit of being easily modified in content and geographic distribution over-time, based on engagement success rate.⁶

Utilization of targeted, web-based advertisement is a relatively new concept in research, yet its popularity is growing. However, few studies have actually compared the effectiveness and utility of such methods to more traditional approaches. Although a recent meta-analysis showed online recruiting to be cost-effective,⁷ questions remain regarding: 1) its effectiveness specifically in older adult populations that may not be as accustomed to using internet advertising, and 2) whether it attracts a less representative population of younger and more highly educated individuals. Therefore, we aimed to compare the effectiveness of targeted web-based recruiting for three clinical trials involving older adults, and determine whether the participants recruited from targeted web-based advertising differed from those recruited using traditional methods. We hypothesized that (1) targeted web-based advertising will achieve a higher recruitment rate for prospective participants compared to traditional methods and (2) the prospective participants recruited by this method will be younger, less diverse, and more highly educated compared to traditional methods.

Methods:

Study Selection:

We compared the recruitment rates of traditional and targeted web-based methods for three representative clinical trials involving older adults: 1) the Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA) (CT.gov Identifier: NCT05422885), 2) the Dietary Intervention Targeting Inflammation, Motivation, and Engagement in Physical Activity in Sedentary, Older Adults with Depressive Symptoms Study (Berries and Steps) (CT.gov Identifier: NCT05735587), and 3) the Individualized Brain Stimulation To Improve Mobility Trial (ISTIM) (CT.gov Identifier: NCT03814304). The study characteristics are summarized in Table 1. All studies were conducted by similarly trained research staff at the Marcus Institute for Aging Research located in Boston, Massachusetts and approved by the Advarra IRB.

Table 1.

Study characteristics for the STAMINA, Berries and Steps, and ISTIM studies.

	STAMINA Study	Berries and Steps Study	ISTIM Trial
Target Population	Older adults aged ≥65 years with slow gait speed and Mild Cognitive Impairment	Older sedentary adults aged ≥65 years with depressive symptoms	Older adults aged ≥60 years with slow gait, mild-to-moderate executive dysfunction, and self-reporting fall risk
Main Study Aim	Determine feasibility and safety of administering intermittent doses of Dasatinib and Quercetin in older adults at risk of Alzheimer’s disease	Determine feasibility and preliminary effectiveness of intervention supplementing fiber and anthocyanins to usual diet compared to control (placebo)	Compare effects of tDCS on dual task costs to physical and cognitive factors on the causal pathway to falls and everyday function
Recruitment Goal (number of completed participants)	12	40	120
Recruitment Time Frame	March 2022 - October 2023	March 2023 - March 2024	June 2019 – May 2024
Utilization of Targeted Web-Based Advertising Method	March 2023 – October 2023	August 2023 – March 2024	February 2022 – May 2024
Study Design	Single-site, Open Label, Pre-Post Pilot study	Single-site, Randomized, Double-Blind, Parallel Pilot Study	Single-site, Randomized, Sham-Controlled, Double-Blinded Study
Study Intervention	Dasatinib (100mg) and Quercetin (1250mg) for 2 consecutive days, every two weeks over a period of 12 weeks (i.e., 12 doses total)	Freeze-dried blueberry powder (48g) (~600 mg of anthocyanins and ~8 g of fiber) or nutritionally matched placebo powder daily for a total 12 weeks	Transcranial direct current stimulation (tDCS) for 20 20-minute sessions over a 4 week period
Type of Study Intervention	Geroscience FDA regulated intervention	Dietary intervention	Brain stimulation intervention
Inclusion/Exclusion Criteria	Included if ≥ 65 years of age with mild cognitive impairment and slow gait, free of chronic heart, liver, kidney, and neurological/ neurodegenerative diseases. Excluded if reported the use of certain medications that interact with Quercetin or Dasatinib or middle cerebral artery could not be insonated using TCD. Extensive details can be found elsewhere.15	Included if ≥ 65 years of age, reported sedentary behavior and mild levels of depressive symptoms, free of diabetes, cognitive impairment, psychotic disorders and inflammatory bowel disorders. Excluded if regularly participated in high intensity exercise and/or had >7,500 steps per day, unstable use of antidepressants, or recent and regular use of antibiotics and/or probiotics.	Included if ≥ 60 years of age with slow gait, mild-to-moderate executive dysfunction, and self-reported fall risk, free of gait disorders, neurological or neurodegenerative diseases, unstable medical conditions, or contraindications to tDCS.
Duration of Study	14 Weeks	14 Weeks	35 Weeks
Total Number of Visits, Number of On-Site Visits	8, 8	6, 3	27, Varies
Average Length of On-Site Visits	2.5 Hours	1 Hour	1 Hour
Total Time Commitment	18 Hours	5.25 Hours	32.5 Hours
Number of Visits Requiring Blood Draw(s), Volume of Blood Collected at each Visit	7, 60mL at 2 visits and 15mL at 5 visits	3, 10mL	N/A
Total Stipend for Completing Entire Study	$300	$300	$510
Monthly Cost for Targeted Web-Based Advertising Services	$1,233	$1,353	$1,241
Cost Per Enrolled Participant Recruited Via Targeted Web-Based Advertising **	$822	$412	$408

^*Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA), Dietary intervention targeting inflammation, motivation, and engagement in physical activity in sedentary, older adults with depressive symptoms Study (Berries and Steps), Individualized brain Stimulation To Improve Mobility (ISTIM) Trial.

^**To compare across all three studies, the cost per enrolled participant was calculated for the entire utilization period of targeted web-based services.

Recruitment Goals and Timelines:

The recruitment goal of the STAMINA study was 12 participants and recruitment occurred during March 2022 – October 2023. The recruitment goal of the Berries and Steps study was 40 participants and recruitment occurred during March 2023 – April 2024. The recruitment goal of the ISTIM study was 120 participants and recruitment occurred during June 2019 – May 2024.

For the purposes of this analysis, the entire recruitment period was used for the STAMINA and Berries and Steps studies. However only a 2-year window (February 2021 – January 2023) of the entire recruitment period (June 2019 – May 2024) was used for the ISTIM trial. This allowed for a comparable analysis during a similar time-frame that included the COVID-19 pandemic.

Traditional Recruitment Methods:

Similar traditional recruitment methods were utilized in all studies. These were defined as recruitment through non-web-based methods, which primarily consisted of a combination of the following: paper study flyers; study presentations at senior housing sites, senior centers and wellness fairs; soliciting volunteers from registries of previous research participants; requesting physician referrals; and geo-targeted mailing of postcards to older adult residents of the greater Boston area. All study advertising materials (e.g., study flyers, presentation PowerPoints, and physician referral letters) were approved by the IRB prior to use.

Study presentations were advertised in advance to encourage attendance and were delivered by study staff. They described the study and participant responsibilities, after which interested attendees were invited to provide their contact information. Additionally, multiple Marcus Institute for Aging Research registries of research volunteers were utilized. Individuals who participated in prior studies at the Institute and provided consent to be added to the registry were contacted if study staff believed they could be eligible. Local physicians were also asked to give information to patients they believed were potentially eligible.

Regardless of type of traditional method used, individuals were contacted via telephone by study staff within approximately 3 days of expressing interest. During these calls they were provided with more details about the study and asked several questions to determine interest and initial eligibility. Individuals who passed this telephone screening were then scheduled to complete an in-person screening where they provided informed consent before completing the remaining in-person screening procedures. Individuals who could not be contacted initially were left voicemails and/or sent emails. Study staff attempted to contact them 2 more times during different times of day within the following 2 weeks. Any individual who could not be contacted after three attempts was deemed “unable to contact.”

Web-based Recruitment Methods:

For a monthly fee, all studies utilized the same external recruitment assistance company for their targeted web-based advertising. This consisted of online study advertisements directed to individuals based on their browsing history and delivered to them when they were online (e.g., through social media platforms such as Facebook or health sites such as WebMD). The recruitment company developed a proprietary algorithm that relied on peoples’ browsing history, and then displayed a study advertisement on the screen when browsing. As more individuals engaged overtime with the advertisements, adjustments were made to the online algorithms to improve its targeting precision for the population of interest. When an individual clicked on an advertisement, they were asked to complete a brief questionnaire including contact information and relevant questions to gauge study eligibility. Advertisements, similar to study flyers, and brief eligibility questionnaires were created by the advertising company with input from the study staff. The company used key information (e.g., target population, study outcomes, important visit details) provided by study staff and created advertisements that were reviewed, modified, and approved by the research teams. Once study staff were satisfied, the advertisements and questionnaires were approved by the IRB prior to being implemented online.

Individuals recruited through targeted web-based advertising who had expressed initial interest in the study, were contacted and screened in the same manner as individuals recruited through traditional methods. However, individuals that disclosed information on their questionnaire that would exclude them from the study were not contacted for a telephone-screening visit.

Recruitment Tracking and Follow-ups:

Traditional and web-based recruitment methods occurred sequentially. All studies utilized traditional recruitment methods initially, but shifted towards primarily targeted web-based advertising after experiencing slow recruitment rates. All individuals expressing interest through both recruitment methods in the studies were added to a recruitment log, including the date they expressed interest and recruitment method used.

Recruitment Rate:

Study recruitment rates were calculated as the number of volunteers engaged in various phases of the recruitment process divided by the number of months each method was utilized. Rates were calculated for the total number of individuals expressing interest in the study (with the exception of ISTIM), those completing telephone-screening visits, those completing in-person screenings, and those deemed eligible at the completion of all phases.

Participant Demographics:

Self-reported demographic information including age, sex, ethnicity, and education level was collected during the telephone-screening visit for all studies. As a result, there is no demographic information for individuals who initially expressed interest but were lost to follow-up prior to completing the telephone screen.

Statistical Analysis:

For participant demographics, mean and standard deviation were calculated for age while total number and percent were calculated for sex, ethnicity, and education level. For recruitment rates, a monthly rate was calculated for each recruitment stage for each recruitment method. The “dead end” proportion (e.g., the number of individuals who were automatically deemed ineligible based on their initial questionnaire responses/ the total number of individuals expressing interest) was calculated for targeted web-based advertising for each study except for the ISTIM study due to data availability. Individuals were deemed ineligible based on responses that indicated they had an exclusionary health condition, were outside our recruitment location, or unwilling to follow the study protocol. Additionally, the “dead end” proportion (e.g., the number of individuals who were deemed ineligible after the telephone screening/ the total number of individuals who completed the telephone screening) was calculated. Given this was a descriptive study, no formal statistical analyses were performed.

Results:

Study Characteristics:

Detailed characteristics of all 3 studies can be found in Table 1. All studies recruited older adults (≥65 years for STAMINA and Berries and Steps, ≥60 years for ISTIM) in the greater Boston area during similar time periods and offered comparable compensation (ISTIM: $510, STAMINA: $300, Berries and Steps: $300) commensurate with the study’s duration and participant burden. The ISTIM trial required the greatest time commitment (ISTIM: 27 visits over 35 weeks, STAMINA: 8 visits over 14 weeks, Berries and Steps: 6 visits over 14 weeks) however, the STAMINA study involved the longest average visit length (2.5 hours). All studies had similar targeted web-based monthly advertising costs ($1,233 for STAMINA, $1,353 for Berries and Steps, and $1,241 for ISTIM). However, the STAMINA study’s cost per enrolled participant via targeted web-based advertising was approximately two times higher compared to the other studies ($822 for STAMINA, $412 for Berries and Steps, and $408 for ISTIM).

Recruitment Methods:

The STAMINA study used traditional recruitment methods for 12 months followed by primarily targeted web-based advertising for 8 months and ultimately reached the recruitment goal of 12 completed study participants. The Berries and Steps study used traditional recruitment methods for 5 months followed by primarily targeted web-based advertising for 9 months and recruited 33 of the 40 individuals proposed as a recruitment goal. The ISTIM trial used traditional recruitment methods for a total of 31 months followed by primarily targeted web-based advertising for a total of 27 months and recruited 118 of the 120 individuals proposed as a recruitment goal.

Recruitment Rate:

In all studies, recruitment rates when utilizing targeted web-based advertising were higher than traditional methods (Table 2). The most noticeable difference between targeted web-based and traditional methods was in the rate of individuals expressing interest in the STAMINA and Berries and Steps studies (9X greater and 8X greater, respectively). Approximately 20% (STAMINA: 55/246; Berries and Steps: 121/517) of these individuals recruited via targeted web-based advertising in both studies that had expressed interest were not contacted due to questionnaire responses. Not only were the rates of targeted web-based methods greater than traditional methods for the various stages of recruitment, but they also appeared similar across all three studies. The rates for telephone and in-person screening completion and subsequent eligibility were approximately 2–3X greater in targeted web-based methods compared to traditional methods. The proportion of individuals excluded after the telephone screening did not differ according to whether targeted web-based advertising (STAMINA: 35/69 (51%); Berries and Steps: 68/110 (62%); ISTIM: 36/183 (20%)) or traditional methods traditional methods (STAMINA: 16/33 (48%); Berries and Steps: 20/29 (69%); ISTIM: 15/66 (23%)) were used within each study. The rate for in-person screening completion in all studies showed a similar inflection point within 2 months of when targeted web-based methods were implemented regardless of year (Figure 1).

Table 2.

Targeted web-based advertising recruitment rate compared to traditional methods recruitment rate for the STAMINA, Berries and Steps, and ISTIM studies.

	STAMINA Study Recruitment Rates			Berries and Steps Study Recruitment Rates			ISTIM Trial Recruitment Rates
	Total N	Traditional Methods (n/Month)	Targeted Web-Based Advertising (n/Month)	Total N	Traditional Methods (n/Month)	Targeted Web-Based Advertising (n/Month)	Total N	Traditional Methods (n/Month)	Targeted Web-Based Advertising (n/Month)
Individuals Expressing Interest	300	4.5	41.0	563	9.2	73.9	N/A	N/A	N/A
Individuals Completing Telephone Screen	102	3.3	11.5	139	5.8	15.7	249	5.5	15.3
Individuals Eligible After Telephone Screen	51	1.7	5.7	51	1.8	6.0	198	4.3	12.3
Individuals Completing In-Person Screen	29	1.2	2.8	39	1.5	4.7	130	2.8	7.1
Individuals Eligible After In-Person Screen	14	0.5	1.5	29	1.5	3.3	65	1.5	3.5

^*Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA), Dietary intervention targeting inflammation, motivation, and engagement in physical activity in sedentary, older adults with depressive symptoms Study (Berries and Steps), Individualized brain Stimulation To Improve Mobility (ISTIM) Trial.

^**The ISTIM trial did not retain data on individuals who expressed interest in the study.

Figure 1.

Number of individuals completing the in-person screening visit by month for the ISTIM, STAMINA, Berries and Steps studies.

*Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA), Dietary intervention targeting inflammation, motivation, and engagement in physical activity in sedentary, older adults with depressive symptoms Study (Berries and Steps), Individualized brain Stimulation To Improve Mobility (ISTIM) Trial

**Arrows indicate the month at which targeted web-based advertising was implemented and became the primary recruitment method utilized.

A majority of individuals who withdrew from the studies before or after they started were recruited using targeted web-based methods (Table 3). All of those who withdrew from the STAMINA study were recruited using targeted web-based methods, whereas those who withdrew from the Berries and Steps and ISTIM studies were recruited from both traditional (Berries and Steps: n= 4, ISTIM: n= 4) and targeted web-based methods (Berries and Steps: n=6, ISTIM: n=10).

Table 3.

Targeted web-based advertising number of participant dropouts compared to traditional methods number of participant dropouts for the STAMINA, Berries and Steps, and ISTIM studies.

	STAMINA Study			Berries and Steps Study			ISTIM Trial
	Total N	Traditional Methods	Targeted Web-Based Advertising	Total N	Traditional Methods	Targeted Web-Based Advertising	Total N	Traditional Methods	Targeted Web-Based Advertising
Number of Individuals who Dropped from Study After being Eligible Before Start of Intervention	1	0	1	3	1	2	8	2	6
Number of Individuals who Dropped from Study After being Eligible After Start of Intervention	2	0	2	7	3	4	6	2	4

^*Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA), Dietary intervention targeting inflammation, motivation, and engagement in physical activity in sedentary, older adults with depressive symptoms Study (Berries and Steps), Individualized brain Stimulation To Improve Mobility (ISTIM) Trial.

Participant Characteristics:

Across all studies, individuals recruited using targeted web-based advertising tended to be younger (STAMINA: 73y ± 7, Berries and Steps: 72y ± 4, ISTIM: 74y ± 6) compared to traditional methods (STAMINA: 77y ±9, Berries and Steps: 74y ± 6, ISTIM: 81y ± 7). All participants in these studies were similar in gender, racial distribution and education level. In all studies, regardless of recruitment method, individuals tended to be mostly female, white, and highly educated (Table 4).

Table 4.

Demographic characteristics of individuals who completed the telephone-screening visit for the STAMINA, Berries and Steps, and ISTIM studies.

	Mean ± SD or n (%)
	STAMINA Study		Berries and Steps Study		ISTIM Trial
	Traditional Methods Screeners, n=46	Targeted Web-Based Advertising Screeners, n=69	Traditional Methods Screeners, n= 36	Targeted Web-Based Advertising Screeners, n= 109	Traditional Methods Screeners, n= 78	Targeted Web-Based Advertising Screeners, n= 183
Age (y)	77 ± 9	73 ± 7	74 ± 6	72 ± 4	81 ± 7	74 ± 6
Sex (female)	26 (57)	36 (52)	24 (67)	80 (73)	49 (63)	121 (66)
Ethnicity
White	40 (87)	61 (88)	34 (94)	95 (87)	67 (86)	170 (93)
African American	3 (7)	7 (10)	2 (6)	9 (8)	5 (6)	8 (4)
Asian	1 (2)	-	-	4 (4)	-	-
American Indian or Alaska Native	1 (2)	-	-	-	-	-
Native Hawaiian or Pacific Islander	1 (2)	1 (2)	-	-	-	-
Other	-	-	-	-	4 (5)	4 (2)
Refused or Unknown	-	-	-	1 (1)	2 (3)	1 (1)
Highest Level of Education
No Diploma	2 (4)	-	-	-	2 (3)	2 (1)
High School Diploma	4 (9)	5 (7)	-	7 (6)	11 (14)	15 (8)
Some College or Associates Degree	9 (20)	15 (22)	6 (16)	26 (25)	13 (17)	44 (24)
Bachelor’s Degree	13 (28)	21 (30)	10 (28)	21 (19)	13 (17)	45 (25)
Master’s Degree or Higher	18 (39)	28 (41)	20 (56)	55 (50)	37 (46)	76 (41)
Refused or Unknown	-	-	-	-	2 (3)	1 (1)

^*Senolytics To Alleviate Mobility Issues and Neurological Impairments in Aging Study (STAMINA), Dietary intervention targeting inflammation, motivation, and engagement in physical activity in sedentary, older adults with depressive symptoms Study (Berries and Steps), Individualized brain Stimulation To Improve Mobility (ISTIM) Trial.

^**For the ISTIM trial, age is approximately calculated as individuals are only asked for year of birth, not exact age, when completing the telephone screen.

Discussion:

Overall Findings:

This descriptive analysis comparing recruitment methods for three clinical trials recruiting older adults indicates that targeted web-based methods attracted a far greater number of individuals per month (i.e., rate of those who expressed interest). They also had 2–3 times higher rates of telephone and in-person screening completion, with no difference in eligibility. This suggests that those expressing interest in the study had similar qualifications, regardless of the method used to recruit them. Moreover, the recruited participants by both methods had similar demographic profiles and levels of education. Thus, our study suggests that targeted web-based recruitment may be a more effective research recruitment strategy for older adults, although like most methods, the people reached by these methods tend to be white and relatively highly educated.

Although other studies^{7, 8} have compared the effectiveness of targeted web-based recruitment methods to traditional methods in a wide age range of individuals (13–74 years old), the trials focusing only on older adults are few and results have been mixed. In addition to the three trials presented in the current study, one trial found targeted web-based recruiting to be effective in older adults,⁹ while others did not.^{10, 11} A plethora of factors may explain these differences, including the type of intervention, targeted demographic, and geographic location. However, it is worth noting that the studies included in our analysis also varied by intervention type, target health demographics, compensation, and required time commitment, yet recruitment rates remained similar throughout each of the recruitment phases. This suggests that the advantages of targeted web-based advertising for interventions recruiting older adults were not study specific. It may seem counter intuitive that web-based recruitment methods would be effective in older adults who may not be as active online as younger generations. However, a recent study suggested that the engagement rate with the online Facebook study advertisements was much higher in older adults compared to younger adults.¹² This supports the notion that online recruitment is a viable strategy for older adults.

Our analysis found that individuals recruited using targeted web-based advertising tended to be slightly younger compared to traditional methods, but were similar in gender, racial distribution and education level. Even though the number of older adults online is growing, younger individuals are still more active online across all social media platforms.⁵ Social media platforms that younger individuals frequently visit are where the targeted web-based advertisements are typically delivered, which may help explain the differences in age. Despite the studies’ best efforts to recruit a diverse population, individuals included in this analysis were primarily white and highly-educated older adults, regardless of type of recruitment method. The demographic similarities of older adults recruited by the two methods suggests that the use of web-based advertisement did not introduce further selection biases. The lack of diversity may have been affected by including only English speakers. Targeted web-based advertising in their native language has been demonstrated to be an effective strategy for Spanish speakers.¹³ Future studies conducted in more diverse populations are imperative.

Targeted web-based advertising may yield more individuals expressing interest per month, as observed in this analysis, but does not necessarily mean more will actually participate in a study compared to traditional methods. Across all studies in the current analysis, there was a large attrition between expressing interest and completing a telephone screening visit, indicating that a substantial portion of those reached do not actually follow through with enrolling in the study. Additionally, a majority of dropped participants across all studies had been recruited using targeted web-based methods. It is reasonable to assume that an individual is less invested, at least initially, in a study that is presented to them passively though an advertisement online (targeted web-based advertising) compared to one that is actively presented in-person and advertised in advance (traditional method).

Targeted web-based advertising may also be more cost effective if the rate at which participants are enrolled is faster. This may save some of the high costs related to the time and effort it takes study staff to use traditional methods such as in-person presentations to small audiences and the active dissemination of study advertisements to various community sites and medical practices. One comparative analysis of primarily younger and middle-aged adult trials found that individuals were recruited at a faster rate through online advertising, making this more cost effective in terms of money spent on advertising. However, when comparing individuals who expressed interest in the study and then enrolled, traditional methods outperformed.⁷ In our study, we found that the cost per enrolled participant recruited via targeted web-based advertising differed significantly between the studies, in particular the STAMINA study was the most costly. This difference may be attributable to the general difficulty of recruiting for this study, regardless of recruitment method, which asked participants to take two experimental medications including a chemotherapeutic drug. Unfortunately our study was not able to calculate the cost per enrolled participant recruited via traditional methods due to the lack of data necessary to estimate this cost (e.g., total time required by study staff to hang fliers, deliverstudy presentations, print materials, etc.).

Study Limitations:

Our analysis included only three studies from one institute. Although this can be seen as a strength in comparing recruitment methods across similar studies, it may not be representative of recruitment for other studies of at other research centers. Unfortunately, many studies do not report the effectiveness of their recruitment strategies, especially for under-represented older adults, despite this being one of the largest challenges in clinical trials research.¹⁴ Additionally, the generalizability of our study is limited and our results can only be applied to relatively highly educated, white older adults. Older adult populations are heterogeneous and therefore warrant further exploration.

An additional limitation in studying targeted web-based advertising is that the total number of individuals presented with advertisements online is unknown and something that cannot be easily calculated. However, this issue is also true of traditional methods. For example, if a study flyer is posted in a physician’s office, the number of individuals reading the flyer cannot be easily determined.

One final limitation is the possibility of confounding due to the onset or waning of the COVID pandemic during the study period. Recruitment rates for clinical trials during the COVID pandemic decreased dramatically. The pandemic coincided with the use of traditional methods in the ISTIM trial and may have contributed to its lower recruitment rate. However, the rate for completion of in-person screening for all studies showed similar inflection points within 2 months of targeted web-based recruitment implementation, regardless of date, suggesting that differences in recruitment rates were due to web-based vs. traditional methods, rather than their temporal relationship to the pandemic.

Conclusions:

Recruitment for clinical trials is difficult and remains one of the largest hurdles in running a successful trial, especially in older adult populations. Our findings in three studies of older adults suggests that targeted web-based advertisement enables recruitment to proceed at a faster rate compared to traditional methods and does not appear to introduce additional selection bias. Such recruitment methods should be considered a viable option, especially in light of growing internet use across all age ranges.

Key Points:

• Targeted web-based recruitment methods are effective in recruiting eligible individuals at a faster rate compared to traditional recruitment methods.

• Individuals recruited using targeted web-based advertising tended to be younger compared to traditional recruitment methods, but were similar in gender, racial distribution, and education level.

Why Does this Paper Matter?

Compared to traditional recruitment methods, targeted web-based advertising may be a more effective option for recruiting older adults for clinical trials.