VICTR CTSA Ethics and Biostatistics Group Meetings

2014 Aug 19

Churchill, Harrell, Byrne

Discussed two research ethics cases:
  • Deep-brain stimulation
  • Cluster-randomized ICU research

2013 Jul 15

Heitman, Gannon, Harrell, Clayton

  • Liz introduced Sam Gannon who is joining the group
  • Sam has been active in research integrity issues; applied for 2y grant to do randomized trial of educational materials for research administrators
  • Submitted to Office for Research Integrity under HHS
  • Will teach VPRED course in October
  • Discussed BERD Ethical Practice in Biostatistics working group; Liz presenting a talk at the Winter ASA Applied Statistics conference
  • Liz became chair of Ethics KFC 1 July
  • Ellen discussed the supplement to the Hastings Report on ethics in a learning health care system; author T. Beauchamp (author of Belmont report); opinions about sharing information and tissue; also see recent paper in Chest by J Ioannidis (2nd author)
  • Topic of future discussion: What is standard of care? (SUPPORT study)
  • Another future topic: Look at RCR plans (at relates specifically to their results) in K proposals

2013 May 21

Churchill, Harrell

  • Discussed ethics of medical decisions about individuals vs. the community; potential pitfalls of personalized medicine

2013 Mar 19

Special session to discuss Weijer et al: The Ottawa Statement on the Ethical Design and Conduct of Cluster Randomized Trials

PLOS Medicine 9:1, Nov. 2012

Li Wang, Kirk Lane, Hui Nian, Frank Harrell, Gordon Bernard, Chang Yu, Liz Heitman, Larry Churchill

  • Example: chlorhexadene bath in ICU, crossover randomization by 8 ICUs at VU
  • We expect a surge in cluster randomized process of care pragmatic clinical trials; relates to EHR
    • Minimal risk interventions with waiver of consent; pts identified easily from EHR; consent process slows down health care team
    • Must have minimal risk and research not being feasible without a waiver or alteration of consent
  • Identification of subjects: what about an EHR popup warning of risk of DVT/prophylactic dose of heparin; prompt randomized to docs, not patients
    • Could argue that exiisting standard of care was not being fulfilled; study question was effect of a reminder; popup also had warnings of downside
  • What is minimum risk? Touchstone is a standard physical exam/blood draw. Should it be relative to the environment at issue, e.g., ICU? Is it better to be in the study or not in the study?
    • Need to consider vulnerability; needs to be context- and trial-dependent
    • Regulations were designed for 1 subject 1 health caregiver
  • Can set up standard processes for ethical, financial, business review; these are just now being developed for performance improvement studies
  • Also need to consider health of the community
  • None of the article's recommendations were felt to be particularly controversial
  • It may be useful to design a mock study that would touch on several of the more problematic aspects of ethics/vulnerability/subject identification to work through some general issues

2013 Feb 19

Nwosu, Heitman, Agboto, Harrell, Churchill, Lane, Nian, Wang

  • Discussed Dan's point about invalid research being conducted in the hospital
    • Approach may reflect a "how do I fix this patient's problems now" focus
    • There are several features that may relate to ethics
    • What are the ethics of "quality improvement" projects not labeled as "research" (not to be published or generalized)
      • What is "publication"?
      • E.g. implementation of unfounded results; deflection of the field
    • Regression to the mean is not well understood
    • Would it work for us to give a course on rigorous quality improvement studies?
    • Need an article on the 5 top ways that QI studies waste resources and are detrimental to quality
  • Need to follow through with Tara regarding sketchy hypothesis generation studio requests

2012 Nov 20

Nian, Agboto, Harrell, Heitman, Nwosu, Lane

  • Discussed upcoming studio on ED chest pain treatment/diagnosis strategy
  • Poster for town hall almost ready; mentions ethics briefly
  • Ethics feedback for studio proposals has been especially expedient and helpful
  • Discussed any need for more ethics input to Meharry investigators
    • Vincent spoke about the bi-weekly Research Design Biostatistics & Ethics one-hour conference
    • May be beneficial to occasionally have combined meetings; we should get to know Lauren Solberg JD and Carolyn Sztela PhD
  • New Ethics in Biostatistics working group in BERD; trying to put together a case database plus resources
  • Meharry Biostatistics just received a US Army contract - post-deployment health care follow-up technology; uses smart phones etc.
    • Vincent will think about whether any VICTR resources can assist with the launch of this project
  • Current Meharry biostat workload is OK; Vincent will continue to evaluate
  • Successful town hall meeting at Meharry at which new Meharry incentive for partial investigator salary support was unveiled
  • Biostat requests have picked up in the past 2 months
  • Studio applications have been more complete as compared with last summer; Frank needs to respond to Tara Helmer when requests come that lack sufficient information
    • Need to revisit the situation and figure out how many examples would be sufficient
    • Only one example of hypothesis study request with no hypothesis should be sufficient
      • Also need to know where samples come from and who are participants
    • Another motivation - make sure investigator is fully prepared for the studio; also need to have the right experts present

2012 Oct 16

Yu, Wang, Nian

Li Wang talked about the poster to be presented at the Dec town hall meeting. One bullet regarding the ethics core needs input from the rest of the group.

Discussed the new arrangement for CTSA vouchers and its impact.

2012 Aug 21

Nwosu, Byrne, Churchill, Yu, Harrell, Jonathan Schildcrout, Heitman, Wang, Nian, Clayton, Lane

Jonathan talked about novel aim he's working on in new 5y grant cycleRisk modeling for diabetes
  • Using EMR definition of type II DM
  • Goal: proactive treatment to prevent precipitating events that predispose to DM
  • Cohort starts with diabetes-free patients in a 3-year span
  • Observation period starts and end of this 3y window
  • Fixed covariates: sex, race, SNPs, family hx, ht
  • Time-varying covariates: wt, bp, glucose
  • HbA1c may tend to be measured only when glucose is high and perhaps after being put on meds
  • Pts being put on meds mask natural hx of A1c
  • Is it worth developing corrections for this to be able to use A1c as a continuous response?
  • Will implementation involve addition of non-standard care? Is that an experiment? Is it ethical at face value?
  • Liz led a discussion on MSCI students' requests for studios (usually hypothesis generation or design studios)
  • Lack of specific hypothesis, not enough specificity of what will be tested
  • Difficult to choose the best experts constituting the studio
  • Studio execution may be inefficient
  • Junior researchers can often be helped the most but need to account for limits of available people resources
  • Need at least brief methods and study design plus nature of measurements, especially dependent variables
  • Also need enough focus with emphasis on testable hypotheses; perhaps emphasize falsifiability, i.e., list ways the study can fail
  • Frequently the issues are related to lack of working closely with mentor
  • Tara is coming to next month's meeting to discuss
Studio discussion: Studio from MSCI can benefit from more preparation. Often hypothesis is general and not testable. Design is often nonexistent. Discussion lead to why it is important for MSCI to get a studio early. This might be a different type of studio, one that helps them focus and tighten up study/design. A suggestion was made to include a line that asks for design methods in the application. Group came down to supporting the early MSCI participation and will try to get some input from MSCI faculty to help tighten up applications. This may be all that needs to happen to have a good studio for MSCI and a positive participation from the experts.

2012 Jul 17

Harrell, Clayton, Heitman

  • Crossover study, issue with long placebo run-in - withdrawal of long-term med
  • Studio applications need to be a bit more detailed
  • Would be good to have Lisa Robbins meet our group
  • Ethics Core - former NSF initiative - wants BERD participation - ethics materials to be deposited into the federal-initiated repository? Are their ownership issues or should we stay focused only on biomedical issues?
    • More intellectual output is being shared from academic institutions

2012 Jun 19

Harrell, Byrne, consultant Mary Banach, Heitman, Tinsley Webster, Lane

  • Ethics review project
    • Tinsley went over a proposed VICTR ethics template; intended for PI to use; will probably be used in place of retrospective trigger phrase search
    • Next step: proposal for VICTR funding from the research on research angle
    • Introduced randomized withdrawal study
  • Fallout from RCR lecture
  • Ethics issue related to translational research
    • Mt Sinai hospital readmission project
    • Bogus result likely due solely to regression to the mean
    • How to deal with the hesitation to randomize
    • Would be interesting to define a VU patient group where readmission is due to an attempt to restore health when the patient is "incurable"; an analysis palliative care at VU would be worthwhile
  • Briefly discussed benefits of version control
  • For future discussion: possible conflict of interest when an investigator serves as her own statistician

2012 May 15

Harrell, Nian, Wang, Lane, Heitman, Tinsley Webster

  • Discussed separating authorship from funding
  • Don't always expect to be co-author; analogy: someone running lab equipment = a statistician not using her brain who is just there to "drop the data into software"
  • Very important to estimate the full cost of a project up front, including manuscript work so that statistician gets involved in the intellectual component

2012 Mar 20

Fisher, Heitman, Tan Ding, Harrell, Clayton, Wang, Lane

  • Discussed reorganization of including relabeling to RCR
    • Material needs to be better known to biostatisticians
    • Would be good to find out about frequency of use
  • Need to link research ethicists with new working group on Ethics in Biostatistics
  • Need to cross-link and
  • Discussed oversupply of biomed sci PhDs and lessening of NIH grant funding and their impact on questionable research practices
  • Need to find out if a final report is needed for first 5y
  • Need to be on the alert for contributing text justifying the need to beef up ethics and biostat re: CTSA renewal; how to measure impact?

2012 Feb 21

Byrne, Clayton, Jill Fisher, Churchill, Lane

  • Discussed the ethical issues of randomizing patients to a control group vs. an intervention group in which the intervention group would have a probability of readmission and a decision support tool.

2012 Jan 17

Heitman, Clayton, Harrell, Jill Fisher

  • Discussed today's clinic on PREDICT
  • Discusssed medical care and research advertising
  • Discussed the contest between proactive consultation/advisory committees and VU's wish for rapid progress
  • Updated our wiki Ethics page

2011 Dec 20

Churchill, Yu, Li, Alvarez, Harrell, Heitman, Lane

  • EAB report; next time it will be beneficial to do less reporting and to leave much more time for discussion
  • What is truly a VICTR product for ethics/biostat?
    • New collaborations started after VICTR; enhanced collaborations; new funding opportunities
    • CTSA CC might do well to draft a definition: couldn't have begun without VICTR, were extended in scope, depth, or quality as a direct result of VICTR resources, collaborators joined because of VICTR resources/structure/procedures
  • Discussed complementary and alternative medicine research at VU, and advertising in general

2011 November 15

Harrell, Alvarez, Chang, Wang, Nian

  • The coordinating center may try and set up a big IRB that would enable expedited/non-committee review at institutional IRBs for those that are approved.
  • Discussed when papers should be retracted.
  • retractiontracker website
  • Think about ways to use non-utilized personnel time.

2011 October 18

Lane, Heitman, Wang, Nian, Alvarez, Yu, Churchill, Harrell

  • Discussed responsibility of PI's to verify results found in literature on which they would like to base a new design.
  • Systematic treatment of ethical pitfalls from biostatistical fallacies.
  • Is it ethical to use human tissue, government dollars, and other researchers' time to advance researchers' careers?
  • Is it an academic institution's responsibility to prevent poor research for ethical reasons.
  • Publicly funded studies should be held to higher standards, but there is an argument that society expects all research to contribute to public good.
  • Discussed an article in Significance calling for all research to require external validation.
  • Hope to schedule Dr. Heitman to cover the 10 items on here recent article at our seminar series.

2011 August 9

Harrell, Byrne, Alvarez

  • Frank discussed recent meeting with Gordon Bernard.
  • Debaun is starting a coordinating center for sickle cell anemia. Frank would like to use VICTR funds for start new centers. Debaun may apply for a large VICTR grant to fund a lot of percent effort of Li's and Dan's or Chang's time.
  • Allocate more of Frank, Chang, and Dan's time to a bridge to the collaboration plan or working toward major grant proposals.
  • Dan also wants to concentrate on the education and quality improvement.
  • Discussed the implications of research in quality improvement.
  • A criterion could be whether the study has the goal of publishing a paper or writing a grant, vs. improving the operation or the hospital.


Attending: Alvarez, Harrell, Scott, Heitman


Attending: Lane, Alvarez, Harrell, Nian, Wang, Churchill, Heitman

  • Renewal
  • Discussed a current stat-ethics case
    • fear of not getting the next grant
    • journals don't publish enough 'negative' studies (and we need to abolish the phrase 'negative study')
    • Consider a special manuscript studio
      • How to best present 'negative' results
      • Which kinds of exploratory analyses are kosher?
  • Opportunity for doing something in the 4th year SOM curriculum?


Attending: Heitman, Lane, Harrell, Tan Ding

  • Updated renewal proposal especially with regard to research ethics


Attending: Heitman, Lane, Harrell, Churchill, Alvarez

  • Case Studies in Ethics in Biostatistics poster to be presented at SCTS meeting next week
  • Discussed plagiarism and questionable research practices
  • Discussed Jill Pulley's feedback about renewal text draft, and graphics
    • One idea was to report on status of all vouchers (difficult to do at present)
    • Need more info on studios involving ethics consideration
    • Keep collaboration plan graphic?
  • Some of the ethics issues are problematic consent, vulnerable populations (including how to recruit), worries about risk benefit, investigator experience/skill, incomplete information (e.g., about compensation), stigmatization, confidentiality
  • Three possible levels of ethics review
  • Went over more input for renewal


Attending: Churchill, Heitman, Harrell, Alvarez, Scott, Byrne (by cell phone)

  • Discussed Ethics portion of (1) progress report and (2) renewal.
  • Discussed conundrum of expanding funding for ethics.


Attending: Churchill, Byrne, Nian, Wang, Harrell, Alvarez, Lane

  • Split some vouchers into two phases: Biostat design (include sample size justification) / Biostat analysis
  • Progress report due 28Feb11
    • Compile paragraphs of ethics and biostat "episodes"
  • Discussed ethics cases related to SRC, studios, and BioVU
    • Question about cluster randomized trials - a major issue re: IRB exemption is whether the risk/benefit of a patient is modified
      • worth an informal ethics/biostat clinic
    • Issue related to national neurosurgery database


Attending: Alvarez, Harrell, Nian, Wang, Scott, Heitman

  • Discussed changes in voucher system
  • Deadlines need to be clearer at StarBrite application time, and we also need to make it clear that we do not guarantee any certain completion time for data analysis
  • Statistical analysis plans need to be in place before analyses proceed
  • Need to always communicate to investigators that the collaboration plan is for long-term collaborations
  • Grant proposal assistance coming out of VICTR must list statisticians as "to be named"
  • Liz talked about the ethics key function committee of the national CTSA possibly having a face-to-face meeting in Nashville
    • This is a good opportunity for biostat to assist, e.g., with respect to authorship, responsible analysis, reproducible research
  • Recent ethics meeting in Chicago discussed ethics consultation
    • Such services have largely been unfruitful (apart from four institutions who did this before CTSA); a better model so far has been the proactive studio
  • New released today


Attending: Lane, Harrell, Scott, Yu, Heitman, Nian

  • Welcomed Hui Nian to the group
  • Discussed new look to and found chances need to ethics-related topic organization
  • Ethics article on how the research ethics component can benefit research - need to jumpstart
    • Liz will contact Tara to get feedback from studio satisfaction surveys; also need to get this on stats
    • Article about to be submitted by Frankel, Heitman, Harrell
    • Think about a visable seminar/workshop on questionable research practices
    • Looked at IMPACT under StarBrite
  • CTSA Coordinating Center
  • Frank talked about the Duke issue and a new Duke model for biostat support from CTSA


Attending: Heitman, Wang, Yu, Scott, Churchill, Lane

  • "The Importance of Organizational Justice in Ensuring Research Integrity" (Martinson, Crain, Vries, Anderson; J Empirical Research in Human Research Ethics, Sept 2010).
    • Wish to do a similar "climate assessment" survey here at Vanderbilt as part of CTSA project.
  • Also higlighted "Overbuilding Research Capacity" (Science Sept 10 Editorial; Bruce Alberts).
  • Liz addressed the questions posed by Dan & Frank in the 17Aug10 meeting - sent email to Dan & Frank.
  • Need to brainstorm a list of metrics to calculate the impact of ethics review.
    • Suggestions: prevention of disasters (PR and otherwise); questions that the PI failed to consider; poor communication of study details.
    • This is related to the Ethics Template project that Liz & group are working on.


Attending: Wang, Yu, Byrne, Harrell

Absence of ethics is not ethics of absence
  • How about a survey of VU investigators to assess their attitudes about research ethics?
    • Can we address the issue of some researchers seeing ethics review as more of a delay than a benefit? Or would this open a can of worms?
    • Should we test investigators' knowledge about responsible conduct of research?
    • Do we need an outside person to design and interpret the survey in order to avoid bias?
  • Let's challenge the ethicists to study their impact, with an eye towards the renewal grant
  • We need a special meeting with ethicists to discuss ethics of cluster randomized trials that are claimed to not require informed consent
  • All biostatisticians need to keep tracking VICTR time through August
  • MPH program is starting to use studios for their grad students
  • Discussed appropriateness of voucher requests that are for a specific statistical method or software use
  • MSCI students are assigned to come to clinics
  • Need to be clear that voucher and pilot study work done after July 1 gets charged to VICTR through BCC, no matter how long ago the project started
  • Protocols needing biostat help must have this in their initial budget; pay attention to this during pre-review


Attending: Lane, Agboto, Harrell, Heitman, Yu, Blakemore

  • Discussed Duke scandal
  • Discussed group of national biostatisticians discussing reproducible research, and its emphasis in BERD
  • NIH doesn't have a similar oversight process as FDA
  • Liz reported on the 2nd World Conference on Research Integrity in Singapore
    • Codes of conduct, standards of practice, nomenclature: research integrity vs. ethical conduct of research
    • Pressure to publish in other countries
  • RFA for CTSA coordinating center
    • Liz will talk to the rest of the Ethics group about related needs


  • Cover BERD characteristics survey


Attending: Blakemore, Scott, Harrell, Yu, Wang, Churchill, Heitmann

  • Face-to-face BERD meeting notes
  • Debriefing of IAB meeting
  • Should we be spending more of our time training junior researchers in research methodology?
  • Proposals to Gordon Bernard
    1. Deal with large number of rejections on first round pre-review biostat considerations. Also, researchers need direction to other core services or to their collaboration plan biostatistician(s), and they need help estimating the amount of biostat support to request. Require all submissions other than those just seeking supplies or CRC support to come to clinic before submitting the VICTR proposal. Pre-assign the ultimate pre-reviewer to be present at clinic at a day the investigator will come.
    2. Redirect database development requests to BCC; same for most special requests to work with particular personnel
    3. Better estimate the total amount of work and get pre-approval for matching funds from their division/department before any work starts on the $2000 portion
    4. Find out best way to announce new policies
    5. Do we need to change the $2000 twice/yr formula?
    6. Request outcomes of previous GCRC/!VICTR funding before an investigator can receive new funding from VICTR for last 2 funded projects. Also follow results after last patient is enrolled.
    7. Find more ways to keep mentors involved when we are assisting fellows


Attending: Yu, Scott, Blakemore, Harrell, Heitman, Churchill

  • Discussion of paper by Brice et al (Academic Med 84:S132, 2009): Publishing ethics in medical education journals
    • Impact of space limitations on salami slicing (also caused by bean counting of publications)
    • How many citations do fraudulent papers get?
  • SCTS meeting
    • David Ransohoff on biomarker research
    • Session on new approaches to therapeutic comparisons
  • Issue about studios: too little detail in some applications for studios (e.g., can't tell if there are ethical issues); for some, investigators are not doing enough background work
    • Frank briefly reviews every studio request and can keep this in mind
  • Discussion of bariatric surgery


Attending: Lane, Heitman, Harrell, Yu, Blakemore, Wang, Phillips

  • Heitman: Survey on research ethics consultation from national CTSA CRE KFC on research
    • 3 surveys from CTSA ethics consortium; doesn't fit very well with our framework
  • Lane & Heitman: Need for assessment of novice clinical researchers' knowledge of basic issues in human research ethics, appropriate ethics/study design education. Where to place in VICTR system? Content? University of Michigan informed consent module? (
    • Some investigators are attempting to jump over various steps to move from basic research into clinical research
    • Mike Stein along with several VICTR personnel is designing a boot camp for clinical research which would give a certificate; it will probably be video-based (needs to better distinguish RCR from IRB)
    • Not enough FTEs funded for ethics experts to review every protocol coming to SRC
    • Should there be ongoing audits of VICTR-funded studies?
    • Should less research be done, putting more resources into good research? Can one predict which research is going to turn out to be "good"?


Attending: Scott, Phillips, Wang, Blakemore, Heitman

  • Leslie Boone, Translational Research Coordinator attended.
  • Discussed ethical issues related to T2 proposals
    • Coercion
    • Follow-ups - patients are consented then not told the results and how their data is being used.
  • Biostatistical involvement with community research proposals
    • Nothing formal, not much coming across our desk
  • Ethics added to the list of check boxes for studio applications. However, it's never checked.
    • Studio applications are so short that it is unclear as to what the ethical issues are.
  • Decided to try a few applications for ethics review and see what happens.


Attending: Blakemore, Scott, Yu, Wang, Churchill, Phillips

  • More questions being asked about ethics studios
    • Ethicists have been more consistently in the loop when studios have been requested.
  • New guidelines for Responsible Conduct of Research for K awards.
    • Liz will be giving a related seminar at the Friday morning CRC Research Skills workshop.
  • JeffreyHorner has been maintaining
    • Frank is looking for input from other ethicists as to what they would like to see on the website and information as to how it is currently being used.
  • Vouchers are to be used for one specific project at a time.
  • Discussed when it is important to have the advisor in attendance in meetings.
    • Important for clear and consistent communication.
    • Difficult when investigator tries to explain what they learned from the statistician/ethicist to their advisor.
  • Teaching Research Ethics four day workshop held at the University of Indiana
  • ACRT (Association for Clinical Research Training) SCTS (Society for Clinical and Translational Science) meeting in DC in April, Ethicists may want to look at the program and consider going.
    • Poster categories include ethics.


Attending: Blakemore, Lane, Harrell, Yu, Scott

  • Discussed ethics training
  • An ethics mini-course as one of the Biostatistics seminars would be welcome
    • could cover general principles of ethics
    • use as a lead-in to ethical conduct of research
  • We need to get more rigorous in writing analytic plans before unveiling the data
"The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data." ~ John Tukey
  • Cf ASA code of ethics
  • Consider ethics training in our grad program
  • Discussed several issues around investigators demanding that continuous variables such as BMI be dichotomized


Attending: Lane, Harrell, Scott, Blume, Wang, Phillips, Heitman

  • Jeffrey updated us on the interim monitoring project
    • Group met to discuss standard analyses/reporting
    • Frustrated about getting data from investigators; there may be issues for future discussion related to scientific quality control
    • Waiting to enter data into a database may cause quality/completeness problems
    • Ethics may be impacted if safety data are not being entered in a timely fashion (this would not pertain to serious adverse events which are immediately reported); trends should be monitored
  • Motion to add to the VICTR StarBrite application a small section on research data management plan; passed unan.
  • RedCAP clinic 2nd and 4th Tuesday 2-4 in Biostat small conference room; perhaps publicize more
  • Potential BERD project to create and fill a RedCap database with characteristics of BERD members
  • Potential BERD Online Resources project with FDA CDER
  • Liz reported that FDA may have funds for ethics education; focus on training of investigators to do good studies
    • Of interest would be an ethical conduct of research for statisticians in drug development
  • Discussed recent studio application by an MSCI student that has ethics issues
  • Terri discussed new certificate program being designed by Michael Stein and Gail Mayo that includes an ethics component
  • Liz discussed planning for biostat/ethics seminar on Crohn's disease (probably for January); may want to involve the investigator
  • Sharon mentioned a surgical study in which inclusion/exclusion criteria were not required by the IRB


Attending: Wang, Yu, Heitman, Churchill, Harrell, Scott, Blakemore, Byrne, Lane

  • Liz spoke about her discussion on the national CTSA ethics KFC Ethics teleconference
    • Case Conference in Ethics & Biostatistics - educational, logic of the combination, getting biostatisticians to recognize some of the ethical issues
    • Questions about content of sessions
  • General issues regarding ethics of scale construction, e.g., depression scales, use of surrogate endpoints
    • arbitrariness, applicability, number of items vs. quality of items, initial foundation
    • possibly do something for SCTS meeting April 2010?
  • How to enhance the impact of translational research?
    • How do you tell something is a bad idea before it has failed?
    • Need to tie to public health needs


Attending: Jeffrey Blume, Ben Saville, Kirk Lane, Shraddha Nigavekar, Terri Scott, Tan Ding, Frank Harrell, Larry Churchill, Dana Blakemore, Sharon Phillips, Dan Byrne, Leena Choi, Janey Wang, Lei Xu

Special Session on Interim Monitoring

  • New CTSA supplement mandating interim monitoring of certain VICTR-funded studies
  • Funds some non-VICTR statisticians
  • Focus on optimum use of funds, early stopping for futility, adequacy of sample size calculations, occasional early stopping for efficacy
  • Reports will be reviewed by the SRC both individually and as portfolio of previously funded studies
  • Some of the criteria for selection of studies for interim monitoring: single-center study, randomized or observational; larger use of resources, long accrual phase; early outcome variable determination
    • FInd studies on the basis of design characteristics then look at resource use
    • Go back 6m-12m to see if studies underway can be selected. But focus is on new applications.
  • Three approaches: blinded sequential confidence intervals, likelihood, Bayesian; preface with accrual monitoring
  • Stat pre-reviewers will note study characteristics in a new RedCap database. We need to make a coded list of study designs.
    • Pre-reviewer's understanding of study design will be fed back to investigator for checking
  • Ethics issues
    • Is an IRB exemption needed?
    • What are ramifications of judging study yield on effect estimates and their uncertainty intervals vs. rejecting H0?
  • Statistical issues:
    • Interim analysis may not have been specified in the proposal, or our interim analysis methods may be at odds with pre-specified statistical analysis plan
    • Alpha spending
    • Adequacy of normal approximation to Bayesian posterior distribution (will speed up calculations if approximation is adequate)
    • Emphasis on estimation vs. signal detection
  • Reporting issues
    • Decide which approach(es) should be presented to the SRC
    • Partially automate reports once we learn more; find out how uniform the reports can be
    • Automatically produce portfolios with snapshots of multiple studies
    • Feedback to investigators that will not unblind final analysis
  • Initial implementation
    • While recruiting a nearly full-time MS biostatistician, each interim monitoring PhD statistician will assume responsibility for one study from start to finish
    • Start making study design dictionary
    • Need to make a list of benefits to investigators
      • Availability of future resources in the future
      • Educational opportunities for investigators to learn more about study design
      • Increasing awareness of advantages of sequential experimentation and better designing their next study
    • Need new text in approval letters
    • Need to finish enumerating new RedCap auxiliary database fields and find ways to pre-populate some of it from StarBrite and other sources


Attending: Vincent Agboto, Kirk Lane, Frank Harrell, Chang Yu, Li Wang, Dana Blakemore, Liz Heitman, Sharon Phillips, Dan Byrne

  • Discussed 2 Meharry submissions to VICTR and need for more detail in proposals
  • Kirk discussed the confusion that many investigators have - the belief that IRB review covers every aspect of ethics
    • Example in which protocol specified that subjects refusing to be enrolled in a study could be gotten around by going back to the IRB to ask for a single exemption to the need for consent
  • Should we raise awareness of the use of non-science-based therapies used and increasingly advertised at VU?
    • Consider offering a training session for the marketing department? This could be delivered by ethics + biostat.
      • Need to cover the issue of marketing clinical trials as benefiting patients as long as a specific trial is not mentioned
      • Also point out that many surveys require IRB approval
  • Liz reported on the national CTSA ethics key function committee, involved in drafting a core competencies document including biostat/study design
  • Liz discussed the climate survey to test the waters at 3 CTSA sites; covers research environment etc.; will use RedCap and will not be limited to CTSA sites
  • Frank reported on the approval of the interim monitoring VICTR supplemental grant application. This is for 2 years. We need to finalize the criteria for selection of VICTR-funded studies for mandated interim monitoring. Consider making a presentation at an upcoming SRC meeting. There are opportunities for research on research.
  • Discussed the possibility of unitizing the cost for ethics reviews
  • Checkbox for ethics was added to studio requests


Attending: Lane, Blakemore, Xu, Wang, Phillips, Yu, Harrell

  • Discussed Nature Neuroscience article
  • Discussed Frank's visit to new Cincinnati CTSA
    • Like us, no methodologists are on their IRB, but they give more input to IRB
  • Clarified coercion vs. incentives in clinical trials, especially in pediatrics
    • Paying for time and inconvenience, not for the deed; are not paying people to take risks


Attending: Churchill, Harrell, Xu, Wang, Phillips

  • Need to check that Ethics has a checkbox in studio requests
  • Need for statisticians to more often insist on a statistical analysis plans
  • Looked at paper in Nature Neuroscience 1May09


Attending: Phillips, Heitman, Harrell, Scott, Byrne

  • Discussed Crohn's disease proposals
    • Could lead to a paper linking ethics to stat
    • IRBs are not aware of some statistical issues. It would be good to address the question of whether waste is harm.
    • This may be suitable for a Ethics/Biostat case presentation in our seminar series
  • Brief update on CTSA supplement application


Attending: Heitman, Churchill, Donahue, Lane, Yu, Wang, Byrne, Harrell

  • Talked about integration of research ethics and biostatistics, while they are separate with regard to national CTSA consortium
  • Liz clarified that there is much of ethics that is not clinical per se, e.g., responsible conduct of research
  • Liz suggested all of us take a close look at the ASA code of ethics, which is not as strong as it could be
    • A joint session would be valuable
    • Kirk suggested that an end result could be a paper circulated to other CTSAs
  • What are points at which research ethics pre-reviews can take place?
    • CTSA research ethicists can come to random biostat clinics, then we can learn to better target participation
    • Studios haven't by default included research ethicists
      • Perhaps triggers for ethicist involvement should be described and given to the studio director
      • Studios officially start 2008-01-07
    • Kirk is involved in GCRC reviews
  • We discussed the strong interface of ethics and biostatistics in the area of study design and study futility/waste
    • Implementation of research ethics and its combination with biostatistics in the context of daily practice is worthy of a paper (a very practical one)
      • Include a section on the ethical implication of design choices
  • We briefly discussed the Palmer paper
    • Larry feels that since a trial intends to affect future collective good, the within-trial ethics may be thought of as side constraints, not as all encompassing.
    • Liz brought up the example of how do you know whether a drug is safe to use on children until you actually test it on children
    • Some informed consent documents say that the trial is not intended to benefit you, the subject
    • To the point that Palmer states it is unethical to make a sample size large enough to estimate an effect with precision (as opposed to being just large enough to detect a nonzero signal), using minimal sample sizes may be unethical to the larger population
      • DMCs have prolonged studies to result in data that can change practice
    • Ethics of use of placebo is worthy of future discussion

Action Items

  1. Read ASA ethics document and think of points to discuss/rework/add
  2. Think about impacting the IRB directly
  3. Consider working on the practical paper described above


Attending: Churchill, Donahue, Yu, Wang, Byrne, Harrell

  • Discussed placebo effects in clinical trials - when is it ethical to have a placebo group?
  • Risk perception - an issue for future discussion
  • Do patients have access to trial protocols?
  • Larry's vision for how we can work together * Determine ways in which it makes sense * Biostatisticians catalog issues that come up with investigators, keep in mind places where ethics makes sense to address * Research ethicists could be involved in biostatistics teaching, e.g. MSCI
  • Liz is designing a document on responsible conduct of research for CTSA
  • Are there issues in need of discussion related to financial incentives to investigators?
  • Will it work for IRB to not have biostatisticians but to rely on the scientific review committee?
    • Not all studies reviewed by IRB go to SRC
  • Review tracking for research ethicists - Larry said we should do this

Action Items

  1. Continue to read ASA ethics document; be ready to discuss
  2. Watch for teaching opportunities
  3. Watch for research ethics-related issues in our biostat collaborations; think of ethics questions that come up or should come up; how does improper interpretation of evidence fit in?
  4. Insure that a research ethicist is always on the SRC
  5. Larry look at what items need to be tracked in a review tracking web page; Li will think about location for information


Attending: Heitman, Churchill, Donahue, Lane, Wang, Byrne, Harrell

  • Discussion of ethics reviews
    • Kirk discussed an example in how conveyance of highly sensitive information to genotyped persons was to be done was not covered in the protocol; originated from VICTR SRC
    • We need a mechanism to trigger ethics reviews more routinely
    • Is there a need for rotating ethicist review of every protocol submitted to VICTR?
    • Suggested a retrospective review of all protocols submitted in Feb 08
      • Count the number of protocols that warranted a fuller ethics review
      • Catalog the reasons
      • For those warranting further review, review minutes or IRB to find out how many were already flagged for a potential ethics concern
    • Reviews could allow ethicists to map criteria for triggering reviews, once a large variety of protocols are seen
    • May be good to coordinate with Jan Zolkower, clinical research advocate
    • Liz will talk to Denise Roe to coordinate with the IRB and find out about confidentiality of minutes
    • Split Feb08 protocols by four ethicists; Kirk has already reviewed 2; will be 4-5 protocols each
    • Need to anonymize with regard to investigators; only need narrative
    • What do we do if we identify an ethical issue not identified by the IRB? Liz will talk to Denise about this.
    • Data would be great for upcoming progress report and for a paper
    • Kirk gave another example in which data were collected with no plan for analysis of that data component
    • In the future we might propose that protocols with CTSA ethicist concern don't go to SRC until the problems are resolved
    • Liz developed an initial outline of study components needing ethical review
      • confidentiality
      • consent, minimal harm
      • selection criteria (inclusion/exclusion) / discriminatory labeling
      • duel use potential
      • deception and disclosure
      • unclear goal of collection of random data / fishing / lack of analysis plan
    • Need to think about how/when studios can help
  • Tabled ASA ethics guidelines discussion for next meeting


Attending: Churchill, Lane, Yu, Byrne, Harrell

  • Debriefed on yesterday's special core meeting to assist Sports Medicine
    • Need to follow-up on assent documents for patients < 18 years old
    • Related to percentage criteria for site inclusion the most important consideration is watching for undue pressure on patients
    • Sometimes a large number of refusals may reflect a good consent form
    • Next time make the context more clear as well as the data presentation
    • We can usually differentiate IRB vs. non-IRB issues
  • Ethics review pilot study for Feb 2008 protocol submissions
    • Liz and Kirk met with Denise Roe of the IRB; IRB was happy for this pilot study to proceed and recognized there are ethical considerations outside the standard regulatory review mode; Ms Roe thought this would help them
    • There are 7 deidentified protocols to review; all have been distributed for review
    • It would be useful to extend this over a few more months
    • Can develop a list of concerns from the reviews, and also look out for known concerns from the literature
    • The bioethicists might educate the biostatisticians about points to look out for in our reviews of all the protocols; this would trigger ethics reviews
      • Cannot do this by looking at specific protocol sections
  • Keep looking out for opportunities for ethics reviews (studios and elsewhere) and ethics-biostat reviews of current studies or those in the planning stage
  • Kirk brought up the issue of subject notification of results of genetic testing in an observational study; the investigators were concerned because of a one-year followup of behaviors linked to the gene


Attending: Churchill, Heitman, Wang, Byrne, Harrell

  • Discussed an ethics issue from Kirk: 2 April 2008, Ongoing study involving children, weight and diabetes. Design was to have participants that were enrolled come to the CRC to have height and weight recorded. These visits were scheduled to coincide with clinic visits so as not to require duplicate trips to the VUMC area. The consent form allows the researchers to contact participants. The physician who sees these patients in the clinic is a co-investigator in the study. The issue is that some “small” number of patients did not come to the CRC to have their data recorded. None of the patients with missing data had officially withdrawn from the study. That is none of these individuals or their parents or guardians (consent provider) had informed the investigators orally or in writing that they wished to withdraw. Since these are pediatric patients it is usual practice to record height and weight at clinic visits. The co-investigator (clinic physician) stated he could “get the missing data from the medical record.” Was this appropriate? The short answer is “no”. However, since there is consent to re-contact the participants, no withdrawal notification has overtly been expressed, the missing data exists in a “private” file, the co-investigator has legitimate access to this information, there seems to be a reasonable way to proceed. The investigators should contact the participants (consent provider) and inform them that there is missing data. That the data exist in the medical record and that their physician is also an investigator in this study. Does the study have permission to obtain the height and weight information from the clinical data (medical record)? An affirmative response is required before proceeding.
    • Feeling of the group is to find out what the protocol says about re-contacting subjects. At any rate a call from the investigator will be confusing to the subject. An up-front solution would have been to mention in the informed consent that the investigator has access to the medical record. Recommended to call IRB and ask the question about the use of the medical record and about re-contacting subjects.
  • Liz is going to a meeting at Washington University sponsored by the Office of Research Integrity:; concerns responsible conduct of research; several interesting topics
  • Discussed how existing courses (e.g., clinical trials methodology) could have a segment on rcr added
  • Ethics reviews for Feburary 2008 have been completed; some wait to be recorded; still intend to add another 2 months' worth of reviews
    • Biostat reviews need to compile a list of trouble areas they think might trigger ethics reviews
  • Next time consider going through the WHO checklist for what informed consent and protocol documents should contain
  • Frank relayed the discussion from today's BERD online resources task force about the potential for having a discussion board for biostat (and possibly ethics?) questions from CTSA researchers nationwide


Attending: Heitman, Wang, Byrne, Harrell, Lane, Xu, Agboto

  • Discussed Liz's biostat seminar presentation
  • Discussed current problem of having an analysis file containing non-consented patients even if a flag can be used to remove them from analysis; non-consented patients were used in very indirect analysis decisions
    • Could bias sample from post-facto consent; subjects unhappy with their outcome may not give consent; need to monitor re-consent acceptance fraction
    • What is the basis for re-contacting patients to re-consent them?
    • Have a file with study IDs of patients who are not known to have consented; however for one site we don't have the list
    • Need to remove non-consented patients from the primary dataset and re-create the analysis file
    • Must exclude all subjects from the site having no list unless a list of definitely consented patients can be obtained; those patients may be included from that site
      • Note: The list of non-consented patients from the site in question may be available soon so this may not be an issue
    • IRB should be notified each time a patient is discovered not to have consented
    • How should this be written in a paper? Need to reflect that non-consented patients were used in preliminary analyses
  • Frank relayed an ethics story from a statistician at another university
  • Ethics review update: files are being saved; need to be distributed


Attending: Churchill, Wang, Byrne, Harrell, Lane, Yu, Heitman

  • Discussed an issue with a surgical study with only 5 controls
  • Update on multi-center orthopedics study
  • Liz summarized the CTSA Clinical Research Ethics Workgroup teleconference from yesterday; she is on the educational curricumum task force of CREW
    • A task force is trying to write a white paper on clinical research ethics consultation; a survey of the 38 CTSAs will be done
    • Goal is to document "best practices"; this is a good time for our input
  • Retrospective ethics review of CTSA protocols is at a standstill
  • Need for research ethicists to give biostatisticians a 'checklist' to alert them of potential ethical issues to note while doing biostat reviews
  • Might be good to bring a grad student in to screen protocols for ethics concerns; could ask for CTSA funding for this as a research project; goal is to produce a validated template that may be useful for biostatisticians; might be done as a voucher
  • Kirk reviews all proposals that go to the SRC


Attending: Lane, Heitman, Xu, Byrne, Harrell, Churchill, Saville, Wang, Phillips

  • Welcomed Sharon Phillips (new member, collaborates with Surgical Sciences) and Ben Saville (new faculty member and visitor to CTSA) to the meeting
  • Reviewed status of extra 2-month retrospective review for ethical questions, and creation of a checklist and template of issues to address
    • Format of voucher funds application does not fit this type of research; asking for $2k
    • Will deal with development and rewview phases
    • Need to deal with clearance of grad student for access to original reviews and protocols
  • Larry reported on multi-center orthopedics study; would be useful to have another roundtable discussion
  • Larry related an issue from Emergency Medicine - questions raised on pink sheets; might suggest clinic
  • A paper should emphasize what is in it for the investigators to have an early ethics consult/review
  • Kirk discussed an example where the coercive nature of monetary/goods payment to a child for participating may be a problem; issue is also related to the source of the goods
    • Compensation vs. inducement


Attending: Lane, Xu, Harrell, Saville, Wang, Phillips, Heitman

  • Griped about the economy
  • Talked about biostat voucher submissions, especially for surgical research
  • Discussed rules about direct participation in grant proposals
  • Problems have arisen for biostat voucher awards that are insufficient for the work required (e.g., for large complete databases)


Lane, Phillips, Xu, Churchill, Harrell, Wang, Yu, Heitman

  • Discussed upcoming External Advisory Committee meeting and new initiatives
  • Discussed application for graduate student funding for retrospective ethics reviews; Diane is still working on scheduling
    • New qualitative research core may help
  • Chang raised an issue of a pilot study for which the investigator wanted an interim analysis
  • We might suggest to CTSA that it more closely monitor subject accrual, in addition to doing sequential analysis of results
    • Perhaps operationalize with Bayesian monitoring
    • Rather than SRC approving only a small (futile) sample size, approve the maximum sample size and demand sequential monitoring
  • Liz reported on the Am Soc Bioethics & Humanities' Clinical Research Ethics Work Group
    • Consultation task force - research ethics consultation relationship with IRB (e.g., service advertised to IRB; legal discoverability; treat as QI or research with PI consented or oversight); distributing templates and best practices in clinical research ethics consultations
    • Need for a statement on financial COI; issue on institutional vs. consortium rules on COI
    • Need to clarify whether CTSA percent efforts cover any of our consortium work
    • New CTSA strategic plan issued by the consortium steering committee creates key function committees which do not seem to include ethics; ethics appears only once in the strategic plan whereas compliance appears often
  • Next meeting: spend most of the time on brainstorming template content; Liz went over some aspects of the WHO ethics committee document that married ethics & biostat


Attending: Byrne, Harrell, Wang, Phillips

  • Discussed progress related to the grant proposal's original aims
  • Regarding putting innovative designs into play, Dan suggested doing a retrospective review of a few completed analyses to assess whether the answer would have been known sooner
  • Ways of assessing impact to MSCI students are being considered
  • Discussed adding a bank of covariables to RedCap
  • Need to check that in proposals listing a statistician the investigator has actually worked with the statistician
  • We need to be vigilant in not approving industry-sponsored proposals that are not good science or have inappropriate analysis plans
  • Proposals using Cohen's effect size need to be improved


Lane, Xu, Harrell, Heitman, Byrne, Yu

  • Need to revive the meeting to discuss CTSA coverage of ethics review
  • Upcoming biostatistics ethics case conference will be on authorship; Kirk discussed another possible example
    • Liz emphasized the importance of determining authorship as early as possible
    • Good subjects for a writing course
    • Need to clarify power of senior and first authors
    • Another good area to cover is delve into how deep a co-author needs to be able to vouch for sections of a manuscript
    • Compare to department grant proposal policies for persons named as key personnel
    • Tried to find authorship criteria for biostat journals, finding only guidelines for Biometrics
  • Can adherence to ethical conduct of research be quantified?
    • A survey of perceived ethical conduct at VU could be done, including surveying the community's perception of the ethics of VU research
  • Kirk discussed a recent studio where patients having medical devices surgically implanted were being studied using data generated by the device
    • At issue is whether patients having computerized devices implanted should be informed at implantation time that they will likely be contacted to enroll in a research study in the future
      • Is it fair to exclude populations from using more expensive devices that can provide research data?
  • We need to find out how many ethics consultations there have been to date.
    • Liz will contact Jan Zolkower one more time to see if she could attend one of our meetings; Jan is reviewing protocols for the IRB with regard to human subjects protection.


Attending: Byrne, Wang, Lane, Scott, Clayton, Harrell, Heitman, Jan Zolkower, Churchill, Phillips, Xu

  • We welcomed Jan, Patient Advocate, and Terri Scott, Biostatistician III, 20% VICTR who is joining the team
  • Jan is a member of the Performance Improvement Team of the IRB; attends VICTR SRC meetings
    • Jan has been involved in handling complaints and educating researchers
    • The CRC was handled separately, now things are rolled into one with advent of VICTR
    • Outreach to VU community, Nashville community, Cancer Center
    • Health fairs for potential study participants; IRB web site has a participants' page
    • Video being made for patient TV channel, working with marketing
    • Last year 17 complaints for 3349 active studies, from participants, families, or study staff
      • Example: lack of compensation, lack of time to review study
    • Studies can be audited with respect to the monitoring safely level assigned to them
    • Beginning to meet with PIs
    • Checks for existing plans for example for what monitoring is done if a patient is taken off a drug
    • Jan has reviewed many VICTR applications
    • Research ethics not directly related to regulation or subject protection is not specifically addressed
    • Larry pointed out the usefulness of getting investigators to think about research ethics from the outset of the development of a new study
    • Need to watch out for small financial requests to VICTR that result in expedited reviews have enough opportunity for research ethics review
  • Liz talked about research ethics consultation; she reviewed some of the issues that have arisen
    • Discussed pack of forethought about ethics in a proposed project for which a studio was recently held; investigators thought of ethics as a "just in time" component even though the intervention was dangerous
      • a good example of the inter-relatedness of ethics and study design
    • Research ethics may need its own space on the StarBrite review web page
    • It would be nice if investigators came to clinics in the initial stage and we could alert ethicists to attend


Attending: Heitman, Churchill, Scott, Harrell, Byrne, Phillips, Wang, Clayton

  • Plan for the meeting about voucher-based funding for our kind of projects (specifically having someone review CRC proposals to identify specific ethical questions they raise).
    • No funding for ethics vouchers per se; no funding for qualitative core functions
    • Can a biostatistician do the screening reviews? Need high sensitivity.
  • Look at the “institutional climate survey” designed by my RCR colleague to consider whether it might be useful as an instrument to measure changes in the ethical climate of VUMC research over the course of the CTSA grant.
    • Two other schools interested (both are members of CTSA)
    • Interested in predisposition to cheating/cutting corners
    • General issue of sensitization and other unrelated interventions in any simple pre-post design
    • A general current climate survey may be beneficial
    • Would have been nice to do this before VICTR began
    • Discussed identifiability of respondents in national and local surveys and whether the survey should be national instead of local, after piloting
    • Effect of nonresponse bias could be large
    • Number of questions is too large
    • Would be good to compare with previous survey instruments that dealt with more specific infractions
    • Another consideration is to classify the type of research (e.g., 'omics' research has more temptation)
  • Progress report - need input very soon


Attending: Heitman, Churchill, Scott, Byrne, Yu, Harrell

  • Industrial Pharma Clinical trial reviewed by Frank - What does it mean to be "dangerous", "unethical" for a statistically underpowered study?
  • Study is powered to detect a difference in a binary endpoint; it is over-powered to detect an effect on the underlying scale
  • Discussed problems with CDAI scale; was never a justification of a cutoff of 150 in the original 1976 paper (which also overfitted the predictors of remission). A score of 100 barely reaches a probability of 0.5 of remission.
  • Ethics issues: patients not consenting to be put at risk for a trial that doesn't yield valid results (scientific integrity is the reason we can treat them as subjects in the research); there is an obligation to reduce hazards in the trial
  • Some patients may be enrolled needlessly; trial is longer than necessary; if the treatment is effective, the trial might delay use of the treatment in the control patients who were accrued last
  • Patients rarely see information about payments to investigators
  • Dan proposed an interim analysis; so far there has been no response from the investigator
  • There are some substantial ethical questions that ought to be addressed

  • Investigators are obligated to reduce the number of subjects exposed to harm and the amount of harm to which each subject is exposed
  • Future topic: ethics of using a less informative response variable in a larger number of subjects vs. a more invasively obtained response variable in a smaller number of subjects
Topic revision: r39 - 19 Aug 2014, FrankHarrell

This site is powered by FoswikiCopyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback