Explore the vast range of titles available.

Online surveys have become a key tool of modern health research, offering a fast, cost-effective, and convenient means of data collection. It enables researchers to access diverse populations, such as those underrepresented in traditional studies, and facilitates the collection of stigmatized or sensitive behaviors through greater anonymity. However, the ease of participation also introduces significant challenges, particularly around data integrity and rigor. As fraudulent responses—whether from bots, repeat responders, or individuals misrepresenting themselves—become more sophisticated and pervasive, ensuring the rigor of online surveys has never been more crucial. This article provides a comprehensive synthesis of practical strategies that help to increase the rigor of online surveys through the detection and removal of fraudulent data. Drawing on recent literature and case studies, we outline several options that address the full research cycle from predata collection strategies to validation post data collection. We emphasize the integration of automated screening techniques (eg, CAPTCHAs and honeypot questions) and attention checks (eg, trap questions) for purposeful survey design. Robust recruitment procedures (eg, concealed eligibility criteria and 2-stage screening) and a proper incentive or compensation structure can also help to deter fraudulent participation. We examine the merits and limitations of different sampling methodologies, including river sampling, online panels, and crowdsourcing platforms, offering guidance on how to select samples based on specific research objectives. Post data collection, we discuss metadata-based techniques to detect fraudulent data (eg, duplicate email or IP addresses, response time analysis), alongside methods to better screen for low-quality responses (eg, inconsistent response patterns and improbable qualitative responses). The escalating sophistication of fraud tactics, particularly with the growth of artificial intelligence (AI), demands that researchers continuously adapt and stay vigilant. We propose the use of dynamic protocols, combining multiple strategies into a multipronged approach that can better filter for fraudulent data and evolve depending on the type of responses received across the data collection process. However, there is still significant room for strategies to develop, and it should be a key focus for upcoming research. As online surveys become increasingly integral to health research, investing in robust strategies to screen for fraudulent data and increasing the rigor of studies is key to upholding scientific integrity.

Digital recruitment methods offer opportunities to address challenges in clinical research participation, particularly in neurology. However, the impact of digital approaches across socioeconomic and demographic groups remains inadequately understood. This study investigates the influence of sociodemographic factors on recruitment and attrition in a remote neurological research cohort, mapping participation pathways and identifying disparities to inform inclusive digital strategies. We conducted a nonexperimental, observational longitudinal cohort study at Mayo Clinic using patient-portal invitations between March and July 2024 as part of a remote speech capture study. Eligibility criteria included age 18 years and older, US residence, and English proficiency. Of 5846 invited patients, progression was tracked across checkpoints (invitation, eligibility screening, electronic consent, and task completion) using Epic (Epic Systems Corporation) to obtain demographic information, Qualtrics (Qualtrics, LLC) for screening, PTrax (a Mayo Clinic-developed Participant Tracking System) for consent tracking, and the recording platform. Socioeconomic context was assessed using the Housing-based Socioeconomic Status (HOUSES) index, where higher values indicate higher socioeconomic status, and the Area Deprivation Index (ADI), where higher values reflect greater neighborhood disadvantage. Data diagnostics included Anderson-Darling tests for non-normality and Little missing completely at random (MCAR) test to characterize missingness. Associations between participation outcomes and age, sex, urbanicity, and socioeconomic indices were examined using nonparametric tests. Exact P values and 95% CIs are reported. Analyses were conducted using BlueSky Statistics (BlueSky Statistics, LLC) and the Python SciPy package. Overall, 415 out of 5846 participants (7.1%) completed all study requirements. Completers were older (median age 66.4, IQR 56.0-72.5; 95% CI 65.1-67.6 years) than noncompleters (median age 62.8, IQR 47.5-72.7; 95% CI 62.2-63.2 years; P<.001). Participants from more socioeconomically disadvantaged neighborhoods were less likely to respond (invitation nonresponder median ADI 45.0, IQR 29.0-63.0 vs interested median ADI 42.0, IQR 27.0-59.0; P<.001), and completers had slightly lower ADI ranks than noncompleters (median 41.0, IQR 27.0-56.0 vs median 44.5, IQR 28.0-62.0; P=.04). Urban participants enrolled faster (median 32.0, IQR 9.0-58.0; 95% CI 31.0-37.0 days) than rural (median 41.0, IQR 22.0-65.0; 95% CI 37.0-49.0 days; P=.01). Female participants responded slower (median 38.5, IQR 14.8-66.3; 95% CI 35.0-41.0 days) than males (median 32.0, IQR 8.0-57.5; 95% CI 29.0-38.0 days; P=.01). No significant differences were observed for the HOUSES index, and device type was unrelated to completion or timelines. Missingness for key variables was completely at random (MCAR χ²3=3.45; P=.24). Digital recruitment does not overcome traditional barriers to participation and may introduce new disparities related to age, urbanicity, and neighborhood disadvantage. These findings inform inclusive digital research strategies, including multichannel outreach, age-specific engagement, and rural technical support. This study applies longitudinal pathway analysis to digital neurology recruitment, offering actionable insights for improving inclusivity in remote research.

Web-based surveys involving self-reported questionnaires are vulnerable to fraudulent responses. Advancements in artificial intelligence and bots have introduced additional challenges to preventing and identifying fraudulent responses to online questionnaires. This study aimed to describe our experiences with fraudulent responses, strategies for preventing and identifying fraudulent responses, lessons learned when conducting a web-based survey with adults living with Long COVID, and recommendations for web-based survey research. The Long COVID and Episodic Disability Study is an international community-engaged study among adults living with Long COVID in Canada, Ireland, the United Kingdom, and the United States. We conducted a longitudinal web-based survey, with online administration of a self-reported questionnaire at 2 timepoints (Time 1 and Time 2), 1 week apart. We recruited through Long COVID community groups using social media, emails, and word of mouth. The survey was disrupted by fraudulent responses, including bots. To defend data integrity, we implemented the following strategies: (1) pausing our initial launch (Wave 1), (2) developing and implementing screening criteria to identify fraudulent responses, and (3) relaunching the web-based survey (Wave 2) with revised recruitment strategies and questionnaire design to prevent and identify fraudulent responses. We received 4663 responses for Time 1 and 1281 responses for Time 2, of which we retained 798 of 4663 (17%) responses and 629 of 1281 (49%) responses. Strategies for preventing fraudulent responses included enabling survey protection features in survey software, shutting down compromised survey links, avoiding recruitment via public social media groups, and removing mention of a financial incentive from recruitment materials. Strategies for identifying fraudulent responses included monitoring response completion times, start and end time stamps, geolocation, and screening for suspicious email address characteristics and duplicates. Our lessons learned fell into the following three areas: (1) survey-design and implementation to prevent and identify fraudulent and bot-generated responses, (2) recruitment strategies to mitigate the risk of disruption by bots, and (3) responding to disruptions caused by fraudulent and bot responses. We recommend the following tactics to prevent and mitigate the risks of fraudulent and bot responses when administering online web-based questionnaires: (1) review current literature and connect with researchers and Research Ethics Boards about strategies before launching, (2) invest in survey software with rigorous information security technology, (3) use bot-detection features available in survey software before launching, (4) design questionnaire items to identify bots and fraudulent actors, (5) tailor criteria for identifying fraudulent and bot responses to the characteristics of the target population, (6) avoid recruitment in public social media groups, (7) engage community leaders in tailored and targeted recruitment, (8) avoid advertising incentives, (9) shut down compromised links rapidly, (10) communicate with the Research Ethics Board about disruptions, and (11) combine automated and manual methods to identify potentially fraudulent responses on time.

Little is known about (1) sociodemographic, psychosocial, or smoking-related differences among individuals recruited to smoking cessation randomized controlled trials (RCTs) using in-person versus online recruitment methods or (2) the relative speed of recruitment using these 2 approaches. This secondary analysis is the first to examine these comparisons in a smoking cessation RCT for people experiencing food insecurity, a vulnerable special population for whom quitting is especially urgent. To compare (1) baseline sociodemographic, smoking-related, and psychosocial characteristics; and (2) screening, eligibility, and enrollment rates of in-person versus online recruits to a smoking cessation RCT for people experiencing food insecurity. Participants completed a brief eligibility questionnaire and a baseline assessment via tablet (in person) or personal electronic device (after clicking an online advertisement). Eligibility required past-30-day food aid use, smoking ≥5 cigarettes per day, and willingness to attempt quitting within 7 days post enrollment. Responses were compared using chi-squared and Fisher exact tests (categorical variables) and 2-tailed t tests (continuous variables). Enrollees recruited online endorsed greater food insecurity (mean 4.5, SD 1.9 vs mean 3.0, SD 2.3; P<.001) and were more likely to be educated beyond high school or equivalent (69% vs 49%; P<.001), have household income of US $20,000 or more (46% vs 36%; P=.03), and be non-Hispanic White (77% vs 50%; P<.001). Online recruits indicated lower motivation to quit smoking (Contemplation Ladder; mean 7.2, SD 2.4 vs mean 8.0, SD 2.8; P<.001) and smoking cessation self-efficacy (mean 20.5, SD 8.0 vs mean 23.2, SD 8.6; P<.001). Online recruits also reported lower subjective social status (mean 4.6, SD 2.0 vs mean 5.9, SD 2.2; P<.001), greater financial strain (mean 17.9, SD 6.3 vs mean 16.2, SD 6.6; P=.004), more depressive symptoms (mean 8.6, SD 6.3 vs mean 7.4, SD 6.1; P=.04), greater loneliness (mean 6.0, SD 2.1 vs mean 5.2, SD 2.0; P<.001), less resilience (mean 19.5, SD 5.1 vs mean 20.5, SD 4.3; P=.02), less alcohol misuse (27% vs 37%; P=.02), and more past-30-day cannabis use (25% vs 15%; P=.01). Enrollment rates were higher online (64.8 per month; n=324) than in-person (7.7 per month; n=178). Although screened eligible at similar rates whether recruited online or in person (79% vs 75%; P=.10), eligible online individuals were more likely to enroll (71% vs 49%; P<.001). This study is the first to compare baseline participant characteristics by recruitment method (in person vs online) in a cessation RCT for people experiencing food insecurity and to evaluate the relative pace of recruitment via those methods. Online and in-person recruits were demographically and psychosocially distinct, and online recruitment was associated with faster accrual than in-person recruitment. These findings inform recruitment strategies for cessation interventions, especially those targeting food-insecure individuals.

Progress in clinical trial research depends on effective recruitment to ensure adequate sample sizes and generalizability of results. Electronic health record (EHR)-based recruitment has been suggested as a potential method to address historical imbalances in trial populations, including underrepresentation of certain racial and ethnic groups. However, disparities in EHR portal uptake and use exist, particularly among these same underrepresented groups. Thus, we hypothesized that EHR-driven recruitment for research may continue to exacerbate these dynamics. In this study, we evaluated whether a financial incentive improved response rates to recruitment messages delivered through an EHR portal in an equitable fashion. A random sample of 1200 patients with diagnoses of breast, colorectal, prostate, or lung cancer and active EHR patient portal accounts received an automated recruitment message for a fictitious cross-sectional survey study, \"The Social Aspects of Cancer Care,\" via their EHR portal. Two strata of patients were sampled, those who identified as non-Hispanic White individuals (n=600, 50%) and those who identified as part of a historically underrepresented racial or ethnic group (n=600, 50%), and randomized to receive an offer of a US $20 monetary incentive (n=400, 33.3%) or no incentive (n=800, 66.7%). Patients were given 14 days to respond. Overall, 39.3% (471/1200) of participants viewed the recruitment message, and 13.3% (159/1200) responded that they were interested. Response rates were higher (P<.001) among non-Hispanic White patients (108/600, 18%) than among African American or Black patients (32/456, 7%) and members of other racial or ethnic groups (18/144, 12.5%). Older patients (≥65 years) were less likely to view the recruitment message (244/720, 33.9% vs 230/480, 47.9%; P=.001) and respond (72/720, 10% vs 82/480, 17.1%; P=.002). Sex was not significantly associated with viewing the message or response outcomes (P=.45 and P=.19, respectively). No differences were observed in response rate for the incentive cohort compared to the cohort with no incentive (60/400, 15% vs 96/800, 12%; P=.21). Patients with cancer who identify as part of a historically underrepresented group were less likely to view and respond to recruitment messages for research studies through their EHR patient portal. Age was also found to significantly impact engagement, with older patients less likely to view and respond. Finally, a modest US $20 financial incentive was not found to significantly impact patient engagement.

Incentivization in clinical trial participation can be challenging, with many studies failing to meet recruitment or retention goals despite traditional compensation strategies. Digital health evolves, and with it, new approaches can emerge to engage participants meaningfully. We propose unique digital images as a novel, symbolic incentive for clinical trials. Digital images combine qualities such as personalization, ownership, and digital visibility, which may drive engagement more effectively than monetary rewards alone. In our illustrative study, participants complete artificial intelligence–personalized digital therapeutics training using CURATE.DTx, generating individualized learning trajectories. These are transformed into digital artworks and minted as nonfungible tokens given as a reward upon trial completion. This concept integrates gamification, personalization, and blockchain technology to support both intrinsic and extrinsic motivation. We explore the implications for decentralized health care, long-term behavior change, and participant recognition in the context of preventive medicine and longevity science. Our aim is to encourage research into the use of digital incentives to transform the trial participant experience and promote sustained engagement in health interventions.

Menstrual complaints are widespread but often stigmatized. The most common is dysmenorrhea, or menstrual cramps, which manifests as mild to severe pain during menstruation and affects >40% of women throughout their reproductive lifespan. Dysmenorrhea is often endured silently or managed through self-medication. Consequently, a vast majority of patients with dysmenorrhea may not be found in medical practices, highlighting the need for direct-to-patient communication to reach a broad and diverse patient population. Primarily, this study aims to reveal the diagnosis status, pain levels, comorbidities, eligibility, and willingness to participate in clinical trials of women affected by dysmenorrhea and menstrual discomfort, based on a broad patient population not necessarily reached in medical practices. Second, this study attempts to test the effectiveness of direct-to-patient communication via online campaigns in engaging patients affected by dysmenorrhea or conditions that may benefit from direct-to-patient communication. Women experiencing menstrual pain were reached through a targeted online campaign using Google Ads (Google LLC) and Facebook (Meta Platforms, Inc) in Germany, Austria, and Poland and were surveyed from April to June 2023. This study is observational. We surveyed 3546 women, 94.6% (3230/3413) of whom reported symptoms consistent with dysmenorrhea, highlighting the high specificity of the Google and Facebook campaigns. Of the affected women in Germany and Austria, 88.5% (874/988) reported pain levels of 6 or higher on a scale of 0 to 10, with even higher pain levels observed in Poland. Elevated pain levels were correlated with dysmenorrhea symptoms but not with premenstrual syndrome (PMS) symptoms. Notably, of the 3230 women reporting symptoms consistent with dysmenorrhea, only 4.6% (n=149) reported being diagnosed with the condition, regardless of elevated pain levels. This can be attributed to two factors: (1) 90.3% (3065/3395) of surveyed women did not seek medical advice, were uncertain about their diagnosis, or their menstrual-related symptoms were not recognized as pathological and (2) among the 9.7% (330/3395) diagnosed, only half of DYS-affected women (149/318, 46.9%) were diagnosed with dysmenorrhea. The other 53.1% (169/318) were diagnosed with PMS but not dysmenorrhea despite regularly experiencing dysmenorrhea symptoms. The situation was better for PMS. Among the 330 diagnosed women, 77.3% (n=255) were diagnosed with PMS, in line with the 80.1% (2729/3409) PMS prevalence in the survey population. Overall, about 8.6% (235/2729) of women with PMS symptoms reported having been diagnosed with PMS, nearly double the diagnosis rate reported for dysmenorrhea. The data reveal a significant diagnostic gap for dysmenorrhea, but not necessarily for PMS, even in high-income countries, as observed in Germany, Austria, and Poland. In these 3 countries, most dysmenorrhea-affected women do not seek medical advice, and up to half of dysmenorrhea diagnoses might be missed. Thus, most affected women might not be found in medical settings (doctors' offices and clinics) despite experiencing significant pain. Online campaigns are shown to effectively reach individuals with menstrual complaints, including those who are undiagnosed or not seeking medical care.

Randomized controlled trials (RCTs) face challenges such as limited generalizability, insufficient recruitment diversity, and high failure rates, often due to restrictive eligibility criteria and inefficient patient selection. Large language models (LLMs) have shown promise in various clinical tasks, but their potential role in RCT design remains underexplored. This study investigates the ability of LLMs, specifically GPT-4-Turbo-Preview, to assist in designing RCTs that enhance generalizability, recruitment diversity, and reduce failure rates, while maintaining clinical safety and ethical standards. We conducted a noninterventional, observational study analyzing 20 parallel-arm RCTs, comprising 10 completed and 10 registered studies published after January 2024 to mitigate pretraining biases. The LLM was tasked with generating RCT designs based on input criteria, including eligibility, recruitment strategies, interventions, and outcomes. The accuracy of LLM-generated designs was quantitatively assessed by 2 independent clinical experts by comparing them to clinically validated ground truth data from ClinicalTrials.gov. We have conducted statistical analysis using natural language processing-based methods, including Bilingual Evaluation Understudy (BLEU), Recall-Oriented Understudy for Gisting Evaluation (ROUGE)-L, and Metric for Evaluation of Translation with Explicit ORdering (METEOR), for objective scoring on corresponding LLM outputs. Qualitative assessments were performed using Likert scale ratings (1-3) for domains such as safety, clinical accuracy, objectivity or bias, pragmatism, inclusivity, and diversity. The LLM achieved an overall accuracy of 72% in replicating RCT designs. Recruitment and intervention designs demonstrated high agreement with the ground truth, achieving 88% and 93% accuracy, respectively. However, LLMs showed lower accuracy in designing eligibility criteria (55%) and outcomes measurement (53%). Natural language processing statistical analysis reported BLEU=0.04, ROUGE-L=0.20, and METEOR=0.18 on average objective scoring of LLM outputs. Qualitative evaluations showed that LLM-generated designs scored above 2 points and closely matched the original designs in scores across all domains, indicating strong clinical alignment. Specifically, both original and LLM-based designs ranked similarly high in safety, clinical accuracy, and objectivity or bias in published RCTs. Moreover, LLM-based design ranked noninferior to original designs in registered RCTs in multiple domains. In particular, LLMs enhanced diversity and pragmatism, which are key factors in improving RCT generalizability and addressing failure rates. LLMs, such as GPT-4-Turbo-Preview, have demonstrated potential in improving RCT design, particularly in recruitment and intervention planning, while enhancing generalizability and addressing diversity. However, expert oversight and regulatory measures are essential to ensure patient safety and ethical standards. The findings support further integration of LLMs into clinical trial design, although continued refinement is necessary to address limitations in eligibility and outcomes measurement.

Annual lung cancer screening (LCS) can decrease lung cancer-related mortality by finding cancer at earlier, more treatable stages, yet uptake remains abysmally low in the United States, especially among adults who seldom interact with the health system. Many eligible individuals are unaware that LCS exists, underscoring the critical need for scalable, population-level communication strategies that increase awareness and engagement. The aim of this study was to evaluate reach, as defined by the Reach, Effectiveness, Adoption, Implementation, and Maintenance framework, as the extent to which the target population comes in contact with a social media-based strategy, Facebook-targeted advertisement (FBTA), designed to connect LCS-eligible individuals in the United States with a digital health communication message. The advertisement served as a digital outreach strategy for promoting engagement with LungTalk, an evidence-based intervention aimed at increasing awareness and informed decision-making about LCS. As part of the INSPIRE-Lung Study (INnovating Social Media for Prevention: LUNG Cancer Screening Awareness, Knowledge, and Uptake), 5 FBTA campaigns were launched over a 79-day period throughout the United States. Advertisements targeted adults aged 50-80 years with interests related to smoking or smoking cessation and linked to a study website where participants could complete an eligibility screener and learn more about the trial. Facebook analytics were used to assess reach, defined by the number, proportion, and demographic characteristics of individuals exposed to and interacting with FBTA content. Key metrics included total reach, impressions, link clicks, and cost-efficiency. The FBTA campaigns reached 1,048,191 unique users and generated 3,109,482 impressions (total advertisement displays, including repeat exposures to the same user). A total of 24,816 individuals clicked on the advertisements (2.37% click-through rate), and 7117 completed the eligibility screener. Of those eligible, 1272 (17.9%) met lung screening criteria, and of these, 483 (38% participation rate) enrolled in the trial. The cost per click was US $0.40, and the cost per enrolled participant was US $19.46. Individuals reached via FBTA were demographically diverse and included many who may be disconnected from traditional health care systems. FBTA is a scalable, cost-effective strategy to achieve population-level reach of LCS-eligible adults. By conceptualizing reach as exposure to an upstream digital message rather than enrollment alone, this study illustrates how social media can broaden population access to evidence-based cancer prevention tools such as LungTalk. Future research should explore embedding intervention content directly into social media platforms and tracking downstream clinical outcomes.

The All of Us Research Program (AoURP) is a prominent precision medicine research initiative committed to diverse participation. The program harnesses digital outreach as a key strategy for recruiting and retaining underrepresented populations, using language that sometimes invokes notions of solidarity. This targeted recruitment of underrepresented groups and potential use of solidaristic language raise concerns about how participation will manifest tangible benefits for these populations and whether institutions assume responsibility for addressing past and present research harms. This study examines how the AoURP conceptualizes \"diversity\" in its social media outreach and how this implementation aligns with the program's stated goals. Specifically, we perform a mixed methods analysis to descriptively capture (1) which underrepresented populations are targeted by the AoURP's social media and (2) how solidaristic messaging is used, if at all, in these calls for participation. AoURP social media posts (n=380) from a 6-month period in 2020-2021 were coded to identify visual depictions and explicit mentions of any \"underrepresented in biomedical research\" (UBR) categories officially targeted by the program. To then characterize UBR-specific appeals, we performed a thematic analysis of UBR-targeted posts, using a coding scheme that identified unsolidaristic language (ie, appeals to individual benefits) and solidaristic language (ie, appeals to benefitting others, attaining shared goals, and addressing injustices). Among the 10 UBR categories officially recognized by the AoURP, \"Race and Ethnicity\" (187/380, 49% of posts) and \"Age\" (71/380, 19%) were the most frequently emphasized, while each of the other remaining categories was rarely invoked (<4/380, 1%). The thematic analysis further identified calls to participate that spanned receiving genetic results (ie, individual benefits), uncovering family and community disease predispositions (ie, benefitting others), improving the future of health (ie, achieving shared goals), and addressing data and health disparities (ie, resolving injustices). In addition to highlighting UBR categories that are more and less emphasized in the AoURP's social media outreach, we also find that the program's messaging indeed resembles a solidaristic appeal to participate. Drawing upon the existing literature on solidarity, we leverage conceptualizations of solidarity as a shared practice grounded in mutuality and bidirectionality to question the AoURP's appeals when institutions fail to fully reciprocate this solidarity. Specifically, we raise concerns about (1) unclear links between participation and addressing health disparities, (2) incomplete acknowledgment of institutions' role in data disparities, and (3) the use of empowerment rhetoric that diverts the onus for correcting these disparities onto participants. Finally, we consider the implications of these issues for future outreach efforts.