Dissemination and Implementation Research at the National Cancer Institute: A Review of Funded Studies (2006-2019) and Opportunities to Advance the Field.

BACKGROUND: To ensure investment in cancer research reaches populations who can benefit, the NCI has funded implementation science grants since the Dissemination and Implementation Research in Health (DIRH) funding opportunities launched in 2006. We analyzed NCI-funded DIRH grants to provide a snapshot of implementation science conducted across the cancer care continuum and highlight areas ripe for exploration. METHODS: NCI-funded DIRH grants between fiscal years 2006 and 2019 were identified using the iSearch database. Two coders classified each grant by topic, populations studied, intervention and setting characteristics, strategies tested, study designs and methods used, and outcomes measured. RESULTS: Seventy-one grants were awarded addressing cancer prevention (n = 33), screening (n = 33), diagnosis (n = 4), treatment (n = 9), and/or survivorship (n = 11). Colorectal (n = 20), breast (n = 15), and cervical (n = 11) were the most studied cancers. Most grants focused on delivery of guidelines (n = 36) or behavioral change interventions (n = 18) in health care settings (n = 47), studying implementation processes (n = 37) and/or testing implementation strategies (n = 43) using experimental (n = 35) and quasi-experimental (n = 6) designs. Few studied sustainability (n = 7), scale-up (n = 2), deimplementation (n = 4), measure development (n = 6), or policy-level implementation (n = 6). CONCLUSIONS: Current funding suggests researchers are studying implementation of cancer control interventions across the care continuum. However, research gaps remain in strategies for sustainability, scale-up, and deimplementation. More emphasis is needed on cancer treatment and survivorship. Additional focus on policy implementation and measure development is warranted. IMPACT: Understanding the breadth of NCI-funded implementation science can inform future efforts to build the knowledge base on how to improve dissemination and implementation of evidence in cancer control.

A Systematic Literature Review of Whole Exome and Genome Sequencing Population Studies of Genetic Susceptibility to Cancer.

The application of next-generation sequencing (NGS) technologies in cancer research has accelerated the discovery of somatic mutations; however, progress in the identification of germline variation associated with cancer risk is less clear. We conducted a systematic literature review of cancer genetic susceptibility studies that used NGS technologies at an exome/genome-wide scale to obtain a fuller understanding of the research landscape to date and to inform future studies. The variability across studies on methodologies and reporting was considerable. Most studies sequenced few high-risk (mainly European) families, used a candidate analysis approach, and identified potential cancer-related germline variants or genes in a small fraction of the sequenced cancer cases. This review highlights the importance of establishing consensus on standards for the application and reporting of variants filtering strategies. It also describes the progress in the identification of cancer-related germline variation to date. These findings point to the untapped potential in conducting studies with appropriately sized and racially diverse families and populations, combining results across studies and expanding beyond a candidate analysis approach to advance the discovery of genetic variation that accounts for the unexplained cancer heritability.

Building capacity: a cross-sectional evaluation of the US Training Institute for Dissemination and Implementation Research in Health.

BACKGROUND: In 2011, the National Institute of Health (NIH) initiated the Training in Dissemination and Implementation Research in Health (TIDIRH) program. Over its first 5 years, TIDIRH provided an in-person, week-long training to 197 investigators who were new to the dissemination and implementation (D&I) field. This paper evaluates the long-term impact of TIDIRH on trainees' use of D&I methods, collaborations, and research funding. METHODS: Trainees were selected to participate through a competitive process. We compared the 197 trainees to 125 unselected applicants (UAs) whose application score was within one standard deviation of the mean for all trainees' scores for the same application year. A portfolio analysis examined electronic applications for NIH peer-reviewed funding submitted by trainees and UAs between 2011 and 2019. A survey of trainees and UAs was conducted in 2016, as was a faculty survey among the 87 individuals who served as TIDIRH instructors. RESULTS: A major goal of TIDIRH was to build the field, at least in part through networking and collaboration. Thirty-eight percent of trainees indicated they had extensive contact with faculty following the training, and an additional 38% indicated they had at least limited contact. Twenty-four percent of trainees had extensive collaboration with other fellows post-TIDIRH, and 43% had at least limited contact. Collaborative activities included the full range of academic activities, including manuscript development, grant writing, and consultation/collaboration on research studies. The portfolio analysis combining grant mechanisms showed that overall, TIDIRH trainees submitted more peer-reviewed NIH grants per person than UA and had significantly better funding outcomes (25% vs 19% funded, respectively). The greatest difference was for large research project, program/center, and cooperative agreement grants mechanisms. CONCLUSIONS: Overall, this evaluation found that TIDIRH is achieving its three primary goals: (1) building a pipeline of D&I investigators, (2) creating a network of scholars to build the field, and (3) improving funding outcomes for D&I grants.

Tracking human genes along the translational continuum.

Understanding the drivers of research on human genes is a critical component to success of translation efforts of genomics into medicine and public health. Using publicly available curated online databases we sought to identify specific genes that are featured in translational genetic research in comparison to all genomics research publications. Articles in the CDC's Public Health Genomics and Precision Health Knowledge Base were stratified into studies that have moved beyond basic research to population and clinical epidemiologic studies (T1: clinical and population human genome epidemiology research), and studies that evaluate, implement, and assess impact of genes in clinical and public health areas (T2+: beyond bench to bedside). We examined gene counts and numbers of publications within these phases of translation in comparison to all genes from Medline. We are able to highlight those genes that are moving from basic research to clinical and public health translational research, namely in cancer and a few genetic diseases with high penetrance and clinical actionability. Identifying human genes of translational value is an important step towards determining an evidence-based trajectory of the human genome in clinical and public health practice over time.

The current state of funded NIH grants in implementation science in genomic medicine: a portfolio analysis.

PURPOSE: Implementation science offers methods to evaluate the translation of genomic medicine research into practice. The extent to which the National Institutes of Health (NIH) human genomics grant portfolio includes implementation science is unknown. This brief report's objective is to describe recently funded implementation science studies in genomic medicine in the NIH grant portfolio, and identify remaining gaps. METHODS: We identified investigator-initiated NIH research grants on implementation science in genomic medicine (funding initiated 2012-2016). A codebook was adapted from the literature, three authors coded grants, and descriptive statistics were calculated for each code. RESULTS: Forty-two grants fit the inclusion criteria (~1.75% of investigator-initiated genomics grants). The majority of included grants proposed qualitative and/or quantitative methods with cross-sectional study designs, and described clinical settings and primarily white, non-Hispanic study populations. Most grants were in oncology and examined genetic testing for risk assessment. Finally, grants lacked the use of implementation science frameworks, and most examined uptake of genomic medicine and/or assessed patient-centeredness. CONCLUSION: We identified large gaps in implementation science studies in genomic medicine in the funded NIH portfolio over the past 5 years. To move the genomics field forward, investigator-initiated research grants should employ rigorous implementation science methods within diverse settings and populations.

Using deep learning to identify translational research in genomic medicine beyond bench to bedside.

Tracking scientific research publications on the evaluation, utility and implementation of genomic applications is critical for the translation of basic research to impact clinical and population health. In this work, we utilize state-of-the-art machine learning approaches to identify translational research in genomics beyond bench to bedside from the biomedical literature. We apply the convolutional neural networks (CNNs) and support vector machines (SVMs) to the bench/bedside article classification on the weekly manual annotation data of the Public Health Genomics Knowledge Base database. Both classifiers employ salient features to determine the probability of curation-eligible publications, which can effectively reduce the workload of manual triage and curation process. We applied the CNNs and SVMs to an independent test set (n = 400), and the models achieved the F-measure of 0.80 and 0.74, respectively. We further tested the CNNs, which perform better results, on the routine annotation pipeline for 2 weeks and significantly reduced the effort and retrieved more appropriate research articles. Our approaches provide direct insight into the automated curation of genomic translational research beyond bench to bedside. The machine learning classifiers are found to be helpful for annotators to enhance the efficiency of manual curation.

HLBS-PopOmics: an online knowledge base to accelerate dissemination and implementation of research advances in population genomics to reduce the burden of heart, lung, blood, and sleep disorders.

Recent dramatic advances in multiomics research coupled with exponentially increasing volume, complexity, and interdisciplinary nature of publications are making it challenging for scientists to stay up-to-date on the literature. Strategies to address this challenge include the creation of online databases and warehouses to support timely and targeted dissemination of research findings. Although most of the early examples have been in cancer genomics and pharmacogenomics, the approaches used can be adapted to support investigators in heart, lung, blood, and sleep (HLBS) disorders research. In this article, we describe the creation of an HLBS population genomics (HLBS-PopOmics) knowledge base as an online, continuously updated, searchable database to support the dissemination and implementation of studies and resources that are relevant to clinical and public health practice. In addition to targeted searches based on the HLBS disease categories, cross-cutting themes reflecting the ethical, legal, and social implications of genomics research; systematic evidence reviews; and clinical practice guidelines supporting screening, detection, evaluation, and treatment are also emphasized in HLBS-PopOmics. Future updates of the knowledge base will include additional emphasis on transcriptomics, proteomics, metabolomics, and other omics research; explore opportunities for leveraging data sets designed to support scientific discovery; and incorporate advanced machine learning bioinformatics capabilities.

Proposed outcomes measures for state public health genomic programs.

PURPOSE: To assess the implementation of evidence-based genomic medicine and its population-level impact on health outcomes and to promote public health genetics interventions, in 2015 the Roundtable on Genomics and Precision Health of the National Academies of Sciences, Engineering, and Medicine formed an action collaborative, the Genomics and Public Health Action Collaborative (GPHAC). This group engaged key stakeholders from public/population health agencies, along with experts in the fields of health disparities, health literacy, implementation science, medical genetics, and patient advocacy. METHODS: In this paper, we present the efforts to identify performance objectives and outcome metrics. Specific attention is placed on measures related to hereditary breast ovarian cancer (HBOC) syndrome and Lynch syndrome (LS), two conditions with existing evidence-based genomic applications that can have immediate impact on morbidity and mortality. RESULTS: Our assessment revealed few existing outcome measures. Therefore, using an implementation research framework, 38 outcome measures were crafted. CONCLUSION: Evidence-based public health requires outcome metrics, yet few exist for genomics. Therefore, we have proposed performance objectives that states might use and provided examples of a few state-level activities already under way, which are designed to collect outcome measures for HBOC and LS.

From public health genomics to precision public health: a 20-year journey.

In this paper, we review the evolution of the field of public health genomics in the United States in the past two decades. Public health genomics focuses on effective and responsible translation of genomic science into population health benefits. We discuss the relationship of the field to the core public health functions and essential services, review its evidentiary foundation, and provide examples of current US public health priorities and applications. We cite examples of publications to illustrate how Genetics in Medicine reflected the evolution of the field. We also reflect on how public-health genomics is contributing to the emergence of "precision public health" with near-term opportunities offered by the US Precision Medicine (AllofUs) Initiative.

Trends in published meta-analyses in cancer research, 2008-2013.

In order to capture trends in the contribution of epidemiology to cancer research, we describe an online meta-analysis database resource for cancer clinical and population research and illustrate trends and descriptive detail of cancer meta-analyses from 2008 through 2013. A total of 4,686 cancer meta-analyses met our inclusion criteria. During this 6-year period, a fivefold increase was observed in the yearly number of meta-analyses. Fifty-six percent of meta-analyses concerned observational studies, mostly of cancer risk, more than half of which were genetic studies. The major cancer sites were breast, colorectal, and digestive. This online database for Cancer Genomics and Epidemiology Navigator will be continuously updated to allow investigators to quickly navigate the meta-analyses emerging from cancer epidemiology studies and cancer clinical trials.

A knowledge base for tracking the impact of genomics on population health.

PURPOSE: We created an online knowledge base (the Public Health Genomics Knowledge Base (PHGKB)) to provide systematically curated and updated information that bridges population-based research on genomics with clinical and public health applications. METHODS: Weekly horizon scanning of a wide variety of online resources is used to retrieve relevant scientific publications, guidelines, and commentaries. After curation by domain experts, links are deposited into Web-based databases. RESULTS: PHGKB currently consists of nine component databases. Users can search the entire knowledge base or search one or more component databases directly and choose options for customizing the display of their search results. CONCLUSION: PHGKB offers researchers, policy makers, practitioners, and the general public a way to find information they need to understand the complicated landscape of genomics and population health.Genet Med 18 12, 1312-1314.

Horizon scanning for translational genomic research beyond bench to bedside.

PURPOSE: The dizzying pace of genomic discoveries is leading to an increasing number of clinical applications. In this report, we provide a method for horizon scanning and 1 year data on translational research beyond bench to bedside to assess the validity, utility, implementation, and outcomes of such applications. METHODS: We compiled cross-sectional results of ongoing horizon scanning of translational genomic research, conducted between 16 May 2012 and 15 May 2013, based on a weekly, systematic query of PubMed. A set of 505 beyond bench to bedside articles were collected and classified, including 312 original research articles; 123 systematic and other reviews; 38 clinical guidelines, policies, and recommendations; and 32 articles describing tools, decision support, and educational materials. RESULTS: Most articles (62%) addressed a specific genomic test or other health application; almost half of these (n = 180) were related to cancer. We estimate that these publications account for 0.5% of reported human genomics and genetics research during the same time. CONCLUSION: These data provide baseline information to track the evolving knowledge base and gaps in genomic medicine. Continuous horizon scanning of the translational genomics literature is crucial for an evidence-based translation of genomics discoveries into improved health care and disease prevention.

Genetic epidemiology with a capital E, ten years after.

More than a decade after Duncan Thomas gave his presidential address at the International Society for Genetic Epidemiology entitled "Genetic Epidemiology with a Capital E," genetic epidemiology has gone mainstream. Epidemiology has taken its place not only in gene discovery studies but also in characterizing genetic effects and gene-environment interactions in populations. Furthermore, epidemiologic principles are being applied to the evaluation of genetic tests. We used an online informatics tool, the HuGE Navigator, to describe the growth in the field in the past decade. We developed the HuGE Navigator as a means to continuously monitor the evolving information obtained from epidemiologic studies of the human genome. Between 2001 and 2010, the HuGE Navigator included 57,005 articles published in 2,396 journals. During that period, the annual number of publications increased almost four-fold. The articles included 986 genome-wide association studies and 1,879 meta-analyses of gene-disease associations. The total number of authors of published studies grew from 12,907 in 2001 to 48,389 in 2010. The number of diseases also increased over time, from 697 medical subject headings in 2001 to 1,404 in 2010. Gene-environment interaction was mentioned explicitly in 17% of published abstracts, almost half of which focused on gene-drug interactions. Clearly, genetic epidemiology has gone "capital E" in the past decade; however, the ever-expanding volume and variety of genomic information poses a formidable challenge for developing appropriate methods for analysis, synthesis, and inference on complex genetic and environmental effects. We extend Duncan Thomas' capital E to include "Evaluation" as the tools of epidemiology are increasingly used to assess how genome-based information can be applied in medicine and public health.

Public knowledge regarding the role of genetic susceptibility to environmentally induced health conditions.

OBJECTIVE: Diseases thought to be caused by exposure to environmental factors are also influenced by genetic susceptibility. It is not clear to what extent the public recognizes the role of genetics in causing these diseases. METHODS: We asked 2,353 people in a national survey to indicate their level of agreement with statements about the genetic contribution to four health conditions typically considered to be environmentally induced. RESULTS: 206 (9%) respondents believed that genetic susceptibility contributes to all four health conditions, while 751 (32%) believed that genetics plays no role in causing any of the conditions. Respondents were more likely to believe that genetics contributes to adverse drug reactions and smoking-related illnesses than to infectious diseases and diseases resulting from exposure to environmental agents. CONCLUSIONS: This study suggests that the public views genetic susceptibility as playing only a limited role in human disease induced by environmental factors. Increasing awareness of the role of genetic factors in these diseases will be necessary for translating gene discovery into effective personal and public health actions.

Reporting, appraising, and integrating data on genotype prevalence and gene-disease associations.

The recent completion of the first draft of the human genome sequence and advances in technologies for genomic analysis are generating tremendous opportunities for epidemiologic studies to evaluate the role of genetic variants in human disease. Many methodological issues apply to the investigation of variation in the frequency of allelic variants of human genes, of the possibility that these influence disease risk, and of assessment of the magnitude of the associated risk. Based on a Human Genome Epidemiology workshop, a checklist for reporting and appraising studies of genotype prevalence and studies of gene-disease associations was developed. This focuses on selection of study subjects, analytic validity of genotyping, population stratification, and statistical issues. Use of the checklist should facilitate the integration of evidence from these studies. The relation between the checklist and grading schemes that have been proposed for the evaluation of observational studies is discussed. Although the limitations of grading schemes are recognized, a robust approach is proposed. Other issues in the synthesis of evidence that are particularly relevant to studies of genotype prevalence and gene-disease association are discussed, notably identification of studies, publication bias, criteria for causal inference, and the appropriateness of quantitative synthesis.