Monday, August 19, 2013
Present: Kat Cheung, Kairn Kelley (by phone), Amanda Kennedy, Rodger Kessler, Ben Littenberg, Charlie MacLean, Connie van Eeghen
Guest: Thomas Joseph visiting from India
1. Start Up: Next semester’s schedule will NOT be Thursday’s at 11:30 but from 12:00 – 1:15, pending Abbie’s approval, starting August 29. Sylvie to send out meeting invitations. Introductions all around. Thomas is a medical graduate from India, now working as a house officer in India, and visiting to observe how research works in the U.S. and at UVM.
2. Discussion: Kat: Proposal to identify whether measures of frailty can help prognosticate older adults with chronic kidney disease, looking for general input about protocol
a. Background: End stage renal disease takes up over $20billion in Medicare funds; it usually affects older people. Interested in how people make decisions around dialysis, especially aged 80 and up. Many older adults will die within 6 months of starting dialysis (~20%); this increases to ~40% at 1 year. How to advise an 80 year old patient on whether to start dialysis?
i. Some studies exist (US/UK) comparing populations that did/didn’t choose dialysis. If an older person does not start dialysis, death usuallyoccurs in 3-4 weeks. No data exist that helps predict the outcome of dying within 6 months. Having the data will help inform not only whether to dialyze but what methods to use (catheters vs. other routes of access)
ii. Will measures of frailty predict these outcomes? Previous study at Stanford focused on veterans with GFR < 20 and are therefore making a decision on dialysis. Six months of follow up outcomes, including mortality, morbidity, decisions about whether to withdraw from dialysis, are pending.
iii. It could be possible to enroll more subjects in Stanford, but it would take an investment in time.
iv. There is no guideline for dialysis re: age; recommendations address only the most extreme cases of patients not able to receive the benefits of dialysis. There is an administrative process to set dialysis up, but as it is covered by Medicare it is a treatment that is often applied more easily than withheld.
i. Effects a large part of the population; decision rules or framework would be welcome by providers
ii. Would support shared-decision making, in an environment that allows decision-making in the on-going process of care (e.g. a patient may start dialysis on a trial basis, and then be helped to make further decisions about continuation)
c. Other resources:
i. Ursula McVeigh: research on deciding whether older patients should get their hips fixed. See Amanda for an introduction.
d. Preliminary research question: Does frailty predict mortality for end stage seniors (65+ years)
i. Is this already known? Or is this question about how to measure and what the scale is for making the decision?
1. Frailty is measured by physical or functional limits; hard to operationalize (e.g. gait speed)
2. There may be other measures of frailty: social, psycho/social, and other aspects of vulnerability who need support – this could lead to a descriptive analysis, with qualitative data, but is a different study
3. The data must be collected prospectively; they don’t exist in our data sets
ii. Does dialysis make the patient more frail?
iii. Is this a prediction model or is the study evaluating an association? “Here are the variables that matter” or “Here is a weighted formula?”
iv. What are the other variables that affect the outcome: age, sex, education, income, polypharmacy… a predictive model can look at more than GFR, frailty, functional status… and considers them all.
1. If interested in frailty, then control for these variables as possible confounders.
2. If interested in predictive advice, then look at the cancer literature/research as a model
v. Other good measures to consider
1. Functionality: SF36
4. Frailty: does it have a crisp definition? Freid definition: exhaustion, weakness, gait speed, weight loss, and other elements – can this be improved on? Flourishing scale, grief/depression scale… ?
vi. Ultimate goal: decision-making about going onto dialysis
1. Deciding what to use as a set of measures depends on the PI’s interest:
a. Test the variable of interest, the outcome, and any confounders
b. Test the predictability of an outcome: start with a large group of variables and filter (a much larger project, with many thousands of cases)
c. Test a new index for prediction (fewer cases)
e. Analysis: Consider survival analysis, which could include the Stanford data. A Kaplan-Meyer analysis provides the opportunity to evaluate censored data: patients who dropped out of the study, patients who withdrew from dialysis, and mortality. May also be able to evaluate hospitalizations.
i. Consider analyzing the first study, and then deciding what the next question will be from that specific population (veterans). By end of 2013, will have 6 months of data on them all and can move quickly into publication.
ii. At FAHC:
1. Obtain list of those with renal replacement therapy
2. Obtain list of GFR < 20
3. Look for the patients on List #2 and not on List #1
iii. Consider the new research direction around “Patient Centered Outcomes Research” (PCOR), in which patients are engaged in defining the research question
iv. Consider the psychological variables of those who cope well with decision making and those who cope poorly, which is related to immediate and long term outcomes
g. Next steps: access the data…
a. August 21: Abby: Opiate project update (no Amanda, Kairn?)
b. August 29: Rodger: Review of Lexicon Checklist
d. Future agenda to consider:
i. Peter Callas or other faculty on multi-level modeling
ii. Charlie MacLean: demonstration of Tableau; or Rodger’s examples of Prezi
iii. Journal article: Gomes, 2013, Opioid Dose and MVA in Canada (Charlie)
iv. Ben: Tukey chapter reading assignments, or other book of general interest
v. Summer plan: each week, one person will send out an article or prezi ahead for review or discussion by all. Alternatively, if a participant is working on a key document for their professional development, this is also welcome (e.g. K awards, F awards, etc.)
Posted by Connie at 8/19/2013 11:38:00 AM