March is National Kidney Month, so Dr. Michael Udwin, VP of Clinical Transformation at Evolent Health, sat down with one of the pioneers of a data-driven program to improve kidney care outcomes by preventing end stage renal disease (ESRD) in the first place. Dr. Julia Andrieni, Vice President of Population Health and Primary Care at Houston Methodist, shares insights from their new chronic kidney disease program, a collaboration with our partner, Evolent Health.
MU: What prompted you to build a population health program focused on chronic kidney disease?
JA: We know that over half of dialysis starts are done on an emergent basis, with infection-prone catheters, at an average cost of $60,000. We know that getting those patients to nephrologists earlier in their disease progression prevents those sub-optimal dialysis starts -- and prevents chronic kidney disease (CKD) from progressing to ESRD in the first place. We also know CKD is a risk factor for developing other costly and mortal conditions – including an increased chance of heart failure and stroke. And it can contribute to behavioral conditions, too, including depression.1
But the physicians who can best disrupt that progression don't always know who those patients are prior to their emergent dialysis need. Up to 96 percent of people1 with kidney damage aren't aware of it. And 40 percent of CKD patients begin dialysis with no prior nephrologist input.
MU: What goals are you working toward?
JA: We needed to find a way to identify and disrupt those earlier-stage CKD patients before they progressed to irreversible ESRD. Ultimately, we're working toward avoiding the need for emergent dialysis, improving referral patterns so patients who do progress to ESRD are seeing nephrologists in time for fistulas to mature or early cannulation devices to be used instead of catheter-based dialysis starts, and reducing dialysis start-related emergency visits and hospitalizations.
To do that, we're using predictive data to identify patients under the care of our primary care physicians (PCPs) who are potentially in stages 4 of 5 of CKD and help inform their care and nephrology referral decisions.
MU: Talk to us about the approach you've taken thus far.
JA: The focus is always on identifying patients at risk. We collaborated with Evolent Health to identify our own institution's challenges, establish care goals for PCPs, nephrologists and for palliative care providers, identify the data we'd need to mine, and build a program that physicians would be excited to leverage.
In our early research, we realized that one-third of our own primary care patients in stages 4 and 5 of CKD received no follow-up. So we worked with Evolent to identify which combination of aggregated health data could point to individuals with CKD whose disease progression had strong potential for disruption. We also asked PCPs what information they most needed, and then created a CKD Action Roster that we share with them – names of patients that our data has indicated as being at-risk for chronic kidney disease. It includes information like recent renal function test results, suspected CKD stage, physician follow-up dates. We also include referral criteria like care management program eligibility and status, so if physicians determine that a handoff is appropriate, they have the information they and the nephrologist need at their fingertips.
MU: What lessons from this process would you pass along to others?
JA: First, as always, it's all about the patient. Financial impacts are the result of high-quality, high-value care, so we worked first to align our primary care providers on the clinical risks of the population we're talking about.
Second, it's unrealistic to think you're going to immediately impact complex provider behaviors and decisions. Keep it simple. Find high-value actions that are aligned with the full care team, and make requests early and clearly.
Finally, empower, don't command. Our role at Houston Methodist is to help physicians identify patients and inform their decisions, not to make those decisions for them. We work with private and employed nephrologists at Houston Methodist, so co-management and collaboration must be part of the process. And change management on this level must be pulled forward by an aligned clinical team, not pushed from behind by a directive. We've seen that building physician partnerships gives the people in the best position to affect patient outcomes the opportunity to run. I know we've succeeded in changing protocols and building programs that will last when our physicians start volunteering for change initiatives and when we hear them say, “This is fun. I love working with our team.”