Risk Analysis Index (RAI) Consortium Tool
A RAPID MULTI-METRIC FRAILTY ASSESSMENT TOOL
During the 2015 Annual meeting of the AVAS we were treated to a panel discussion on ‘THE PREOPERATIVE ASSESSMENT AND PREPARATION OF THE SURGICAL PATIENT’. During this presentation Daniel Hall, MD, MDiv, MHSc, FACS (Department of Surgery University of Pittsburgh and VA Pittsburgh) presented his group’s view on “THE FRAILTY ASSESSMENT: WHAT TOOLS TO USE”.
At the 2016 annual meeting Dan Hall, et al, presented two papers on the utility of the RAI consortium tool:
‘Preliminary validation of the Risk Analysis Index (RAI) of frailty’
‘Feasibility of a rapid, objective multi-metric frailty assessment for clinical practice setting2.
Dr. Hall acknowledged that the “eyeball” test has likely been a fairly accurate assessment of the ability of a geriatric patient to tolerate surgery and the recovery process. It is likely even more accurate if one takes the time to perform a detailed geriatric assessment, but that is often not possible in a busy clinic. Other issues with a simple “eyeball” assessment include lack of reproducibility and the inability to provide objective data for charting, data mining or training purposes. It is simply not enough to write that “I looked the patient over and he should do fine” or “he is not going to do well with this operation as he is too frail”.
To help guide clinical judgments during preoperative assessment, several meaningful frailty indices and scores have been developed. However, we still need a quick test which can be performed in clinic to eliminate the majority of patients who would do fine with surgery and would not need further preoperative preparation. Unfortunately, there does not seem to be enough data to support one test or the other. Additionally, the majority of research is focused on precise and reproducible tests that simply do not transition to the busy clinical environment we face. There has been some merit in the gait speed and timed get up and go, but it is not clear if by themselves, these will afford the accuracy we need.
Dr. Hall, went on to explain that one promising option is the Risk Analysis Index (RAI) adapted from the Minimum Data Set Mortality Risk Index (MMRI). The MMRI originally was developed to assess predicted mortality in patient admitted to the Nursing Home. Dr. Jason Johanning, from the Omaha VA, developed a revision of the MMRI called the RAI. The RAI is a 14-item survey that measures 11 variables and takes less than 2 minutes to generate. Dr. Johanning has utilized the RAI since 2011 in over 11,000 patients scheduled for surgery. It is now in use in the Omaha, Pittsburgh, Atlanta, Indianapolis, Phoenix, Nashville VA's, as well as several private hospitals, including all 23 hospitals in the University of Pittsburgh Medical Centers (UPMC). It has been successful in its predictive value of Mortality and Morbidity and identifies the 5-10% of patients who are likely to be frail.
Using the RAI, the team was able to predict increasing mortality and morbidity rates with increasing scores. For instance in patients with scores less than 15 (77% of the cohort) the 30-day and one year mortality rates were 0.3% and 1.6%. For each 10 point rise in the RAI score, they saw a rise in one year mortality from 1.6% to 7.3% to 17% and up to 25%. Morbidity rates also saw a similar climb with increasing RAI scores.
During their presentation the team released a link to a trial version of their online tool designed to assist providers in obtaining meaningful RAI scores. The survey takes less than a 2 minutes to complete, and comes in two forms. The first version includes only the RAI (https://vhacdwweb05.vha.med.va.gov/surveys/?s=ELMTJMLFT7). The second begins with the RAI, and if the patient scores >21, an additional module opens that permits calculation of the gait speed, grip strength, and both the Hopkins and Edmonton Frailty Scores (https://vhacdwweb05.vha.med.va.gov/surveys/?s=NLXYDKJALT). Please note that the link is accessible only from VA computers, and should be used only for trial purposes (please do not use for clinical care). A version suitable for clinical care can be obtained on request to Dr. Hall
Dr. Hall agreed to answer a few questions of ours.
In considering the practicing physician, I would need to know what is the best methodology I have in my armamentarium for assessing a ninety year old patient who has a large reducible inguinal hernia that he says is getting more and more painful and more difficult to reduce as time goes on so I can explain my decision to schedule the operation right then or to hold off for further assessment and potential preoperative treatment strategies (assuming he is otherwise fit for surgery).
You stated that once you have used your screening tool and you have identified the 5-10% of patients who “Might be Frail” you proceed with a more rigorous and precise assessment. What exactly does that entail for your group?
ANSWER: If your 90 year old patient scores below 15 on the RAI, I would be confident scheduling the operation. If, however, he scores higher, I would want to take a closer look. No existing measure of frailty captures the overarching syndrome, but if multiple measures all indicate significant frailty, that is information that would make me reluctant to operate, and it would inform the shared decision-making process with the patient. It’s not feasible to measure grip strength or walking speed on everybody that walks into your clinic, but we have developed an integrated and rapid protocol to “confirm” frailty for patients scoring above 21 on the RAI. It takes about 6 minutes and involves answering a few more survey questions as well as measuring grip strength, gait speed, and a timed up and go—all you need is a stop watch, a chair, a hallway and a grip dynamometer. The protocol results in 5 validated measures of frailty that can help clinicians understand the nature and severity of frailty in the specific patient. You state that options include 1. a phenotypic characterization using Linda Fried’s (Harvard) methodology of walking speed, grip strength, shrinking, exhaustion and low activity or 2. Kenneth Rockwood’s (Dalhousie Univ. Halifax Nova Scotia) use of seventy variables. Can you expand on which approach you use and describe how you do this in practice?
ANSWER: The multi-mode frailty protocol described above renders gait speed, grip strength, Fried’s Frailty Score as well as the Edmonton Frail Scale that is an example of Rockwood’s accumulation of deficits model. If the patient flags as frail on the RAI, but none of these other measures indicate increased risk, I think it is likely safe to proceed. If, however, the RAI, Fried & Edmonton frailty scores all confirm frailty, the decision to operate becomes much more risky.
On a side note, it may be helpful to know that Rockwood has found that you don’t need all 70 variables to predict frailty. In fact any well-chosen 10-12 variables is almost as predictive as the entire collection of all 70 variables. Based on that research, the RAI, Edmonton and modified Frailty Index (mFI) all employ a more parsimonious group of frailty-relevant variables.
What is the advantage of this approach?
ANSWER: Frailty can be a bit like the elephant in the room of blind men (and women)—each feeling a part of the animal, but none is able to see the entire beast. You may not be a “believer” in any given frailty measure because they all have limitations. But if the patient (or surgeon) is skeptical that the RAI score is accurate, the multi-mode assessment let’s you test their hypothesis. If none of the other measures show frailty, their skepticism is vindicated. But if multiple measures show frailty, it’s harder to ignore, and it’s harder to “blame” the identified frailty on the limitations of a single score. There’s an elephant in the room, and both you and your patient are going to have to do something about it.
If indeed you now determine that the patient is frail, what is your next step in your algorithm in preparing the patient and their family in a decision for surgery?
ANSWER: Talk to the patient. Share the diagnosis of frailty and its significance, and see if that changes the patient’s mind. For example, if your 90 year old with the hernia has a 25% chance of dying in the next year (because his RAI score is 35), he may be more willing to endure the hernia-related pain. But who know, maybe the pain is so life-limiting that it is worth the risk.
It is important to remember that we can and do operate safely on even the most frail patients. For example, our data show that fully 60% of patients with RAI>35 are alive five years after their surgery. But many of these are at dramatically increased risk of death and disability. And the risk of disability is key. Studies show that older patients frequently receive more invasive treatment than they want. They also show that if a treatment has a high risk for disability or loss of independence, older patients frequently prefer treatments that trade quantity of life for better quality of life. In our practice, we take the diagnosis of frailty as an opportunity to invest time in clarifying the goals of care, establishing advance directives, and identifying surrogate decision makers. We’ve found that palliative care colleagues can be helpful with this—so long as you make it clear to them that you aren’t asking them to shift the goals of care so much as you are asking them to help clarify what patients really want—their hopes, their fears, the things that make life worth living, and the outcomes they consider unacceptable.
Finally, and especially if the patient elects to proceed with surgery, we think it is critical to get everyone on the same page. If at all possible, the decision making process should include not only the patient and surgeon, but the patient’s surrogate and representatives from the team of clinicians who will provide perioperative care to include anesthesiologists, critical care specialists, nurses and palliative care. It may be possible to mitigate frailty-associated risks through preoperative strength training or intraoperative anesthetic management. Getting everyone on board early helps identify these opportunities. It also helps reduce what I call “premature withdrawal of care”—by which I mean those times when the surrogate decision maker gives up too early because they are distressed by the nature of ICU care required to rescue patients from the kinds of complications that were predictable by virtue of frailty. Even if the patient and I chose surgery with our eyes wide open to the risks, and even if we negotiated an agreement for a 2-week time limited trial of postoperative rescue therapies, if we don’t involve the surrogate and other involved clinicians in this preoperative discussion, the long-shot of surgical therapy in the context of frailty can succumb to a failure of nerve. Don’t get me wrong: patients and their surrogates can and should change their minds in the setting of changing clinical realities. But we know that recovery will be long and difficult for the frail. They need to know this up front. Many will choose alternative options. But for those who want to proceed to surgery, the ultimate success of achieving their goal (such as return to a the nursing home to play scrabble with grandchildren) depends on a commitment from the entire team and family to give the patient a chance to recover from the index operation and the predictable (but not devastating) complications that will likely follow.
We were told that the NSO was developing a Frailty score. Do you have enough information to expand on that and explain its merits? How will your team’s RAI tool help with this process?
ANSWER: We do not have enough information regarding the NSO frailty assessment parameters for us to know how this will mesh with existing knowledge but we are looking forwards to the addition to Frailty measures to the NSO toolbox .
Let’s say I am a novice in the field of geriatric patient assessment and want to understand more. How do the American College of Surgeons Geriatric Guidelines fit into this equation? What do I do with their information regarding a patient in my office for a preoperative appointment? (link below).
ANSWER: The ACS Geriatric Guidelines are a great resource to learn more about the enormous field of geriatrics as it pertains to surgery. It presents a comprehensive review of the range of issues that require particular attention in geriatric populations. Although some surgeons may master this material, it is likely beyond the scope of most. And thus timely consultation to our geriatrics colleagues may be in order to complete the kind of comprehensive geriatric assessment recommended by the ACS guidelines.
In addition, it’s worth noting that the ACS is currently partnered with the John Hartford Foundation in a multi-year project to develop a ACS-sponsored quality program for Geriatric Surgery that would establish standards of excellence similar to existing ACS-sponsored programs for trauma, cancer, or bariatrics. I think that program will develop more concrete and actionable guidance. I’ve participated in the first two rounds of that program development and it looks very promising. You can learn more here: https://www.facs.org/quality-programs/geriatric-coalition
Mohanty S, Rosenthal RA, Russell MM, Neuman MD, Ko CY, Esnaola NF. Optimal Perioperative Management of the Geriatric Patient: A Best Practices Guideline from the ACS NSQIP/American Geriatrics Society. Journal American College of Surgeons 2016;222(5):930-947.