Abstract
Background
Atrial fibrillation (AF) clinical trials are essential for informing evidence-based care, yet many trials are discontinued or remain unpublished. This contributes to research inefficiency and reduced transparency. This study evaluated characteristics associated with discontinuation and non-publication of AF clinical trials registered on ClinicalTrials.gov.
Methods
We performed a retrospective analysis of interventional AF trials registered between 1999 and 2025 on ClinicalTrials.gov. Trials were categorized as completed or discontinued, and as published or unpublished. Trial characteristics were extracted, and multivariable logistic regression was used to identify predictors of discontinuation and non-publication.
Results
Among 538 AF trials, 423 trials (78.6 percent) were completed and 115 trials (21.4 percent) were discontinued. Only 153 trials (36.2 percent) were published. The estimated median time from primary completion to publication was 18.0 months (95% confidence interval, 14.0–21.0 months). Trials enrolling 100 participants or more had significantly lower odds of both discontinuation and non-publication. Multicenter trials showed higher odds of discontinuation but lower odds of non-publication. The other characteristics studied were not significant predictors.
Conclusions
A substantial proportion of AF trials are discontinued or remain unpublished, revealing ongoing gaps in research transparency. Larger sample sizes support higher completion and publication rates, while a multicenter design increases discontinuation risk but improves dissemination. These findings underscore the need for strategies that strengthen trial feasibility and reporting practices.
Keywords: Atrial fibrillation, Clinical trials, Trial discontinuation, Publication bias, Research transparency, ClinicalTrials.gov
1. Introduction
Atrial fibrillation (AF) is the most prevalent sustained cardiac arrhythmia, affecting an estimated 33 million individuals worldwide [1]. It is associated with a five-fold increased risk of stroke and doubles the risk of mortality [2]. The global burden of AF continues to rise owing to aging populations and the growing prevalence of risk factors such as hypertension, diabetes, and obesity [3]. Clinical trials play a crucial role in shaping therapeutic strategies for AF by evaluating the safety and efficacy of pharmacologic and non-pharmacologic interventions.
Despite the critical role of clinical trials, a significant proportion of them are either discontinued prematurely or remain unpublished after completion. Trial discontinuation not only results in wasted resources but also exposes participants to potential risk without contributing to the scientific body of knowledge [4]. Similarly, non-publication of trial results undermines research transparency and may lead to publication bias, where positive results are preferentially disseminated over null or negative outcomes [5].
Understanding the factors associated with trial discontinuation and non-publication is essential for improving the design, execution, and reporting of clinical trials. While prior investigations have examined these issues across broader fields of medicine, few have focused specifically on AF trials, despite their high volume and clinical significance. This study aims to fill the gap by systematically analyzing interventional AF trials registered on ClinicalTrials.gov, identifying trial characteristics associated with discontinuation and non-publication, and offering recommendations to enhance trial completion and transparency in AF research.
2. Methods
We conducted a retrospective observational study using data obtained from ClinicalTrials.gov, a publicly accessible registry of clinical trials. We included all interventional studies registered under the condition “atrial fibrillation” from inception through August 2025 (Fig. 1). Trials were included if they investigated therapeutic, diagnostic, or preventive interventions for AF in human subjects. Observational studies, trials with missing status or publication data, and non-interventional studies were excluded. All included studies were human interventional clinical trials; no animal or preclinical studies were included, as such studies are not registered on ClinicalTrials.gov.
Fig. 1.
Number of registered atrial fibrillation trials at ClinicalTrials.gov between 1999 and 2025. The apparent decline in recent years reflects administrative right-censoring and publication lag for recently registered trials, not a true reduction in trial activity.
Each included trial was categorized based on completion status (completed vs. discontinued) and publication status (published vs. unpublished). Publication status was determined by checking for linked publications on ClinicalTrials.gov or searching indexed articles in PubMed using trial identifiers (NCT numbers).
We extracted a range of trial characteristics, including participant sex and age group, sample size (<100 vs. ≥100), funding source (federal, industry, mixed, or other), number of centers (single vs. multicenter), intervention type (drug, device, behavioral, etc.), allocation method (randomized vs. non-randomized), intervention model (parallel, single-group, crossover, etc.), masking (open-label, single, double, triple, or quadruple), and trial phase. The threshold of 100 participants for sample size dichotomization was selected a priori based on its established use in prior registry-based studies evaluating trial conduct and publication patterns as well as its pragmatic clinical relevance as an approximate boundary distinguishing small from adequately powered trials. Sensitivity analyses treating sample size as a log-transformed continuous variable and using enrollment tertiles were also performed and are reported in the Supplementary Appendix.
Descriptive statistics were computed to summarize trial characteristics. Time to publication was defined as the interval between the primary completion date and the first publication date (online ahead of print or print publication, whichever occurred earlier), expressed in months. Median time to publication and corresponding 95% confidence interval (CI) were estimated using the Kaplan–Meier method. Chi-square tests were used to compare categorical variables. To identify independent predictors of trial discontinuation and non-publication, multivariable logistic regression models were constructed. Adjusted odds ratios (aOR) with 95% CIs and p-values were reported. A p-value <0.05 was considered statistically significant. All analyses were conducted using jamovi statistical software. Sensitivity analyses involving log-transformed continuous variables, enrollment tertiles, and binary funding classification were performed using Python (statsmodels v0.14, scipy v1.11).
This study analyzed publicly available, de-identified data obtained from ClinicalTrials.gov. Because no human subjects were directly involved and no identifiable private information was used, institutional review board approval and informed consent were not required. The study complies with the principles of the Declaration of Helsinki.
3. Results
We analyzed 538 interventional clinical trials focusing on AF. Of these, 423 trials (78.6%) were completed, whereas 115 (21.4%) were discontinued before reaching completion. In terms of dissemination, only 153 of the 423 completed trials (36.2%) had their results published (publication analysis was restricted to completed trials only; discontinued trials were excluded from this denominator), while 270 trials (63.8%) remained unpublished. The majority of trials enrolled adult participants (97.0%) and included both male and female individuals (98.0%). Regarding sample size, 42.5% of trials enrolled fewer than 100 participants, whereas 57.5% enrolled 100 or more. Most studies were funded by non-specific or institutional sources (56.5%), followed by industry (39.0%), mixed sources (3.3%), and federal funding (1.1%). A large proportion of trials were randomized (92.8%). The parallel group model was the most commonly used intervention model (80.6%), with fewer trials using single-group (14.0%), crossover (4.2%), factorial (0.8%), or sequential (0.6%) designs. More than half of the studies were open-label (51.3%), while the rest employed quadruple (15.6%), double (12.4%), single (12.4%), or triple (8.3%) masking. Over half of the trials (52.1%) were single-center, and the remaining 47.9% were conducted across multiple centers. Drug-based interventions predominated (72.3%), followed by device-based (17.1%) and procedural interventions (14.5%), with biological, genetic, dietary, and behavioral interventions comprising a small minority (Table 1).
Table 1.
Characteristics of completed vs discontinued trials and published vs nonpublished trials (n=538).
| Characteristic | Total | Completion | p-value | Publication | p-value | ||
|---|---|---|---|---|---|---|---|
| Completed (n = 423, 78.6%) | Discontinued (n = 115, 21.4%) | Published (n = 153, 36.2%) | Unpublished (n = 270, 63.8%) | ||||
| Sex | 0.569 | 0.565 | |||||
| Female | 2 (0.4) | 2 (0.5) | 0 (0.0) | 0 (0.0) | 2 (0.7) | ||
| Male | 9 (1.7) | 8 (1.9) | 1 (0.9) | 3 (2.0) | 5 (1.9) | ||
| Female+male | 527 (98.0) | 413 (97.6) | 114 (99.1) | 150 (98.0) | 263 (97.4) | ||
| Age group | 0.379 | 0.083 | |||||
| Adult | 522 (97.0) | 409 (96.7) | 113 (98.3) | 151 (98.7) | 258 (95.6) | ||
| Adults+children | 16 (3.0) | 14 (3.3) | 2 (1.7) | 2 (1.3) | 12 (4.4) | ||
| Sample size | <0.001 | <0.001 | |||||
| <100 | 225 (42.5) | 147 (35.3) | 78 (69.0) | 24 (15.8) | 123 (46.4) | ||
| ≥100 | 305 (57.5) | 270 (64.7) | 35 (31.0) | 128 (84.2) | 142 (53.6) | ||
| Funding | 0.295 | 0.080 | |||||
| Federal | 6 (1.1) | 5 (1.2) | 1 (0.9) | 4 (2.6) | 1 (0.4) | ||
| Industry | 210 (39.0) | 168 (39.7) | 42 (36.5) | 60 (39.2) | 108 (40.0) | ||
| Mixed | 18 (3.3) | 17 (4.0) | 1 (0.9) | 3 (2.0) | 14 (5.2) | ||
| Other | 304 (56.5) | 233 (55.1) | 71 (61.7) | 86 (56.2) | 147 (54.4) | ||
| Number of centers | 0.564 | <0.001 | |||||
| Single center | 259 (52.1) | 211 (52.8) | 48 (49.5) | 55 (37.4) | 156 (61.7) | ||
| Multicenter | 238 (47.9) | 189 (47.3) | 49 (50.5) | 92 (62.6) | 97 (38.3) | ||
| Drug | 0.503 | 0.445 | |||||
| No | 149 (27.7) | 120 (28.4) | 29 (25.2) | 40 (26.1) | 80 (29.6) | ||
| Yes | 389 (72.3) | 303 (71.6) | 86 (74.8) | 113 (73.9) | 190 (70.4) | ||
| Behavioral | 0.241 | 0.265 | |||||
| No | 533 (99.1) | 418 (98.8) | 115 (100.0) | 150 (98.0) | 268 (99.3) | ||
| Yes | 5 (0.9) | 5 (1.2) | 0 (0.0) | 3 (2.0) | 2 (0.7) | ||
| Biological | 0.020 | 0.191 | |||||
| No | 531 (98.7) | 420 (99.3) | 111 (96.5) | 153 (100.0) | 267 (98.9) | ||
| Yes | 7 (1.3) | 3 (0.7) | 4 (3.5) | 0 (0.0) | 3 (1.1) | ||
| Device | 0.514 | 0.240 | |||||
| No | 446 (82.9) | 353 (83.5) | 93 (80.9) | 132 (86.3) | 221 (81.9) | ||
| Yes | 92 (17.1) | 70 (16.5) | 22 (19.1) | 21 (13.7) | 49 (18.1) | ||
| Dietary | 0.859 | 0.270 | |||||
| No | 534 (99.3) | 420 (99.3) | 114 (99.1) | 151 (98.7) | 269 (99.6) | ||
| Yes | 4 (0.7) | 3 (0.7) | 1 (0.9) | 2 (1.3) | 1 (0.4) | ||
| Genetic | 0.365 | 0.270 | |||||
| No | 535 (99.4) | 420 (99.3) | 115 (100.0) | 151 (98.7) | 269 (99.6) | ||
| Yes | 3 (0.6) | 3 (0.7) | 0 (0.0) | 2 (1.3) | 1 (0.4) | ||
| Procedure | 0.273 | 0.206 | |||||
| No | 460 (85.5) | 358 (84.6) | 102 (88.7) | 134 (87.6) | 224 (83.0) | ||
| Yes | 78 (14.5) | 65 (15.4) | 13 (11.3) | 19 (12.4) | 46 (17.0) | ||
| Other | 0.230 | 0.011 | |||||
| No | 488 (90.7) | 387 (91.5) | 101 (87.8) | 133 (86.9) | 254 (94.1) | ||
| Yes | 50 (9.3) | 36 (8.5) | 14 (12.2) | 20 (13.1) | 16 (5.9) | ||
| Allocation | 0.108 | 0.003 | |||||
| Randomized | 449 (92.8) | 345 (91.8) | 104 (96.3) | 136 (97.1) | 209 (88.6) | ||
| Non-randomized | 35 (7.2) | 31 (8.2) | 4 (3.7) | 4 (2.9) | 27 (11.4) | ||
| Intervention model | 0.089 | 0.022 | |||||
| Single | 74 (14.0) | 66 (15.8) | 8 (7.2) | 14 (9.3) | 52 (19.4) | ||
| Parallel | 427 (80.6) | 328 (78.3) | 99 (89.2) | 128 (84.8) | 200 (74.6) | ||
| Sequential | 3 (0.6) | 2 (0.5) | 1 (0.9) | 2 (1.3) | 0 (0.0) | ||
| Crossover | 22 (4.2) | 20 (4.8) | 2 (1.8) | 6 (4.0) | 14 (5.2) | ||
| Factorial | 4 (0.8) | 3 (0.7) | 1 (0.9) | 1 (0.7) | 2 (0.7) | ||
| Masking | 0.006 | 0.047 | |||||
| Open-label | 273 (51.3) | 225 (53.7) | 48 (42.5) | 78 (51.7) | 147 (54.9) | ||
| Single | 66 (12.4) | 49 (11.7) | 17 (15.0) | 21 (13.9) | 28 (10.4) | ||
| Double | 66 (12.4) | 58 (13.8) | 8 (7.1) | 17 (11.3) | 41 (15.3) | ||
| Triple | 44 (8.3) | 31 (7.4) | 13 (11.5) | 18 (11.9) | 13 (4.9) | ||
| Quadruple | 83 (15.6) | 56 (13.4) | 27 (23.9) | 17 (11.3) | 39 (14.6) | ||
| Phase | 0.522 | 0.002 | |||||
| 1 | 34 (6.3) | 25 (5.9) | 9 (7.8) | 3 (2.0) | 22 (8.1) | ||
| 2 | 109 (20.3) | 91 (21.5) | 18 (15.7) | 23 (15.0) | 68 (25.2) | ||
| 3 | 183 (34.0) | 142 (33.6) | 41 (35.7) | 62 (40.5) | 80 (29.6) | ||
| 4 | 212 (39.4) | 165 (39.0) | 47 (40.9) | 65 (42.5) | 100 (37.0) | ||
Data presented as n (%).
Of the 115 discontinued trials, reasons for discontinuation were available in the ClinicalTrials.gov registry for all trials based on their recorded study status. The most common category was termination (n = 84, 73.0%), followed by withdrawal (n = 29, 25.2%) and suspension (n = 2, 1.7%). Registry entries did not consistently provide granular reasons within these status categories (e.g., specific causes such as low accrual, funding withdrawal, or safety concerns), limiting further subgroup classification. These findings highlight the incomplete and non-standardized nature of discontinuation reporting in clinical trial registries.
When comparing completed and discontinued trials, several characteristics showed statistically significant differences. Trials enrolling ≥100 participants were significantly more likely to reach completion than smaller trials (64.7% vs 35.3%, p < 0.001). Trials incorporating biological interventions had a higher discontinuation rate than those without (3.5% vs 0.7%, p = 0.020). Moreover, quadruple masking was more frequently used in discontinued trials than completed ones (23.9% vs 13.4%, p = 0.006) (Table 1).
With regard to publication status, trials with larger sample sizes were considerably more likely to be published than those with fewer participants (84.2% vs 53.6%, p<0.001). Similarly, multicenter trials demonstrated higher publication rates than single-center trials (62.6% vs 37.4%, p<0.001). A higher proportion of trials that employed randomization were published compared with those that were non-randomized (97.1% vs 88.6%, p = 0.003). The parallel design model was more common among published trials (84.8% vs 74.6%, p = 0.022), whereas the single-group model was more prevalent in unpublished studies (19.4% vs 9.3%, p = 0.022). In terms of masking, double-masked trials were more likely to be unpublished than their counterparts (15.3% vs 11.3%, p = 0.047). Non-publication was also associated with the trial phase, as Phase 2 trials were more frequently unpublished (25.2% vs 15.0%, p = 0.002). Additionally, trials testing “other” types of interventions were more likely to be published (13.1% vs 5.9%, p = 0.011) (Table 1).
The estimated median time from primary study completion to publication was 18.0 months (95% CI, 14.0–21.0 months), indicating that 50% of published trials achieved publication within approximately 18 months after completion (Fig. 2).
Fig. 2.
Time to publication analysis using Kaplan–Meier estimation, depicting the cumulative probability of publication from primary study completion to publication (months).
In sensitivity analyses, treating sample size as a log-transformed continuous variable yielded consistent results: each unit increase in log-transformed enrollment was associated with significantly lower odds of discontinuation (OR: 0.629; 95% CI: 0.527–0.751; p<0.001) and non-publication (OR: 0.432; 95% CI: 0.352–0.530; p<0.001), confirming a dose-response relationship between trial size and both outcomes (Supplementary Table S1). Tertile analysis further supported this trend: discontinuation rates declined from 27.4% in the smallest tertile (≤75 participants) to 12.5% in the middle tertile (76–220 participants) and 11.4% in the largest tertile (>220 participants; p for trend <0.001). Similarly, non-publication rates fell from 82.8% in the smallest tertile to 74.1% in the middle and 37.2% in the largest tertile (p for trend <0.001) (Supplementary Table S2). A sensitivity analysis using binary funding classification (industry vs. non-industry) showed no significant association with discontinuation (OR: 0.873; 95% CI: 0.570–1.338; p = 0.534) or non-publication (OR: 1.033; 95% CI: 0.689–1.550; p = 0.874), consistent with the primary four-category model (Supplementary Table S3).
Multivariable logistic regression supported these findings. Trials with sample sizes of 100 or more had significantly lower odds of being discontinued (aOR: 0.193; 95% CI: 0.115–0.324; p<0.001) or remaining unpublished (aOR: 0.281; 95% CI: 0.166–0.476; p<0.001). Furthermore, multicenter trials were more than twice as likely as single-center trials to be discontinued (aOR: 2.43; 95% CI: 1.41–4.17; p = 0.001) and significantly less likely to remain unpublished (aOR: 0.311; 95% CI: 0.177–0.545; p<0.001). Other characteristics such as funding source and age group were not statistically significant predictors of trial discontinuation or publication in the adjusted models (Table 2, Table 3). Estimates derived from sparse covariate categories—specifically the pediatric-inclusive age group (3% of trials) and federal and mixed funding categories—should be interpreted with caution given their wide confidence intervals.
Table 2.
Logistic regression of trial discontinuation.
| Predictor | aOR (95% CI) | p-value |
|---|---|---|
| Sample size | ||
| ≥100 - <100 | 0.1932 (0.11536–0.324) | < 0.001 |
| Funder | ||
| Industry - federal | 0.2988 (0.02997–2.98) | 0.303 |
| Mixed - federal | 0.0772 (0.00358–1.662) | 0.102 |
| Other - federal | 0.6196 (0.06326–6.068) | 0.681 |
| Number of centers | ||
| Multicentric - Single center | 2.4287 (1.41497–4.169) | 0.001 |
| Age | ||
| Adult - adult+children | 0.7854 (0.15377–4.012) | 0.772 |
Model coefficients - discontinuation vs completion.
aOR, adjusted odds ratio; CI, confidence interval.
Table 3.
Logistic regression of trial non-publication.
| Predictor | aOR (95% CI) | p-value |
|---|---|---|
| Sample size | ||
| ≥100 - <100 | 0.281 (0.166–0.476) | <0.001 |
| Funder | ||
| Industry - federal | 6.926 (0.585–82.016) | 0.125 |
| Mixed - federal | 8.721 (0.535–142.167) | 0.128 |
| Other - federal | 3.005 (0.256–35.285) | 0.381 |
| Number of centers | ||
| Multicentric - Single center | 0.311 (0.177–0.545) | <0.001 |
| Age | ||
| Adult - adult+children | 4.968 (0.899–27.455) | 0.066 |
Model coefficients - non-publication vs publication.
aOR, adjusted odds ratio; CI, confidence interval.
4. Discussion
In this large, registry-based study of 538 interventional trials in AF, we found that more than one in five trials were discontinued and nearly two-thirds of completed trials remained unpublished. Larger trials (≥100 participants) exhibited markedly lower odds of discontinuation and non-publication, whereas multicenter trials, although more vulnerable to discontinuation, were more likely to yield published results. These findings highlight enduring challenges in clinical trial conduct and reporting in AF, with implications for evidence synthesis, ethics, and research efficiency.
4.1. Interpretation in the context of existing literature
The discontinuation rate of 21.4% in our AF cohort aligns with prior reports across various medical fields, which have found that 20–25% of registered trials were prematurely terminated [6], [7]. Similarly, the high proportion of unpublished AF trials is consistent with broader evidence demonstrating persistent non-publication in clinical research, despite policies mandating trial registration and results reporting [4], [8].
Our finding that larger trials are less likely to be discontinued or unpublished supports earlier work indicating that small trials face higher risks of recruitment failure, underfunding, and perceived lower impact, which decrease the likelihood of successful completion and publication [5], [9].
The relationship between multicenter design and outcomes is complex. Multicenter trials face increased logistical and regulatory burdens, which may explain their higher discontinuation rates [10]. However, when completed, their broader generalizability and stronger evidentiary weight increase the likelihood of publication [11].
Other associations in our analysis provide important insights. Double-masked trials showed higher rates of non-publication, potentially reflecting greater operational challenges and increased risk of inconclusive outcomes. The elevated non-publication rate among Phase 2 trials is consistent with the well-recognized “valley of death” in translational research, where early-phase studies frequently produce unfavorable or neutral results that remain unpublished [12]. Conversely, the higher publication rates of “other” intervention trials may reflect novelty or editorial interest, highlighting biases in publication practices.
4.2. Implications
Discontinuation and non-publication have ethical and scientific consequences. Participants in discontinued or unpublished trials are exposed to potential risks without contributing to generalizable knowledge, undermining trust in research [13]. Non-publication drives reporting bias, leading to inflated estimates of treatment efficacy and compromised clinical guidelines [8], [14].
Practical strategies to reduce discontinuation include feasibility assessments, improved recruitment planning, and adaptive trial designs. Predictive models using registry data have been developed to identify trials at high risk of poor accrual and early termination [15]. To address non-publication, stronger enforcement of reporting requirements, institutional incentives, and support for publishing negative or neutral results are urgently needed [16]. Dedicated journals or repositories for negative outcomes could mitigate selective reporting and enhance transparency [14].
4.3. Limitations
This study has several limitations. Reliance on ClinicalTrials.gov means the results depend on the accuracy of registry data, which may be incomplete or inconsistently updated. Registry entries rarely provide detailed reasons for discontinuation, limiting causal interpretation. Some trials may have been published without NCT identifiers, leading to underestimation of publication rates. Additionally, generalizability to other disease areas should be made cautiously, as patterns may vary across specialties. Furthermore, the apparent decline in the number of reported trials in recent years likely reflects publication lag and incomplete follow-up for recently registered trials, rather than a true reduction in trial initiation or reporting activity. The study period encompasses the COVID-19 pandemic (2020–2021), which likely disrupted trial enrollment and operations for an indeterminate number of AF trials; because ClinicalTrials.gov does not record cause of discontinuation as a structured field, pandemic-attributable trial failures cannot be reliably quantified, representing an important contextual limitation. Lastly, AF type (paroxysmal vs. persistent) was not consistently reported as a structured registry field, precluding reliable AF-subtype stratification; detailed procedural characteristics including ablation strategy and energy modality were similarly unavailable in structured form; geographic analysis was constrained by multi-country trial designations and uneven regional representation; and regression estimates for sparse covariate categories (pediatric-inclusive age group, federal and mixed funding) should be interpreted with caution given their wide confidence intervals.
4.4. Future directions
Future research should incorporate qualitative data to explore reasons for trial discontinuation and barriers to publication, and AF-type stratification should be a priority for future registry-based analyses, particularly as ClinicalTrials.gov continues to refine its structured data fields. Studies with standardized procedural coding or that use supplementary data sources (e.g., device approval registries or society-maintained trial databases) are needed to address the question of whether AF ablation modality influenced trial discontinuation or publication rates, and studies focused on geographic stratification, particularly in datasets enriched with region-specific trial infrastructure information or multinational trial network data, might also lend greater understanding to trial discontinuation and non-publication rates. Prospective monitoring and predictive analytics could allow earlier detection of at-risk trials. Additionally, interventions designed to promote dissemination of all outcomes—including null and negative findings—should be rigorously evaluated.
4.5. Conclusion
Our analysis reveals substantial discontinuation and non-publication in AF clinical trials, strongly influenced by trial size, design, and phase. To protect participants, optimize resource allocation, and uphold research integrity, AF trialists and stakeholders must prioritize robust planning, enforce reporting standards, and commit to transparent dissemination of all trial outcomes.
Authorship statement
All authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
CRediT authorship contribution statement
Own Khraisat: Writing – original draft, Visualization, Methodology, Formal analysis, Data curation, Conceptualization. Triston Messer: Writing – review & editing, Validation, Methodology. Khaled Harmouch: Writing – review & editing, Investigation, Data curation. Ahmad Alqaseer: Writing – review & editing, Investigation, Data curation. Waleed Ismail: Writing – review & editing, Validation. Hashem Samardali: Writing – review & editing, Resources, Investigation. Ehab Alkhawaldeh: Writing – review & editing, Supervision, Methodology. Amer Hammad: Writing – review & editing, Visualization, Software, Formal analysis. Dipal Patel: Writing – review & editing, Validation. Jason Sayanlar: Writing – review & editing, Resources. Muddasir Ashraf: Writing – review & editing, Supervision, Project administration. Ahmad Khraisat: Writing – review & editing, Project administration, Supervision, Conceptualization..
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The authors thank the following from Aurora Cardiovascular and Thoracic Services: Jennifer Pfaff and Sarah Kennedy for editorial preparation of the manuscript and Brian Schurrer for assistance with the figures.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.ijcha.2026.101949.
Appendix A. Supplementary data
The following are the Supplementary data to this article:
References
- 1.Chugh S.S., Havmoeller R., Narayanan K., et al. Worldwide epidemiology of atrial fibrillation: a Global Burden of Disease 2010 Study. Circulation. 2014;129:837–847. doi: 10.1161/circulationaha.113.005119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Krijthe B.P., Kunst A., Benjamin E.J., et al. Projections on the number of individuals with atrial fibrillation in the European Union, from 2000 to 2060. Eur. Heart J. 2013;34:2746–2751. doi: 10.1093/eurheartj/eht280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Colilla S., Crow A., Petkun W., Singer D.E., Simon T., Liu X. Estimates of current and future incidence and prevalence of atrial fibrillation in the U.S. adult population. Am. J. Cardiol. 2013;112:1142–1147. doi: 10.1016/j.amjcard.2013.05.063. [DOI] [PubMed] [Google Scholar]
- 4.Ross J.S., Mulvey G.K., Hines E.M., Nissen S.E., Krumholz H.M. Trial publication after registration in ClinicalTrials.Gov: a cross-sectional analysis. PLoS Med. 2009;6 doi: 10.1371/journal.pmed.1000144. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Carlisle B., Kimmelman J., Ramsay T., MacKinnon N. Unsuccessful trial accrual and human subjects protections: an empirical analysis of recently closed trials. Clin. Trials. 2015;12:77–83. doi: 10.1177/1740774514558307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Al-Durra M., Nolan R.P., Seto E., Cafazzo J.A., Eysenbach G. Nonpublication rates and characteristics of registered randomized clinical trials in digital health: cross-sectional analysis. J. Med. Internet Res. 2018;20 doi: 10.2196/11924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kasenda B., von Elm E., You J., et al. Prevalence, characteristics, and publication of discontinued randomized trials. JAMA. 2014;311:1045–1051. doi: 10.1001/jama.2014.1361. [DOI] [PubMed] [Google Scholar]
- 8.Chen R., Desai N.R., Ross J.S., et al. Publication and reporting of clinical trial results: cross sectional analysis across academic medical centers. BMJ. 2016;352 doi: 10.1136/bmj.i637. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Walters S.J., Bonacho Dos Anjos Henriques-Cadby I., Bortolami O., et al. Recruitment and retention of participants in randomised controlled trials: a review of trials funded and published by the United Kingdom Health Technology Assessment Programme. BMJ Open. 2017;7 doi: 10.1136/bmjopen-2016-015276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.McDonald A.M., Knight R.C., Campbell M.K., et al. What influences recruitment to randomised controlled trials? A review of trials funded by two UK funding agencies. Trials. 2006;7:9. doi: 10.1186/1745-6215-7-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ioannidis J.P. Effect of the statistical significance of results on the time to completion and publication of randomized efficacy trials. JAMA. 1998;279:281–286. doi: 10.1001/jama.279.4.281. [DOI] [PubMed] [Google Scholar]
- 12.Seruga B., Sterling L., Wang L., Tannock I.F. Reporting of serious adverse drug reactions of targeted anticancer agents in pivotal phase III clinical trials. J. Clin. Oncol. 2011;29:174–185. doi: 10.1200/jco.2010.31.9624. [DOI] [PubMed] [Google Scholar]
- 13.Emanuel E.J., Wendler D., Grady C. What makes clinical research ethical? JAMA. 2000;283:2701–2711. doi: 10.1001/jama.283.20.2701. [DOI] [PubMed] [Google Scholar]
- 14.De Angelis C., Drazen J.M., Frizelle F.A., et al. Clinical trial registration: a statement from the International Committee of Medical Journal Editors. N. Engl. J. Med. 2004;351:1250–1251. doi: 10.1056/NEJMe048225. [DOI] [PubMed] [Google Scholar]
- 15.Ma S., Wang Y., Wagner J., Johnson S., Pakhomov S., Aliferis C. Predicting accrual success for better clinical trial resource allocation. Sci. Rep. 2025;15:3879. doi: 10.1038/s41598-025-88400-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schmucker C., Schell L.K., Portalupi S., et al. Extent of non-publication in cohorts of studies approved by research ethics committees or included in trial registries. PLoS One. 2014;9 doi: 10.1371/journal.pone.0114023. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.