Asymptotic properties of covariate-adaptive randomization

ABSTRACTPocock and Simon’s minimization method is a very popular covariate-adaptive randomization procedure intended to balance the allocations of two treatments across a set of covariates without compromising randomness. Additional covariate-adaptive schemes have been proposed in the literature, such as Atkinson’s -optimum Biased Coin Design and the Covariate-Adaptive Biased Coin Design (CA-BCD), and their properties were analyzed and compared in terms of imbalance and predictability. The aim of this paper is to push forward these comparisons by also taking into account other randomization methods, such as the Permuted Block Design, the Big Stick Design, a generalization of the CA-BCD that can be implemented when the covariate distribution is unknown, and the Covariate-Adaptive Dominant Biased Coin Design, which is a new class of stratified randomization methods that forces the balance increasingly as the joint imbalance grows and improves the degree of randomness as the size of every stratum increases. The performance of covariate-adaptive procedures is strictly related to the considered factors and the number of patients in the trial as well, which makes it hard to find a dominant rule, namely a design that is more balanced and less predictable with respect to other schemes. In general, stratified randomization methods perform very well when the number of strata is small, showing also some dominance structure with respect to the other designs. Nevertheless, the evolution and the performance of stratified designs are strictly related to the random entries of the subjects. Thus, these rules become less efficient in the case of both (i) limited samples and (ii) large number of factors/levels.

Let C be a collection of particles, each of which is independently undergoing the same Markov chain, and let d be a metric on the state space. Then, using transition probabilities, for distinct p, q in C, any time t and real x, we can calculate F (t)pq (x) = Pr [d (p,q)<x at time t]. For each time t ≧0, the collection C is shown to be a probabilistic metric space under the triangle function \(\tau _{T_m } \). In this paper we study the structure and limiting behavior of PM spaces so constructed. We show that whenever the transition probabilities have non-degenerate limits then the limit of the family of PM spaces exists and is a PM space under the same triangle function. For an irreducible, aperiodic, positive recurrent Markov chain, the limiting PM space is equilateral. For an irreducible, positive recurrent Markov chain with period p, the limiting PM space has at most only [p/2]+2 distinct distance distribution functions. Finally, we exhibit a class of Markov chains in which all of the states are transient, so that P ij(t)→0 for all states i, j, but for which the {F ttpq } all have non-trivial limits and hence a non-trivial limiting PM space does exist.

Let D denote the generator of a continuous time positive recurrent Markov process with state space N (or R +). Sufficient conditions are given to imply the existence of β>0 such that if λ≠0 is a point of the spectrum of D considered as an operator on the L 2 space of the equilibrium distribution, then Re(λ)≦−β. A related result is given for discrete time Markov chains.

St. Louis, Missouri, sits astride the confluence of the Mississippi and Missouri Rivers, a position that made the city the “Gateway to the West.” Its most famous monument, the Gateway Arch, memorializes the millions of pioneers who traveled to the city seeking transportation to the frontier lands of the West, and who were outfitted with the tools and knowledge they needed on the journey to these new lands of opportunity. The Clinical & Translational Science Awards (CTSA) program has enabled Washington University (WU) and its regional partners to create a gateway of another sort, aimed at helping researchers toward new opportunities and discoveries in clinical and translational research. The Washington University Institute of Clinical and Translational Sciences (ICTS) was founded in September 2007 to serve as the intellectual and physical home for clinical and translational research and clinical research training at WU and its partner institutions. ICTS was founded on WU's 4 prior clinical research and training grants—GCRC, K12, K30, and T32—and on an existing university-wide strategic initiative in multidisciplinary collaborative research emphasizing genome sciences, biological imaging, and clinical investigation. The formation of the ICTS resulted in a new level of regional partnership with other academic, health care, and industrial and community partners. In its second year of existence, the ICTS is increasingly able to integrate the different strengths of its partners to reach its goals of reinventing and reinvigorating CTS research and training. In designing the ICTS, we used a conceptual model that identified important phases in the progress of a clinical or translational research project (Figure 1). For each phase, we identifed needed resources and potential barriers to completion, and have structured the ICTS in a way that attempts to provide needed resources and expertise for each phase of research from development and refinement of research ideas through translation of results to clinical practice and community implementation. Education, training, and career development are at the heart of our ICTS, and are linked to all research phases and programmatic activities. Our ICTS has 15 key Program Functions, each linked to 1 or more research phases, and each with specific goals, milestones, and evaluation criteria for various services, cores, and programs. We will highlight 2 new and important program functions below: our community-based research efforts and our programs in biomedical informatics. Phases in the evolution of clinical research projects. The 3 CCBR arms serve as a model for community participation to improve research relevant to public health. A central transformative element offour ICTS was the establishment of the Center for Community-Based Research (CCBR), which has initiated community engagement ef orts in 3 important groups: community organizations, community-based health care providers, and individuals from medically underserved populations. Each of the 3 arms is designed to solicit community input to increase the relevance of research to community needs, and to encourage partnerships between academic researchers and community groups. Our Community Advisory Board, established by the 3 arms, meets quarterly, to ensure relevance of the CCBR programs and the ICTS programs as a whole. The Community Organizations program has focused on cultivating partnerships with community organizations and conducting community needs assessments. This new model has spearheaded a series of conferences with our Regional Health Commission to address how our academic medical center can better serve the needs of our local community and improve the public health. In the fall of 2009, we will begin a series of shared conferences with community organizations highlighting successful community-based participatory research projects. These conferences will include joint presentations by academics and community partners and will strive to facilitate the formation of new partnerships. To further stimulate new community-based participatory research projects, we will soon initiate a research grants program, supported by a regional coalition of universities and health care institutions, to fund research projects that include a high degree of cooperation between community organizations and academic researchers. The program has also initiated a traveling poster presentation forum that translates data in a community relevant fashion. The Community Health Care Practitioners program seeks to expand community networks of practitioners to increase opportunities for clinical and translational (T2 and T3) research. One such effort is the Washington University Pediatric and Adolescent Ambulatory Research Consortium (WU PAARC). Currently comprised of approximately 60 community pediatricians and 5 pediatric nurse practitioners, the network provides a formal structure to develop research ef orts that identify best practices and translate research findings into practice. Community pediatricians participate in the consortium in different ways, from contributing ideas and receiving study updates to helping develop study protocols and recruit patients from their practices. Some studies are initiated by WU PAARC physicians, such as a recent study on a Telephone Asthma Program that resulted from physicians’ frustration with patients not using their asthma maintenance medications. For this randomized controlled trial, 362 families with asthmatic children were recruited from 95 community pediatric offices over 13 months. The 12-month telephone asthma coaching intervention helped parents better manage their child's asthma care and improved children's asthma control and parents’ quality-of-life scores. An upcoming study will compare 2 corticosteroid treatments for croup, a disease commonly seen in community practice but never before studied in this setting. In addition to WU PAARC-sponsored studies, community pediatricians have opportunities to participate in other studies developed by ICTS researchers including participating in study development and implementation. An example is a prevalence study for community-acquired methicillin-resistant Staph aureu s for which 1,300 children were recruited from 11 WU PAARC practices over an 8-month period. WU PAARC serves as a model for cooperative research between academic and community physicians, and these examples show how those involved have a real opportunity to improve the health of children and adolescents. As our ICTS matures, we plan to expand our practice-based research network activities to include new patient and provider populations. The CCBR has also started HealthStreet, a project linking populations that have been underrepresented in research to health information, referrals to social, medical, and psychiatric services, as well as opportunities to participate in research studies. HealthStreet utilizes Community Health Workers and a comprehensive outreach approach to establish trust between the lay and medical communities. The hub of CCBR activity is a new storefront facility in an underserved neighborhood near our medical center where community members can receive free health screenings, health referrals, and information about WU clinical studies. Community Health Workers also reach people in sites such as beauty shops, laundromats, bus stops, and community agencies to spread the word about services offered, and opportunities for research. The 3 CCBR arms serve as a model for community participation to improve research relevant to public health. CCBR investigators are actively training and engaging other ICTS investigators to collaborate on community-based projects with the goal of addressing “outcomes that matter.” Major institutional goals for improving CTS research are the creation of efficient communication and coordination across key function activities; the creation of an informatics infrastructure capable of integrating ICTS program components for data sharing; and the development of electronic resources of existing clinical and biospecimen data that may be used to develop new clinical studies, identify potential research participants, and execute novel molecular correlative studies. The ICTS Center for Biomedical Informatics (CBMI) is addressing these complex core issues through industry-standard hardware infrastructure and software tools to store, integrate, query, analyze, and visualize complex clinical and molecular data sets. Among the several activities of CBMI are the integration and implementation of the research management tool ClinPortal, development and management of biospecimen tracking tools, and creation of the Clinical Investigation Data Exploration Repository (CIDER), a research repository of medical records. ClinPortal is a research management tool that allows investigators and study coordinators to integrate the data collection and management needs of a clinical study with critical regulatory compliance requirements. Investigators create case report forms using data elements (or variables) that are defined using controlled vocabularies, and ClinPortal then automatically generates the backend database schema as well as the forms to enter the data and conduct queries. Use of ClinPortal offers a number of advantages to the institution and to the investigator. By default, data stored in ClinPortal is available for sharing while following all HIPAA regulations, although the investigator that owns the data may turn of sharing if so desired. As a result, once an investigator has established his or her study within the ClinPortal system, there is potential for sharing relevant data with and from other studies. Data access is controlled both by investigator permissions and by regulatory requirements. Future phases include automated access to clinical data from our hospital and outpatient systems. Assuming appropriate release by the patient and access rights for the researcher, investigators using ClinPortal will eventually have access to clinical data from hospital and clinic sources for their study participants. Tissue Suite, a Web-based biospecimen inventory, tracking, and annotation tool, may be used for study-based biospecimen collection efforts and by other non-study investigators interested in identifying existing biospecimens throughout the medical campus that may be available for translational research. The ICTS Translational Pathology and Molecular Phenotyping (TPMP) Core supports the physical storage of biospecimens, while Tissue Suite is the application supported by both the CBMI and the TPMP to provide the informatics to manage data about those biospecimens. Investigators who are managing their own biospecimen collection efforts outside of the TPMP Core may still make use of the Tissue Suite application to manage their inventory in compliance with caBIG standards. Web accessibility, the ability to share biospecimens and biospecimen-related data, advanced search functionality, and capture of clinical, pathological, and quality assurance data for collected biospecimens are key features of Tissue Suite. Together, ClinPortal and Tissue Suite facilitate collaboration between investigators and data integration across studies. As clinical data becomes available through the ClinPortal tool, biospecimens referenced in the complementary Tissue Suite tool will gain increasing clinical annotation, making them even more valuable for translational research studies. Another major focus of the CBMI has been the development of a comprehensive research patient data warehouse. CIDER contains a wealth of data from the electronic medical records of most inpatients and outpatients seen at the WU medical center and affiliated hospitals in the BJC HealthCare system. This system will become available to ICTS investigators in 2009 and will initially contain data on patient encounters, laboratory testing, and medications. We sought community input in designing the patient notification and opt-out procedures and in strictly designating access and use of this information by qualified investigators. We expect that CIDER will become a very useful tool for many clinical and translational researchers to perform a review preparatory to research, identify potential participants to recruit to active clinical studies both by querying CIDER and as they are encountered in the health care system, and implement quality assurance and improvement initiatives. In addition to these highlighted functions, the CBMI is working on other ways to simplify communication and tracking within the ICTS. One new project is the creation of an electronic Research Concierge function to serve as the focal point to help clinical and translational researchers navigate through the study planning, design, review, and execution processes. This Research Concierge function will be added to WU's existing site that currently provides single sign-on access to the resources necessary to propose, manage, and close research projects and satisfy compliance requirements for research. In a nod to St. Louis history, these resources can be found at the WU Research Gateway Web site (http://research.wustl.edu). More information about the WU ICTS can be found at our main site http://icts.wustl.edu.

Practice: To avoid unintended harms, cancer screening efforts should be tailored to the characteristics of individuals, and efforts should be directed to informing those who should avoid screening as well as those who might benefit from screening. Policy: Cancer screening guideline developers should consider the potential for harmful unintended consequences when cancer screening is not targeted to the characteristics of individual patients. Research: Funding mechanisms are needed to continue to support independent health services research to optimize appropriate use of cancer screening in clinical practice. The Veterans Health Administration (VHA) has many unique strengths, such as its national electronic medical record and large clinically rich databases, which make it an ideal system for conducting health services research to improve patient care nationally. Health services research examines how people access health care, how much care costs, and what happens to patients as a result of this care under everyday conditions [1]. For several reasons, it has generally not been considered “translational behavioral research.” First, most translational research focuses on the development and implementation of interventions. In contrast, most health services research takes a broader view by studying the real-world use and outcomes of interventions to affect changes in current practice. Second, translational behavioral research has traditionally focused on increasing the use of interventions in clinical practice, such as increasing the use of cancer screening. In contrast, a core task of health services research has been to identify problematic clinical practices in which interventions are not appropriately targeted to the patients for whom they are intended. These approaches can be at odds when the push to increase the use of an intervention causes a high uptake among patients in whom the harms likely outweigh the benefits. For example, efforts to increase cancer screening may increase screening among patients with advanced age and severe comorbid illness [2]. Third, research focused on discovering the harms of interventions in every day practice is generally more difficult for health system leaders to accept than research to broadly implement new interventions. However, VHA has been an unwavering leader in health services research that evaluates the appropriate use of interventions and informs patients and clinicians about the real-world benefits and harms of interventions.

<b>MRC-NIHR Efficacy and Mechanism Evaluation (EME) Programme Logic Model</b><br /> <br /> Juliana Callaghan&sup1;, Robert Gray&sup1;, Louise Jones&sup2;, Adam Lockwood&sup1;, Danny McAuley&sup3;, John Norrie⁴, Danielle Preedy&sup1;, Sarah Thomas&sup1;, Insa Wemheuer&sup1;<br /> &sup1;National Institute for Health and Care Research (NIHR), School of Healthcare Enterprise and Innovation, University of Southampton, &sup2;UK Research and Innovation (UKRI) Medical Research Council (MRC), &sup3;Queen's University of Belfast, ⁴The University of Edinburgh<br /> February 2023<br /> <br /> The EME Programme is a partnership between the Medical Research Council (MRC) and the National Institute for Health and Care Research (NIHR). The aim of the programme is to improve the health and wealth of the nation by funding clinical efficacy studies that test if an intervention or technology which has shown early promise in patients works as expected in a well-defined group of patients. This may include using such studies to understand treatment mechanisms. The programme plays a key role in translational research by testing novel or repurposed interventions/technologies and determining whether their development should progress to later-phase clinical trials, or inform new earlier-stage research. <a href="https://www.nihr.ac.uk/explore-nihr/funding-programmes/efficacy-and-mechanism-evaluation.htm">More information can be found on the programme page on the NIHR website</a>.<br /> <br /> A logic model is a graphical way to show how an activity, programme or intervention is expected to work and bring about the benefits and changes it intends to achieve. By summarising the core elements, a logic model can then be used to support programme planning, implementation and evaluation. NIHR logic models represent in a linear flow diagram the key activities, outputs, outcomes and impacts of each funding programme as a series of logical steps.<br /> <br /> <br /> <br /> <b>Inputs</b><br /> The first step outlined in the logic model focuses on ‘inputs’, i.e., the resources needed to undertake programme activities. The inputs for the EME Programme are: <ul> <li>MRC and NIHR funding</li> <li>funding from the devolved nations</li> <li>NIHR coordinating centre resources</li> <li>stakeholder time</li> <li>the existing UK research infrastructure in terms of expertise, capacity and funding</li> </ul> <br /> <br /> <b>Activities&nbsp; </b><br /> Inputs feed into activities, the second stage of the logic model. Activities are the actions that NIHR and the funded research community undertake to help achieve the programme’s aims and objectives. Together, inputs and activities represent NIHR’s planned work.<br /> <br /> The initial focus for the programme is identification of key research questions, issues and programme priorities. A NIHR team develops targeted <a href="https://www.nihr.ac.uk/researchers/funding-opportunities/">funding opportunities</a> and activities through horizon scanning and collaboration with key stakeholders to stimulate activity or applications in an area of strategic priority. The programme also invites researcher-led proposals, which enables research into investigator-driven areas. All submitted project proposals are reviewed by harnessing detailed feedback from experts, including academics and other professional experts (for example, clinical staff and health care professionals) as well as patients, carers, service users, specific communities and/or members of the general public. A panel recommends high-quality research projects for funding according to set criteria.<br /> <br /> Funded projects are then actively monitored by a NIHR team that supports the project by providing advice and expertise regarding, for example, risks, appropriate methodologies, stakeholder management or patient and public involvement and engagement.<br /> <br /> NIHR support also includes guidance on the dissemination of research outputs. NIHR supports transparent research management and publication of knowledge, with information on and results of projects being openly accessible via the NIHR <a href="https://fundingawards.nihr.ac.uk/">website</a> and <a href="https://www.journalslibrary.nihr.ac.uk/#/">NIHR Journals Library</a> that hosts the <a href="https://www.journalslibrary.nihr.ac.uk/EME/#/">NIHR EME journal</a>.<br /> <br /> Further, the EME Programme’s application and funding process ensures that funded research projects test interventions/technologies independently and rigorously for efficacy, as well as mechanistic insight. The research process often includes collecting samples, for example in the form of biobanks. By funding projects that follow on from clinical studies supported by the NIHR or other funders, the programme also actively supports researchers in reusing data or samples.<br /> <br /> The EME Programme also builds research capacity through initiatives to broaden, strengthen and upskill the translational research community. &nbsp;For example, the programme provides tailored funding opportunities and activities to support the development of early-career clinical trialists and to stimulate cross-sector and international collaboration.<br /> <br /> <br /> <b>Outputs </b><br /> The next step of the flow diagram focuses on the range of ‘outputs’ that result directly from the undertaken activities. For the EME Programme, these include: <ul> <li>academic outputs such as articles in the <a href="https://www.journalslibrary.nihr.ac.uk/eme/#/">NIHR EME journal</a> and other peer-reviewed journals,</li> <li>communications tailored to key audiences (e.g. patient groups, industry, policy-makers and health-care professionals)</li> <li>study data from the testing of interventions/technologies</li> <li>samples collected as part of the research</li> <li>new intellectual property (IP) being registered as a result of the funded research</li> </ul> <br /> <br /> <br /> <b>Cross-cutting activities</b><br /> Some activities that enable the intended change cut across several steps of the logic model: <ul> <li>Stakeholder collaboration: the programme works closely with stakeholders such as other funders, industrial partners, patients, carers, service users, specific communities and the public to support translational research across the ecosystem</li> <li>Targeted knowledge exchange and dissemination takes place across the project lifecycle</li> <li>Activities, outputs and outcomes produced through research contribute to an increasing pool of knowledge which feeds into and influences both the identification of new questions and methods for answering those questions.</li> </ul> <br /> <br /> <b>Outcomes</b><br /> Outcomes are the changes or benefits the programme intends to produce through its activities and by funding a portfolio of research. Due to the programme’s position on the translational pathway and its role being to inform further research, outcomes are typically seen on a short- to medium-term.<br /> <br /> <b>Scientific advancements</b><br /> Research outputs document and disseminate the scientific advancements made by NIHR-funded projects. Scientific advancements include improved knowledge of the potential health benefits of interventions/ technologies and their mechanisms of action. They also include methodological advancements in the field of clinical trials and mechanistic studies. Improved knowledge can also stimulate further research. For research funded by the EME Programme, this can occur in both directions of the research and development pathway, with, for example, research teams continuing to study the tested interventions/ technologies in later-phase clinical trials or the results stimulating further earlier stage research. This all contributes to the UK’s world-leading reputation in delivering translational health research.<br /> <br /> <b>More efficient and effective health research</b><br /> Developing sample banks and establishing the effectiveness of interventions and technologies results in more efficient and effective health research. For example, trials showing ‘no effect’ of the tested intervention/ technology leads to researchers switching to other approaches, which reduces waste in UK scientific research and increases the efficiency of investments in Phase II and Phase III trials. Involving patients, carers, service users, specific communities and/or members of the general public in all steps of the research process results in increased relevance of research to patient and public need. Increased international and cross-sector collaboration in research enables aligned and complementary research aims to be addressed and optimises delivery of the translational research ecosystem.<br /> <br /> <b>De-risking future investment in research</b><br /> In the medium-term, research funded by the EME programme is expected to result in the de-risking of future investment in research by establishing which interventions and technologies are effective, how they work, and progressing those that are effective for practical use as measured on the Technology Readiness Level (TRL) scale. Increased knowledge around interventions/technologies and their mechanisms of action can also enable increased investment in associated development.<br /> <br /> <b>Enhanced research capacity </b><br /> Research projects funded by the EME Programme also contribute to enhancing research capacity in the form of training and skill-building of project team members through specialist skills and career progression, for example. Working on EME-funded projects is also expected to improve understanding of and interest in translational research for both researchers and trial sites. Collectively, this is expected to result in increased trial activity, capacity and networks, which in turn contributes to international recognition of the UK’s research capabilities.<br /> <br /> <br /> <b>Impacts</b><br /> Impacts, or long-term outcomes, are the anticipated broader (direct and indirect) changes or benefits for organisations, communities, systems and wider society expected to result from the programme's contribution via its activities and portfolio of funded research. These are expected to become apparent in approximately 10-25 years.&nbsp;<br /> For the EME Programme, the overarching long-term benefits are further research and the development of new and repurposed interventions and technologies. These contribute to health benefits through new or repurposed interventions and technologies; improved wealth of the nation through job creation; more effective and efficient use of healthcare resources through targeted investment into promising interventions and technologies; and ultimately, to improving the health and wealth of the nation.&nbsp;<br /> <br /> <br /> <b>Contributions and acknowledgements</b><br /> The NIHR supports the principles of open research, including full and appropriate recognition of the many varied contributions to the creation of knowledge. To support this, we use the<a href="https://credit.niso.org/"> CRediT taxonomy</a> to accurately reflect how each team member has brought their knowledge and skills to the development and delivery of this work. Those that have contributed to this work are listed alphabetically.&nbsp; <ul> <li>Juliana Callaghan:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Writing – review &amp; editing</li> <li>Robert Gray:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Writing – review &amp; editing&nbsp;</li> <li>Louise Jones:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization</li> <li>Adam Lockwood:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Project administration, Funding acquisition, Methodology, Supervision, Writing – review &amp; editing</li> <li>Danny McAuley:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Writing – review &amp; editing</li> <li>John Norrie:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Writing – review &amp; editing</li> <li>Danielle Preedy: &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Writing – review &amp; editing</li> <li>Sarah Thomas:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Conceptualization, Funding acquisition, Methodology, Supervision</li> <li>Insa Wemheuer:&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; Project administration, Visualization, Writing – original draft, Writing – review &amp; editing</li> </ul> <br /> In addition, we would specifically like to thank Rebecca Adler, Principal Consultant at NIRAS, and David Salisbury, Independent Consultant and NIRAS Associate, for facilitating the initial development of a Theory of Change for the EME Programme that formed the basis of developing the logic model presented in this document. We would also like to thank the Communications Team at the School of Healthcare Enterprise and Innovation, University of Southampton, for their advice and support in undertaking the visualisation of the model.&nbsp;<br /> <br /> The Logic Model presented in this document also builds on the following document:&nbsp;<br /> <br /> Rentel M C, Simpson K, Dav&eacute; A, Carter S, Blake M, Franke J, <i>et al.</i> A 10-year impact assessment of the Efficacy and Mechanism Evaluation (EME) programme: an independent mixed-method evaluation study. <i>Efficacy Mech Eval</i> 2021;8(20)<br /> <br /> <b>Competing interests&nbsp;</b><br /> This work has been undertaken as part of the delivery of the National Institute for Health and Care Research (NIHR), which is funded by the Department of Health and Social Care. The EME programme is a partnership between the Medical Research Council (MRC) and the NIHR. It is funded by the MRC and the NIHR, with contributions from the CSO in Scotland, Health and Care Research Wales and the HSC R&amp;D Division, Public Health Agency in Northern Ireland. All authors of this document have contributed to it as part of work paid for by the NIHR. No competing interests were disclosed.

Implementation and experience of a web-based allocation system with Pocock and Simon's minimization methods

Let Φ = {Φ n} be an aperiodic, positive recurrent Markov chain on a general state space, π its invariant probability measure and f ≧ 1. We consider the rate of (uniform) convergence of Ex[g(Φ n)] to the stationary limit π (g) for |g| ≦ f: specifically, we find conditions under which as n →∞, for suitable subgeometric rate functions r. We give sufficient conditions for this convergence to hold in terms of(i) the existence of suitably regular sets, i.e. sets on which (f, r)-modulated hitting time moments are bounded, and(ii) the existence of (f, r)-modulated drift conditions (Foster–Lyapunov conditions).The results are illustrated for random walks and for more general state space models.

The authors prove an almost sure central limit theorem for partial sums based on an irreducible and positive recurrent Markov chain using logarithmic means, which realizes the extension of the almost sure central limit theorem for partial sums from an i.i.d. sequence of random variables to a Markov chain.

This paper considers positive recurrent Markov chains where the probability of remaining in the current state is arbitrarily close to 1. Specifically, conditions are given which ensure the non-existence of central limit theorems for ergodic averages of functionals of the chain. The results are motivated by applications for Metropolis–Hastings algorithms which are constructed in terms of a rejection probability (where a rejection involves remaining at the current state). Two examples for commonly used algorithms are given, for the independence sampler and the Metropolis-adjusted Langevin algorithm. The examples are rather specialized, although, in both cases, the problems which arise are typical of problems commonly occurring for the particular algorithm being used.

We use matrices over formal power series to represent irreducible and positive recurrent Markov chains and identify a natural class of good finitary isomorphisms (magic word isomorphisms) as those arising from elementary matrix operations. This extends the positive K-theory framework which has been used for other classification problems in symbolic dynamics.

Three-Dimensional Cross-Sectional Light-Sheet Microscopy Imaging of the Inflamed Mouse Gut

In 2006, the National Institutes of Health (NIH), guided by the leadership of Elias Zerhouni, launched the Clinical and Translational Science Award (CTSA) program to jump-start transformative change in the nation’s research agenda to emphasize translating discovery from the bench to implementation in the community.1-5 Much of the funding for the CTSAs was garnered by rolling NIH funding of General Clinical Research Centers and educational K and T awards. Successful applicants were then asked to utilize these funds to (as stated in the Request for Applications) “transform the local, regional and national environments for clinical and translational science and to increase the safety, efficiency and speed of clinical and translational research.”4 Each of the 60 CTSA institutions has approached addressing this transformation in slightly different ways; however, it has become clear over the last 5 years that substantive changes, as a result of CTSA funding at academic institutions, are transforming medical research and the infrastructure that supports it. Each of the CTSA’s is required to address specific key functions to advance the consortium goals. These required key functions are (1) expertise in Biomedical Informatics, (2) a program in Research Education, Training, and Career Development, (3) a program in Community Engagement and Research, (4) an Evaluation component, (5) Pilot Projects in Translational and Clinical Studies, and (6) a Regulatory Knowledge and Support component. Optional CTSA key functions include (1) support of the Development of Novel Clinical and Translational Methodologies, (2) components or cores to support Research Design, Epidemiology, Biostatistics, and Clinical Research Ethics, (3) Clinical Research Resources and Facilities, and (4) components or cores that develop Translational Technologies and Resources. It is also encouraged for institutions to support research in Child Health. Addressing these key functions at the University of Michigan, for example, has led to transformational change in the structure and culture of the institution. After receiving a CTSA award in 2007, the University of Michigan (U-M) conducted a strategic assessment of the translational research enterprise through over 100 extensive interviews of faculty and staff providing the infrastructure for translational research, as well as the investigators conducting that research. The institution then conducted a survey of research team members involved in human subject research to identify and describe the current challenges facing translational research—specifically, the administrative processes associated with its planning and execution. In addition to identifying the issues, the team rated the technical and cultural/political levels of difficulty in implementing the solutions, as well as the corresponding impact each solution would potentially bring. These were then graphically mapped on a three-dimensional bubble chart (Figure 1). This stratification assisted in the planning, design, and prioritization of solutions, offering insights that guide the refinement efforts of a clinical research enterprise’s organization, business, and information technology (IT) processes. Bubble chart of implementation complexity. The chart orders tasks by level of execution difficulty—both technical and fiscal execution difficulty—as well as organizational and cultural complexity. The bubble size represents impact such that the bigger the bubble, the bigger the impact the change will have on improving translational research. For instance, enhancing study coordinator programs is relatively easy both technically and from an institutional buy-in perspective and was judged to provide a relatively large impact. Creating integrated informational technology systems involves significant investment, is more technically complex, requiring buy-in from multiple parties within the institution but is pivotal for achieving the goal of doing more effective translational research. The visual mapping provides context and definition to the challenge of enhancing the efficiency and effectiveness of translational research not only at the University of Michigan but all of the CTSAs. IT = Information Technology, CTSU = Clinical Trials Support Unit. Eight areas were identified as solution categories. The topics include organizational, procedural, financial, structural, educational, technological, and cultural elements. Transform and expedite study and contract approval process. The contract approval process, especially in regard to industry protocols, was recognized by the faculty as their number one area of discontent. The approval process was not clear to research team members and varied widely by unit and department as well as by type of sponsor. Educational and process improvement opportunities were identified. To address these issues, U-M has set a goal to expedite study approval process so that the contract and regulatory processes are executed within 60 days from submission for all studies and 45 days for industry-sponsored studies. The primary strategy to accomplish this aggressive goal was to create a team consisting of members of the units involved and to reengineer processes and conduct a demonstration project for feasibility within 1 year. Secondary strategies included increasing the transparency of contracting within the system to allow better tracking of proposals by the faculty and study teams, publishing desirable contract terms on a website for faculty and staff to use as protocols are developed, and to create an interactive, web-based research road map to guide faculty and staff through the necessary administrative steps to conduct translational research.6 Establish a clinical research support unit with best practice tools. The current translational research enterprise consists of multiple units with redundant functions of varying quality. The primary strategy was to create a leveraging of the resources of the CTSA to develop a franchisable model built on best practice modules that could then be easily adopted by other units throughout the university. Develop a customer friendly, transparent financial system that spans clinical trial budget creation to account reconciliation and study close out. The lack of transparent and integrated financial systems to track costs and other financial aspects of clinical research including budgeting and billing remains a challenge for study team members. The University of Michigan Office of Research has created the Calendar Review and Analysis Office (CRAO) to assist researchers with this process and is currently implementing a clinical study budget tool, eThority, to create standard billing calenders and processes across the organization. CRAO will also champion the evaluation of financial feasibility prior to proposal initiation; provide more assistance to PIs and research team members with study budgets, through increased training on best practices, budget negotiators, and standardized internal budget; and generate a new reconciliation report that will make study account statements more easily understood by research teams. Improve regulatory approval process and reduce Institutional Review Board (IRB) turn-around time. Significant gains have been made, resulting in a 40% improvement in IRB cycle time. Study teams will be advantaged by continuous improvements to the process. Embedded within the goal to expedite study approval is the goal to reduce IRB turnaround time (TAT) to 21 days for full convened board approval, including ancillary committees, and reduce no more than minimal risk studies to 5 days TAT. U-M is investigating strategies for greater IRB board participation of faculty, instituting another IRB board specifically for industry-based studies, as well as dedicating a board to minimal risk studies. The CTSA is also identifying various research teams to extract best practices. The CTSA has created an IND/IDE consultation and support team for researchers that have already helped over 100 investigators with IND/IDE applications. Create a home for study coordinators to promote education and sharing of best practices. A strong research team, anchored by a solid study coordinator, is critical to clinical research success. Study coordinator turnover was noted to be 18% a year. Of all study coordinators, 75% have been at U-M less than 3 years, compared to 44% of the rest of U-M staff and 56% of study coordinators nationally. Study coordinators noted that there was no coherent program or “home” for them within the system. Several strategies have been developed and are being enacted to address this problem. The CTSA created a virtual, web-based site for study coordinator support that will eventually include standardized resources such as workflow steps, templates, forms, and answers to common questions. A monthly study coordinator forum has been enacted to share challenges and best practices. Human resources has been engaged to develop uniform study coordinator job descriptions and to mentor the coordinators to develop a meaningful career path. Increase patient participation (enrollment) in research. The institution has set an ambitious goal to double patient participation in research in 5 years, while decreasing the number of studies that do not enroll a subject to 5%. A primary strategy to accomplish this was the creation of a web-based participant registry (UMclinicalstudies.org) to help match interested patients and volunteers with open studies. Currently, over 7,000 participants are registered. U-M is also investing in an electronic medical record system that better enables queries and “flags” potentially eligible patients. Further investigation by a study team is necessary to determine the underlying reasons for poor accrual to studies, whether due to missed competitive enrollment periods, rare disease studies that inherently will be challenged to accrue, registry studies with open enrollment, lack of ability to find enough patients within the patient population, or unwillingness to ask patients to participate. Understanding and addressing the underlying issues provides an opportunity for the CTSA to improve the quality, and potentially reduce “waste,” in the proposal and regulatory systems. Achieve efficiencies through integration of multiple, fragmented IT systems. More than 30 IT systems, as well as paper records, were identified that are currently utilized by investigators to perform clinical and translational research. Moreover, multiple IT service providers exist across the institution with poorly defined roles, responsibilities, and accountability. Alignment of key IT systems and personnel that support clinical and translational research is imperative. The CTSA is playing a central role in developing and leading a cohesive strategy that integrates IT for researchers’ benefits. For example, the institution is investing in a system-wide approach to develop an overall architecture that will create a clinical data repository. Evolve the culture of U-M to be more supportive and enthusiastic of clinical research. Like many academic institutions, the University of Michigan has struggled to value clinical research to the same extent as the basic sciences. In interviews and surveys, many faculty noted that choosing a career in clinical research can be seen as “falling between the cracks” or not as respected as a career in basic research. While improving infrastructure demonstrates value to these researchers, it is not enough. The CTSA has advocated to create a working group to make recommendations to University leadership to increase recognition for excellence in clinical research, to refine promotion and incentive criteria to value scholarly clinical activity and emphasize collaboration and innovation across all disciplines (team science). Furthermore, a plan is being developed to expand training programs to support training of those who want to become clinical and translational researchers, beyond the few slots that the CTSA (K and T programs) can currently support. The mission of the CTSA to enhance an investigator’s ability to conduct translational research at the University of Michigan has resulted in concrete organizational plans and investments to improve infrastructure of translational research and increase the satisfaction of participants, faculty, and staff who are engaged in these activities.7 Similar changes are occurring at CTSA institutions all over the country, validating the vision of Dr. Zerhouni and the investment in these vehicles of change by the NIH. Supported by UL 1RR024986 NIH/NCRR—Clinical and Translation Science Award.

In clinical trials, several covariate-adaptive designs have been proposed to balance treatment arms with respect to key covariates. Although some argue that conventional asymptotic tests are still appropriate when covariate-adaptive randomization is used, others think that re-randomization tests should be used. In this manuscript, we compare by simulation the performance of asymptotic and re-randomization tests under covariate-adaptive randomization. Our simulation study confirms results expected by the existing theory (e.g. asymptotic tests do not control type I error when the model is miss-specified). Furthermore, it shows that (i) re-randomization tests are as powerful as the asymptotic tests if the model is correct; (ii) re-randomization tests are more powerful when adjusting for covariates; (iii) minimization and permuted blocks provide similar results.

Background: A common issue in rare disease stratified clinical trials with multi-component endpoints is allocation bias, as they frequently lack blinding. Allocation bias arises when future treatment allocations can be predicted from prior ones, potentially leading to patients with specific characteristics being preferentially assigned to either the treatment or control group. Despite its potential impact, the effect of allocation bias on inference in these trials remain unstudied.Methods: To model biased patient responses, we derived an allocation biasing policy tailored to stratified trials with multi-component endpoints. Using this policy, we assessed the impact of allocation bias by evaluating type I error rates of a stratified version of the Wei-Lachin test, integrating Fleiss’s stratified test with the Wei-Lachin test, when allocation bias was present but ignored during inference.Results: Ignoring allocation bias when applying the stratified Wei-Lachin test results in an inflation of the type I error rate, exceeding the 5% significance level. The amount of inflation depends on the number of strata, number of endpoint components and the chosen randomization procedure. Less restrictive randomization procedures, such as the stratified Big Stick Design, exhibited the lowest type I error inflation, while stratified Permuted Block Randomization, results in highest inflation. The inflation of the type I error increases with the number of strata included. An increasing number of independent endpoint components is also associated with higher inflation.Conclusion: Allocation bias threatens the validity of stratified clinical trials with multi-component endpoints evaluated using the stratified Wei-Lachin test and should be mitigated through careful study planning. Ensure the number of patients in each stratum is not smaller than the number of strata, restrict the number of endpoint components to those essential for the study’s objectives, and use randomization procedures that allow for some imbalances, such as the Big Stick Design, to reduce allocation predictability.