Assessment of transparency indicators across the biomedical literature: How open is open?

Recent concerns about the reproducibility of science have led to several calls for more open and transparent research practices and for the monitoring of potential improvements over time. However, with tens of thousands of new biomedical articles published per week, manually mapping and monitoring changes in transparency is unrealistic. We present an open-source, automated approach to identify 5 indicators of transparency (data sharing, code sharing, conflicts of interest disclosures, funding disclosures, and protocol registration) and apply it across the entire open access biomedical literature of 2.75 million articles on PubMed Central (PMC). Our results indicate remarkable improvements in some (e.g., conflict of interest [COI] disclosures and funding disclosures), but not other (e.g., protocol registration and code sharing) areas of transparency over time, and map transparency across fields of science, countries, journals, and publishers. This work has enabled the creation of a large, integrated, and openly available database to expedite further efforts to monitor, understand, and promote transparency and reproducibility in science.

Deep Learning for Epidemiologists: An Introduction to Neural Networks.

Deep learning methods are increasingly being applied to problems in medicine and health care. However, few epidemiologists have received formal training in these methods. To bridge this gap, this article introduces the fundamentals of deep learning from an epidemiologic perspective. Specifically, this article reviews core concepts in machine learning (e.g., overfitting, regularization, and hyperparameters); explains several fundamental deep learning architectures (convolutional neural networks, recurrent neural networks); and summarizes training, evaluation, and deployment of models. Conceptual understanding of supervised learning algorithms is the focus of the article; instructions on the training of deep learning models and applications of deep learning to causal learning are out of this article's scope. We aim to provide an accessible first step towards enabling the reader to read and assess research on the medical applications of deep learning and to familiarize readers with deep learning terminology and concepts to facilitate communication with computer scientists and machine learning engineers.

Identifying COVID-19 Vaccine Deserts and Ways to Reduce Them: A Digital Tool to Support Public Health Decision-Making.

A private-academic partnership built the Vaccine Equity Planner (VEP) to help decision-makers improve geographic access to COVID-19 vaccinations across the United States by identifying vaccine deserts and facilities that could fill those deserts. The VEP presented complex, updated data in an intuitive form during a rapidly changing pandemic situation. The persistence of vaccine deserts in every state as COVID-19 booster recommendations develop suggests that vaccine delivery can be improved. Underresourced public health systems benefit from tools providing real-time, accurate, actionable data. (Am J Public Health. 2023;113(4):363-367. https://doi.org/10.2105/AJPH.2022.307198).

Evaluation of confounding in epidemiologic studies assessing alcohol consumption on the risk of ischemic heart disease.

BACKGROUND: Among different investigators studying the same exposures and outcomes, there may be a lack of consensus about potential confounders that should be considered as matching, adjustment, or stratification variables in observational studies. Concerns have been raised that confounding factors may affect the results obtained for the alcohol-ischemic heart disease relationship, as well as their consistency and reproducibility across different studies. Therefore, we assessed how confounders are defined, operationalized, and discussed across individual studies evaluating the impact of alcohol on ischemic heart disease risk. METHODS: For observational studies included in a recent alcohol-ischemic heart disease meta-analysis, we identified all variables adjusted, matched, or stratified for in the largest reported multivariate model (i.e. potential confounders). We recorded how the variables were measured and grouped them into higher-level confounder domains. Abstracts and Discussion sections were then assessed to determine whether authors considered confounding when interpreting their study findings. RESULTS: 85 of 87 (97.7%) studies reported multivariate analyses for an alcohol-ischemic heart disease relationship. The most common higher-level confounder domains included were smoking (79, 92.9%), age (74, 87.1%), and BMI, height, and/or weight (57, 67.1%). However, no two models adjusted, matched, or stratified for the same higher-level confounder domains. Most (74/87, 85.1%) articles mentioned or alluded to "confounding" in their Abstract or Discussion sections, but only one stated that their main findings were likely to be affected by residual confounding. There were five (5/87, 5.7%) authors that explicitly asked for caution when interpreting results. CONCLUSION: There is large variation in the confounders considered across observational studies evaluating the impact of alcohol on ischemic heart disease risk and almost all studies spuriously ignore or eventually dismiss confounding in their conclusions. Given that study results and interpretations may be affected by the mix of potential confounders included within multivariate models, efforts are necessary to standardize approaches for selecting and accounting for confounders in observational studies.

Variation in Interleukin 6 Receptor Gene Associates With Risk of Crohn's Disease and Ulcerative Colitis.

Interleukin 6 (IL6) is an inflammatory cytokine; signaling via its receptor (IL6R) is believed to contribute to development of inflammatory bowel diseases (IBD). The single nucleotide polymorphism rs2228145 in IL6R associates with increased levels of soluble IL6R (s-IL6R), as well as reduced IL6R signaling and risk of inflammatory disorders; its effects are similar to those of a therapeutic monoclonal antibody that blocks IL6R signaling. We used the effect of rs2228145 on s-IL6R level as an indirect marker to investigate whether reduced IL6R signaling associates with risk of ulcerative colitis (UC) or Crohn's disease (CD). In a genome-wide meta-analysis of 20,550 patients with CD, 17,647 patients with UC, and more than 40,000 individuals without IBD (controls), we found that rs2228145 (scaled to a 2-fold increase in s-IL6R) was associated with reduced risk of CD (odds ratio 0.876; 95% confidence interval 0.822-0.933; P = .00003) or UC (odds ratio 0.932; 95% confidence interval 0.875-0.996; P = .036). These findings indicate that therapeutics designed to block IL6R signaling might be effective in treatment of IBD.

Correction: Risk of Bias in Reports of In Vivo Research: A Focus for Improvement.

[This corrects the article DOI: 10.1371/journal.pbio.1002273.].

Risk of Bias in Reports of In Vivo Research: A Focus for Improvement.

The reliability of experimental findings depends on the rigour of experimental design. Here we show limited reporting of measures to reduce the risk of bias in a random sample of life sciences publications, significantly lower reporting of randomisation in work published in journals of high impact, and very limited reporting of measures to reduce the risk of bias in publications from leading United Kingdom institutions. Ascertainment of differences between institutions might serve both as a measure of research quality and as a tool for institutional efforts to improve research quality.

Long noncoding RNAs as novel predictors of survival in human cancer: a systematic review and meta-analysis.

BACKGROUND: Expression of various long noncoding RNAs (lncRNAs) may affect cancer prognosis. Here, we aim to gather and examine all evidence on the potential role of lncRNAs as novel predictors of survival in human cancer. METHODS: We systematically searched through PubMed, to identify all published studies reporting on the association between any individual lncRNA or group of lncRNAs with prognosis in human cancer (death or other clinical outcomes). Where appropriate, we then performed quantitative synthesis of those results using meta-analytic methods to identify the true effect size of lncRNAs on cancer prognosis. The reliability of those results was then examined using measures of heterogeneity and testing for selective reporting biases. RESULTS: Three hundred ninety-two studies were screened to eventually identify 111 eligible studies on 127 datasets. In total, these represented 16,754 independent participants pertaining to 53 individual and 6 grouped lncRNAs within a total of 19 cancer sites. Overall, 83 % of the studies we identified addressed overall survival and 32 % of the studies addressed recurrence-free survival. For overall survival, 96 % (88/92) of studies identified a statistically significant association of lncRNA expression to prognosis. Meta-analysis of 6 out of 7 lncRNAs for which three or more studies were available, identified statistically significant associations with overall survival. The lncRNA HOTAIR was by far the most broadly studied lncRNA (n = 29; of 111 studies) and featured a summary hazard ratio (HR) of 2.22 (95 % confidence interval (CI), 1.86-2.65) with modest heterogeneity (I(2) = 49 %; 95 % CI, 14-79 %). Prominent excess significance was demonstrated across all meta-analyses (p-value = 0.0003), raising the possibility of substantial selective reporting biases. CONCLUSIONS: Multiple lncRNAs have been shown to be strongly associated with prognosis in diverse cancers, but substantial bias cannot be excluded in this field and larger studies are needed to understand whether these prognostic information may eventually be useful.

Comparative assessment of phototherapy protocols for reduction of oxidative stress in partially transected spinal cord slices undergoing secondary degeneration.

BACKGROUND: Red/near-infrared light therapy (R/NIR-LT) has been developed as a treatment for a range of conditions, including injury to the central nervous system (CNS). However, clinical trials have reported variable or sub-optimal outcomes, possibly because there are few optimized treatment protocols for the different target tissues. Moreover, the low absolute, and wavelength dependent, transmission of light by tissues overlying the target site make accurate dosing problematic. RESULTS: In order to optimize light therapy treatment parameters, we adapted a mouse spinal cord organotypic culture model to the rat, and characterized myelination and oxidative stress following a partial transection injury. The ex vivo model allows a more accurate assessment of the relative effect of different illumination wavelengths (adjusted for equal quantal intensity) on the target tissue. Using this model, we assessed oxidative stress following treatment with four different wavelengths of light: 450 nm (blue); 510 nm (green); 660 nm (red) or 860 nm (infrared) at three different intensities: 1.93 × 10(16) (low); 3.85 × 10(16) (intermediate) and 7.70 × 10(16) (high) photons/cm(2)/s. We demonstrate that the most effective of the tested wavelengths to reduce immunoreactivity of the oxidative stress indicator 3-nitrotyrosine (3NT) was 660 nm. 860 nm also provided beneficial effects at all tested intensities, significantly reducing oxidative stress levels relative to control (p ≤ 0.05). CONCLUSIONS: Our results indicate that R/NIR-LT is an effective antioxidant therapy, and indicate that effective wavelengths and ranges of intensities of treatment can be adapted for a variety of CNS injuries and conditions, depending upon the transmission properties of the tissue to be treated.

Quantifying urban park use in the USA at scale: empirical estimates of realised park usage using smartphone location data.

BACKGROUND: A large body of evidence connects access to greenspace with substantial benefits to physical and mental health. In urban settings where access to greenspace can be limited, park access and use have been associated with higher levels of physical activity, improved physical health, and lower levels of markers of mental distress. Despite the potential health benefits of urban parks, little is known about how park usage varies across locations (between or within cities) or over time. METHODS: We estimated park usage among urban residents (identified as residents of urban census tracts) in 498 US cities from 2019 to 2021 from aggregated and anonymised opted-in smartphone location history data. We used descriptive statistics to quantify differences in park usage over time, between cities, and across census tracts within cities, and used generalised linear models to estimate the associations between park usage and census tract level descriptors. FINDINGS: In spring (March 1 to May 31) 2019, 18·9% of urban residents visited a park at least once per week, with average use higher in northwest and southwest USA, and lowest in the southeast. Park usage varied substantially both within and between cities; was unequally distributed across census tract-level markers of race, ethnicity, income, and social vulnerability; and was only moderately correlated with established markers of census tract greenspace. In spring 2019, a doubling of walking time to parks was associated with a 10·1% (95% CI 5·6-14·3) lower average weekly park usage, adjusting for city and social vulnerability index. The median decline in park usage from spring 2019 to spring 2020 was 38·0% (IQR 28·4-46·5), coincident with the onset of physical distancing policies across much of the country. We estimated that the COVID-19-related decline in park usage was more pronounced for those living further from a park and those living in areas of higher social vulnerability. INTERPRETATION: These estimates provide novel insights into the patterns and correlates of park use and could enable new studies of the health benefits of urban greenspace. In addition, the availability of an empirical park usage metric that varies over time could be a useful tool for assessing the effectiveness of policies intended to increase such activities. FUNDING: Google.

Systemic Complement Activation Profiles in Nonexudative Age-Related Macular Degeneration: A Systematic Review.

Topic: To evaluate whether differences exist in systemic complement activation profiles in patients with early to intermediate nonexudative age-related macular degeneration (AMD) or geographic atrophy (GA) compared with control participants without AMD. Clinical Relevance: Complement inhibition has emerged as a therapeutic strategy for GA, although clinical trials to date have yielded mixed results. Despite these efforts, no clear consensus exists regarding what portions of the complement pathway are dysregulated in AMD or when this dysregulation occurs relative to AMD stage. Although past studies have compared systemic complement activation profiles in patients with AMD versus in control participants without AMD, differences in AMD case definition and differing analytical approaches complicate their interpretation. Methods: We performed a systematic review by identifying articles from database inception through October 11, 2020, that reported systemic complement activation profiles in patients with early or intermediate nonexudative AMD or GA versus control participants without AMD by searching PubMed, Google Scholar, and Embase. Risk of bias was assessed using a modified Newcastle-Ottawa score. Results: The 8 reviewed studies included 2131 independent participants. Most studies report significantly higher systemic levels of products associated with complement activation and significantly lower systemic levels of products associated with complement inhibition in patients with early and advanced nonexudative AMD compared with control participants without AMD. Discussion: Evidence suggests that systemic complement overactivation is a feature of early or intermediate and advanced nonexudative AMD. However, given significant heterogeneity, these findings are not conclusive and warrant further investigation.

Systemic Complement Activation Profiles in Nonexudative Age-Related Macular Degeneration: A Meta-Analysis.

Although complement inhibition has emerged as a possible therapeutic strategy for age-related macular degeneration (AMD), there is not a clear consensus regarding what aspects of the complement pathway are dysregulated in AMD and when this occurs relative to disease stage. We recently published a systematic review describing systemic complement activation profiles in patients with early/intermediate AMD or geographic atrophy (GA) compared to non-AMD controls. Here, we sought to meta-analyze these results to estimate the magnitude of complement dysregulation in AMD using restricted maximum likelihood estimation. The seven meta-analyzed studies included 710 independent participants with 23 effect sizes. Compared with non-AMD controls, patients with early/intermediate nonexudative AMD (N = 246) had significantly higher systemic complement activation, as quantified by the levels of complement proteins generated by common final pathway activation, and significantly lower systemic complement inhibition. In contrast, there were no statistically significant differences in the systemic levels of complement common final pathway activation products or complement inhibition in patients with GA (N = 178) versus non-AMD controls. We provide evidence that systemic complement over-activation is a feature of early/intermediate nonexudative AMD; no such evidence was identified for patients with GA. These findings provide mechanistic insights and inform future clinical trials.

Media and social media attention to retracted articles according to Altmetric.

The number of retracted articles has grown fast. However, the extent to which researchers and the public are made adequately aware of these retractions and how the media and social media respond to them remains unknown. Here, we aimed to evaluate the media and social media attention received by retracted articles and assess also the attention they receive post-retraction versus pre-retraction. We downloaded all records of retracted literature maintained by the Retraction Watch Database and originally published between January 1, 2010 to December 31, 2015. For all 3,008 retracted articles with a separate DOI for the original and its retraction, we downloaded the respective Altmetric Attention Score (AAS) (from Altmetric) and citation count (from Crossref), for the original article and its retraction notice on June 6, 2018. We also compared the AAS of a random sample of 572 retracted full journal articles available on PubMed to that of unretracted full articles matched from the same issue and journal. 1,687 (56.1%) of retracted research articles received some amount of Altmetric attention, and 165 (5.5%) were even considered popular (AAS>20). 31 (1.0%) of 2,953 with a record on Crossref received >100 citations by June 6, 2018. Popular articles received substantially more attention than their retraction, even after adjusting for attention received post-retraction (Median difference, 29; 95% CI, 17-61). Unreliable results were the most frequent reason for retraction of popular articles (32; 19%), while fake peer review was the most common reason (421; 15%) for the retraction of other articles. In comparison to matched articles, retracted articles tended to receive more Altmetric attention (23/31 matched groups; P-value, 0.01), even after adjusting for attention received post-retraction. Our findings reveal that retracted articles may receive high attention from media and social media and that for popular articles, pre-retraction attention far outweighs post-retraction attention.

Large Pediatric Randomized Clinical Trials in ClinicalTrials.gov.

BACKGROUND: Large, randomized controlled trials (RCTs) are essential in answering pivotal questions in child health. METHODS: We created a bird's eye view of all large, noncluster, nonvaccine pediatric RCTs with ≥1000 participants registered in ClinicalTrials.gov (last search January 9, 2020). We analyzed the funding sources, countries, outcomes, publication status, and correlation with the pediatric global burden of disease (GBD) for eligible trials. RESULTS: We identified 247 large, nonvaccine, noncluster pediatric RCTs. Only 17 mega-trials with ≥5000 participants existed. Industry funding was involved in only 52 (21%) and exclusively funded 47 (19%) trials. Participants were from high-income countries (HICs) in 100 (40%) trials, from lower-middle-income countries (LMICs) in 122 (49%) trials, and from both HICs and LMICs in 19 (8%) trials; 6 trials did not report participants' country location. Of trials conducted in LMIC, 43% of investigators were from HICs. Of non-LMIC participants trials (HIC or HIC and LMIC), 39% were multicountry trials versus 11% of exclusively LMIC participants trials. Few trials (18%; 44 of 247) targeted mortality as an outcome. 35% (58 of 164) of the trials completed ≥12 months were unpublished at the time of our assessment. The number of trials per disease category correlated well with pediatric GBD overall (ρ = 0.76) and in LMICs (ρ = 0.69), but not in HICs (ρ = 0.29). CONCLUSIONS: Incentivization of investigator collaborations across diverse country settings, timely publication of results of large pediatric RCTs, and alignment with the pediatric GBD are of pivotal importance to ultimately improve child health globally.

Privacy-first health research with federated learning.

Privacy protection is paramount in conducting health research. However, studies often rely on data stored in a centralized repository, where analysis is done with full access to the sensitive underlying content. Recent advances in federated learning enable building complex machine-learned models that are trained in a distributed fashion. These techniques facilitate the calculation of research study endpoints such that private data never leaves a given device or healthcare system. We show-on a diverse set of single and multi-site health studies-that federated models can achieve similar accuracy, precision, and generalizability, and lead to the same interpretation as standard centralized statistical models while achieving considerably stronger privacy protections and without significantly raising computational costs. This work is the first to apply modern and general federated learning methods that explicitly incorporate differential privacy to clinical and epidemiological research-across a spectrum of units of federation, model architectures, complexity of learning tasks and diseases. As a result, it enables health research participants to remain in control of their data and still contribute to advancing science-aspects that used to be at odds with each other.

Antenatal imaging and clinical outcome in congenital CMV infection: A field-wide systematic review and meta-analysis.

OBJECTIVES: Postnatal outcome in fetuses with congenital cytomegalovirus infection (cCMV) varies from asymptomatic infection to severe neurodevelopmental impairment. Αntenatal biomarkers of long-term clinical outcome, have yet to be established. Α systematic review and meta-analysis was performed to examine whether prenatal cerebral ultrasonography (US) and magnetic resonance imaging (MRI) findings in cCMV fetuses may predict clinical outcome. METHODS: PubMed and the Web of Science were systematically searched to identify studies reporting on any prenatal US and/or MRI imaging of fetuses with cCMV as well as their postnatal clinical outcome. All reported associations between imaging and postnatal clinical outcome were systematically extracted. Where appropriate, the reported associations were quantitatively synthesized within Bayesian random-effects meta-analyses. RESULTS: A total of 1336 studies were screened to identify 26 eligible observational studies. Overall, 4181 fetuses were studied, of which 1518 had been diagnosed with cCMV. All studies performed fetal US while in 14 (54%) MRI was also performed. Studies substantially varied in timing of fetal imaging, reporting of abnormalities, definition of poor outcome and statistical analysis. Among studies reporting on statistical significance, 6/6 for US and 3/4 for MRI identified significant associations between imaging findings and outcome. In our meta-analyses, within isolated abnormalities, only microcephaly had greater than 95% probability of being associated with poor outcome (OR 26.7; 95% CI, 1.44-1464.5; I2, 19%). Effect sizes for US were higher than those for MRI findings. CONCLUSIONS: Although studies displayed significant heterogeneity in both methodology and analytical decisions, it became evident that when both prenatal cerebral US and MRI are normal the negative predictive value of poor outcome is high. This is important for clinicians when consulting pregnant women. Need to standardize practices and definitions become evident. FUNDING: There was no source of funding.

New clinical trial designs in the era of precision medicine: An overview of definitions, strengths, weaknesses, and current use in oncology.

With expanding knowledge in tumor biology and biomarkers, oncology therapies are increasingly moving away from the "one-size-fits-all" rationale onto biomarker-driven therapies tailored according to patient-specific characteristics, most commonly the tumor's molecular profile. The advent of precision medicine in oncology has been accompanied by the introduction of novel clinical trial designs that aim to identify biomarker-matched subgroups of patients that will benefit the most from targeted therapies. This innovation comes with the promise of answering more treatment questions, more efficiently and in less time. In this article, we give an overview of the different biomarker-based designs, comparing the features of enrichment, randomize-all, umbrella, and basket trials, and highlighting their advantages and disadvantages. We focus more on the novel designs known as master protocols, which include umbrella and basket trials. We have also conducted a search in ClinicalTrials.gov for registered oncology-related protocols of ongoing or completed trials labeled as umbrella or basket trials for solid tumors; we also included additional relevant trials retrieved from other reviews. We present and discuss the key features of the 30 eligible basket trials and 27 eligible umbrella trials. Only a minority of them are randomized (2 and 9, respectively), including three trials with adaptive randomization. Five of these trials have been completed as of July 2018. Precision medicine trial designs fuel new hopes for identifying best treatments, but there is also the potential for hype. The benefits and challenges associated with their use will need continued monitoring.