Value-Based Reimbursement—once focused on incentives and shared savings—now more often means capitation. Whether adopting Medicare Alternative Payment Models (APMs) or contracting with health plans, physician groups and health systems have signaled greater willingness to adopt these new Risk payment models with their guaranteed payments for attributed patients.
But here’s the problem: If you don’t have good data on your costs, you are flying blind.
Let’s look at the myths and realities of what you really need both to measure your success under global capitation and to ensure that your Risk program is on track to be successful.
Do You Need Claims Data for Calculating and Evaluating Costs?
The short answer is that the more open your system, with patients using practices and facilities outside your own, the more claims data is necessary for calculating key cost indicators. The most important of these indicators is your Per Patient Per Year Costs. That is a key benchmark that you can compare against global capitation levels. While that number may indicate you have a problem, however, it won’t tell you what that problem is.
Your status with payers currently dictates whether you get claims data. ACOs receive claims data from Medicare, and so will provider groups that are Direct Contracting Entities under the Medicare Direct Contracting program. But groups contracting with commercial carriers rarely—if ever—have access to full, patient-identified claims data. That includes Medicare Advantage plans, covering over a third of Medicare beneficiaries.
Organizations that intend to provide care to patients under a global capitation payment, particularly under private insurance plans or employer-direct contracts, must negotiate provisions for regular claims data feeds. For ACOs, Independent Practice Associations (IPAs), and similar Clinically Integrated Networks (CIN) that include many private practices, this data is critical for you to understand how patients are flowing into your system and where they are going for specialty care.
Having claims data is a big advantage in two critical areas: you can calculate the cost of services that have been generated by your patients and your providers, and also see services patients have used from other providers outside your organization.
But, Is Claims Data Enough?
Here’s the other problem that people rarely acknowledge: if you aren’t already in Value-Based reimbursement like capitation, you won’t get claims data until after you’ve signed the agreements and services are flowing. So your retrospective view of patient costs has two flaws. First, it won’t help you decide whether you can survive under global capitation, and second, it can’t help you enough to manage costs prospectively.
A lot of ACOs have built their shared savings strategies on claims data alone; others have simply dug into strategies that seem to make sense, without really focusing on data at all. That may work when no downside Risk is included in the arrangement, but it’s a mistake when real revenues—which pay for provider salaries, rent, record systems, and direct services to patients—are on the line through a fixed fee structure like capitation.
Let’s examine what you can’t see with claims data. Especially if your view of cost data is based on annual costs reflected by claims data, your view of these cost drivers is obscured:
- Longitudinal view of patient conditions, services, and costs;
- Problem lists from the EHR that indicate patient issues with diagnoses outside the claims data parameters;
- Clinical patient data such as lab values, functionality scores, other critical outcome information;
- Referred versus non-referred, or patient choice, services;
- Diagnoses that are filtered out by the payer because they are not primary to the service;
- Quality measure data;
- Trended outcome data.
While claims data may be more comprehensive because it comes from all sources of services (theoretically, which we will address next), the quality of the data is poor from the standpoint of engaging in activities to improve cost performance. In fact, the poor substance of the data virtually limits these improvement activities to service elimination, prior authorization, or coordination of services.
Over-dependence on claims data leads to the perspective that costs are high because of “patient overuse.” This explains why ACOs have focused their attention on limiting post-acute services and emergency room services, and have over-relied on coordination of care activities for patients with high utilization. Claims data is not helpful in engaging providers in a longer-term focus on improving care design, medical and patient decision-making, and clinical cost control, because it lacks the clinical data needed for providers to engage.
ACOs and groups often bemoan the lack of provider engagement in cost, while, in fact, meaningful cost data is rarely provided to physicians at all, especially data of value for clinical improvement.
Collecting EHR Clinical Data Is Feasible for Providers Going into Global Capitation
Years ago, the aggregation of data from provider systems was a big challenge. As a pioneer in such efforts during the early 2000’s, Roji Health Intelligence modeled data using the backup databases of many clinical record systems to create a common for collection of data. Additionally, more private practices used cloud-based or low price EMRs that made data exports almost impossible. Aggregating data from networks having over 300 different clinical data systems, we were able, even then, to integrate all patient transactional and clinical data into a patient-centric database for evaluation of cost and quality.
Now, health systems are bigger and more consolidated, with fewer EHRs. But those clinical systems have interoperability capabilities and are also supported by dedicated staffs of the health systems, enabling reporting of the data. This has supported easy and more current collection of data from Epic and other large systems.
Current data collection from providers is seamless and fast, rarely taking more than a few days for providers to create reports using Roji data specifications, and to submit securely to us. When FHIR v4 becomes more broadly adopted, data sufficiency will continue to improve.
Analytics and data aggregation vendors should be able to easily support clinical and transactional data collection from providers either anticipating or participating in global capitation. Even if not all participating providers have easily accessible systems, the capability to include clinical data from the majority of providers and supplement other data feeds is indispensable.
But Data Does Lie
Aggregating provider data, however, is not enough to support global capitation cost management. The data needs a full vetting and restructuring for clinical engagement.
Anyone who espouses the catchphrase “data doesn’t lie” hasn’t looked at clinical data as it comes from EHRs. Data does lie because it depends on human delivery via both technology implementation and actual data input. Data goes missing, is not collected or entered, is miscoded, uses local generic or outdated codes, or is not coded at all, and is not reported into the export files because it is hiding in some unknown table of the database. Data can also lead people astray, creating inconsistencies in the timeline of the patient—such as diagnoses that are coded after a procedure that imply a complication rather than a preexisting condition.
Data also can be false at point of original capture. The variation in blood pressure values is an extraordinary example of wildly variable values over a single episode of care. Some blood pressures are taken incorrectly by staff, some reflect patient “white coat” syndrome, fear, pain, differing equipment, and practices used in taking the measurement.
When aggregated, such mistakes in data are amplified. That’s why data must be reviewed and validated. But it is only when patient histories are combined in episodes that mistakes come to light. This is a critical function for providers while they review and draw conclusions about cost data that is built on clinical diagnoses and other aspects of the patient. Again, episodes facilitate that analysis by creating the sequence of patient history and events in a way that illuminates what happened and what should be questioned.
The truth of cost data cannot be revealed through claims data because it lacks the robustness needed to create the story. Clinical decisions require supporting clinical data to make sense.
Facilitating the Clinical View of Costs
We have previously addressed several mechanisms that help groups analyze cost information in a way that illuminates clinical processes and treatment decisions. Patient and procedural episodes are just one way to evaluate patient outcomes and costs in the context of such clinical decisions, but they are essential to the development of an effective and efficient specialty network.
Specialty physicians will drive most of the downstream costs for patients attributed to the globally capitated entity. If those physicians cannot see their own procedures and medical episodes along a cost curve, their ability to change that curve is diminished. Episodes have the power to illuminate how medical decisions—such as overlapping multiple procedures in a single patient event—have affected costs and outcomes for individual patients. That can and should lead to internal discussion about best practices that will optimize affordability and produce best outcomes.
There are no shortcuts to successful participation in global capitation. You will need experts to aggregate and structure data for you so that your clinical expertise is not wasted on reviewing false or insufficient data. As providers, only you can create better clinical care design and physician-patient decision-making to produce better outcomes at an affordable cost.
Don’t fly blind into Risk without a clear view of what you need to work on to make your health care achieve more.
Founded in 2002, Roji Health Intelligence guides health care systems, providers and patients on the path to better health through Solutions that help providers improve their value and succeed in Risk.
Image: Jukan Tateisi
