Ingolv Urnes and Eric Bender have had time to draw breath and think about an exciting NAACOS (National Association of ACOs) 2024 Spring Conference.
VBC momentum, initially slow, now seems unstoppable, primary care is key
Accelerating growth in VBC contracts is apparent looking at revenue figures from large health systems. For example, Iron Mountain has gone from zero to roughly 40% of revenue from VBC contracts in the last ten years.
NAACOS chair, Clif Gaus, noted his surprise by how interested NAACOS-affiliated organizations have been in taking risk.
The ‘2030 Target’ was brought up many times; this is CMS’ target that by 2030 all Medicare and most Medicaid should be based on VBC contracts. In line with the 2030 Target, several participants noted that CMS is chipping away and reducing fee for service (FFS) codes.
The US spends a smaller proportion on primary care than other developed countries; going forward it is recognized that most of the VBC action will be in primary care (see below on Care Management) and the integration between primary care and pharmacy. CMS is also of this mindset with the introduction of the ACO Primary Care Flex Model, a voluntary 5-year test model starting 1 Jan 2025, which is aimed at growing participation in ACOs and MSSP and increasing the number of people with Medicare in an accountable care relationship. It focuses on primary care delivery in MSSP and will test how prospective payments and increased funding for primary care in ACOs impact health outcomes, quality and costs of care.
Multiple Payers a challenge for Providers
It was universally recognized that the work of providers is made complicated by serving patients covered by multiple payers and programs – from state-managed Medicaid of many flavors and various Medicare programs to commercial insurance.
The take-away was to enable providers to treat all patients the same and for ACOs and other payers to deal with the billing complexity behind the scenes.
This payer agnostic approach would allow an effective, single approach to population health and care management.
Different ACOs have different needs and strategies re Provider Engagement
Provider engagement was the hot topic. Clearly, different organizations have varying needs to engage providers depending on strategy and target patient cohort. For example, organizations focusing on high-risk or high needs (such as Cityblock) will focus all energy on the individual patients, whereas others (such as Aledade or Pearl) will provide primary care with analytics and a more hands-off, recommendation-based approach.
Care Management is core to VBC
It seems universally accepted that care coordination results in better engagement and clinical outcomes and reduces cost.
However, several research papers suggest that human-led care management is too expensive to run. Many delegates were interested in how our AI Nurse, costing a tenth to twentieth of human-care coordinators, could deliver care coordination.
Another care management challenge is the risk that multiple parties – payers, providers, primary care – are all trying to manage the patient simultaneously leaving patients confused.
Start with Attribution… ‘Who are my patients?’
It was suggested that when developing a VBC or ACO strategy, the first question to ask is ‘Who are the patients attributed to us?’.
Big picture there are two approaches: retrospective (most MSSP) and prospective (e.g. ACO Reach). From the panel debates, it appears attribution is a complex issue which takes up a lot of time.
MSSP vs ACO REACH
Several sessions addressed the differences between MSSP and REACH. One consultant concluded ‘MSSP is reliable, REACH is exciting’.
The current REACH program is ending in 2026; consensus is that REACH will be folded into MSSP, and that CMS is working to more tightly align reporting requirements between the various programs.
ACO REACH – Three Key Measures
Success in the REACH program is driven by three key measures: (i) all cause admissions, (ii) unplanned admissions (COPD, CHF, Diabetes etc.) and (iii) timely follow-up.
According to several REACH organizations, a real challenge is that a small number of unplanned admissions can move you from top 80-percentile to bottom 20-percentile and significantly negatively impact contract economics.
The challenge of collecting SDoH has also mentioned by several REACH organizations. From 2024 REACH ACOs are required to collect and submit data on SDoH (on top of the demographic data required from 2023) for aligned beneficiaries. The HEDR adjustment is applied to determine the REACH ACO’s Total Quality Score and final Quality Withold Earn Back.
As an aside, several delegates noted that for example BP readings can be collected any-which-way, as long as the process is auditable – opening the way for using AI Nurses to reach out and collect this very cost-effectively.
Medicaid: Patients harder to engage
Not surprisingly, the conclusion of a session dedicated to Medicaid was that this population is even harder to access and engage. Accessibility, Accessibility, and valuable impact of Social Workers was repeated by all presenters.
From Alaska we heard about high no-show rates having a particular impact on rural providers.
The general poor maternity service in the US was also discussed, with presentations from several panel members discussing both heart-warming and heart-wrenching stories. One success story was at Jefferson (PA) where the provision of meals for high-risk pregnancies for 12 weeks; another was the engagement of patients in laundromats.
FQHC at the center of VBC?
Several presenters suggested that FQHCs have the infrastructure and mindset to deliver VBC at scale, provided they are supported with data and risk management tools.
One presenter noted that ‘it is about care management (CM) and utilization management (UM)’ and challenged the audience to consider what are actual impactable costs. As an example, for patients with COPD or high SDoH burden the impactable costs are substantial, for patients on dialysis maybe less so.
As regards the SDOH burden, an interesting point was made that this correlates strongly with high utilization for old patients, but less so for single moms (who, it is assumed, prioritize staying at home looking after the children).
Pop Health Fundamentals
Everyone agrees that Pop Health is critical, however, several delegates pointed out that one does not need AI for risk identification; existing statistical methods are more than strong enough.
There was significant focus on health equity. The introduction of a Health Equity Index (HEI) was discussed and seen as a valuable tool as the ‘right thing to do’ and it has the potential to enable better cost management.
There were several presentations on ‘passive monitoring’ of EHRs to detect risk, particularly that of hospital admissions. One delegate talked about eFI – Electronic Frailty Index – as a proven, strong indicator of admissions risk. Pearl Health’s signal concept was similar; here Pearl is looking for signals to then commence care management.
Brain Health and its impact on effective VBC is a topic several delegates see becoming an important consideration. With an ageing population and increasing prevalence of dementia, Alzheimer’s, and other related conditions, ‘brain health’ is, as noted by several delegates, going to become an important consideration in providing good, low-cost healthcare.
Fundamentally Pop Health needs to move from an obsession with analytics and data to ‘so what?’ What actions can we take to bend the cost curve? Clearly, using Pop Health signals to trigger effective care management and programs to drive patient engagement and self-management is key. Given staffing shortages and cost, it is also clear that this is an area where the AI Nurse will become critical.
Future of VBC? AI Nurse Florence provides Automated Care Management
Better care coordination, better patient engagement, frequent nudging to support self-management is crucial to deliver better outcomes and lower cost. Given staff shortages, we must find more effective, lower-cost ways of delivering Care Management.
Our AI Nurse Florence automatically engages patients with ‘simple’ text powered by AI. She combines RPM, medication reminders, personalized coaching in English and Spanish and is proven to be often easier to answer candidly to. Florence is non-judgmental but persistent, where other clinicians might be reluctant or too busy.
The AI Nurse will save you significant clinical and admin time while delivering better patient experience and improved clinical outcomes:
- Hypertension: Clinical time down by 75%, 10-fold reduction in admin tasks, 97% medication adherence, sustained BP control
- Diabetes: 64% less time spent by clinicians, A1c reduced by 1 point
- Heart failure: Hospital admissions reduced by over 65%
- Patient Net Promoter Score: Consistently above 75%
For further details see www.generatedhealth.com
Or contact
US: Eric Bender ebender@generatedhealth.com
UK: Kylie Dentith kylie.dentith@generatedhealth.com
Australia: John Griffiths john.griffiths@generatedhealth.com
