2015 and Beyond: 6 Predictions for Healthcare and Population Health
In 2015, some shared accountability sponsors will start looking to participating physicians and advanced practice clinicians to assume some downside risk along with the hospital or integrated delivery network sponsor.
This is just one of the predictions I have for the coming year. But predicting trends in healthcare and population health management for 2015 is somewhat difficult because of a number of unknowns. For example, what the Supreme Court does vis-a-vis the Affordable Care Act (ACA), could have major impact. How successful the now Republican-controlled Congress is in obtaining sufficient override-veto support from their Democratic counterparts is another important factor.
But that’s the nature of a prediction – it’s what we think will happen. So, caveats aside, here’s what I predict we’ll see happen in healthcare in 2015 in terms of the building blocks of population health and other factors.
Health insurers and delivery systems seem to be falling into three camps so far as at-risk contracting is concerned:
- “Strategic thinkers.” Those who have strategically embraced the concept of value-based payment and are fully committed to making a transition from fee-for-volume to value-based payment. They are working through the tactical implications (e.g., cash flow bathtub during the transition), but consider such challenges subordinate to the strategic direction. Metaphorically, they have decided that the future is no longer in railroads – they must get into the transportation business. This first group represents a relatively small minority of payers and providers.
- “Bet hedgers.” Those who are taking a tactical, defensive approach to value-based payment. They are dabbling in it, hedging their bets, fearing that if they do not, they may be left standing on the railway platform as the train pulls out of the station; i.e., they are hopeful, metaphorically, that railroads still have a bright future. This group is a larger minority than the first group. When this second group sees that they may lose money on shared risk schemas (e.g., ACO agreements), they threaten to leave, which is most likely why Medicare has decided to delay implementation of downside risk in their ACO program for three more years.
- “Ostriches.” Those who are in denial have stuck their head in the sand, doing only what is necessary to avoid financial penalties imposed by the federal (e.g., CMS) or state governments (e.g., Medicaid or state exchanges).
Because CMS apparently fears that ACO sponsors from the second category will bail out if they are put at downside risk, I do not expect to see much growth in federally sponsored at-risk contracting.
Because the strategic thinker group is looking down the road rather than in the rear-view mirror, they will use the lull in the federal environment to acquire tools and expertise in critical success factors for at-risk contracting, including applications to measure leakage by provider, by procedure, budget to actual financial performance in bundles and capitation and voluntary disenrollment (churn) in their government and commercial at-risk contracts. They will continue to seek out at-risk contracts with trustworthy state and commercial partners who are willing to enter into more enlightened health benefit program arrangements (e.g., ones that include the end-users in healthcare decision-making rather than buffering them from the reality of the market as CMS has done in its ACO offerings).
Another trend that bears close watching is the shift from defined benefit (historical means by which employer-sponsored insurance is provided) to defined contribution (future means of providing insurance) comes into play. We may be at a tipping point (depending in some measure what happens with the ACA) where employers shift from sponsoring a buffet of health insurance options for their employees (defined benefit) to a system where they give each employee a voucher with which to purchase the insurance plan that best meets their needs (defined contribution). This is where the commercial exchanges (e.g., Towers-Watson, Mercer and Aon [Hewitt]) become the stars of the show. They are the best positioned in the commercial insurance market to respond to such a shift to defined contribution. If Congress and/or the Supreme Court eliminate the mandated minimum benefits required under the ACA, then free market forces would be allowed to operate with dramatic lowering of the cost of care.
Population Risk Evaluation
Payers and providers have both jumped into “shared risk” without much analytic ability to predict how well either of them will do. Critical success factors include better analytics infrastructure, transparency and trust.
Traditional insurance underwriting has severe limitations because of the coarseness of the grain (e.g., inpatient, outpatient, procedures, pharmaceuticals, durable medical equipment, et al) and the fact that these actuarial buckets apply aptly to workflow expenses, but not so well to population expenses.
Enlightened payers and providers will be seeking better ways to evaluate risk in a payer population so they can strike deals between payers and providers that are more fair and sustainable. Underwriting will begin to shift from actuarially driven to clinically driven. The latter seeks to reflect the cost of clinically relevant underwriting buckets so the delivery system is taking risk for buckets of cost around which it can organize clinical improvement projects.
These same enlightened participants will begin to appreciate that there is a better schema for allocation of risk and division of labor under which payers and providers assume responsibility for what they do best. There used to be a sign in the old Shakey’s Pizza Parlors that read: “We made a deal with the bank. We don’t cash checks and they don’t bake pizza.”
Insurers are much better equipped and have more experience managing financial risk for outliers, and the larger ones have a broad book of business (risk pool) across which to spread that risk. So, they should take responsibility of outlier/stochastic risk.
Providers are much better equipped to manage utilization of services (e.g., tests ordered in a diagnostic work-up, triage to treatment venue, treatment algorithms for chronic disease patients, clinical indications for referral, clinical indications for intervention, choices of medications and fluids, choices of devices, prosthetics, stents, implants and grafts). So clinicians should be responsible for variation in the ordering of care and the implementation of the orders.
Payer, providers and analytics vendors can team up to construct better risk models to which both payer and provider can relate. Progressive vendors will figure out how to link the administrative codes providers use to generate bills and insurers use to pay claims to clinically relevant diagnostic and procedure groupings of those codes transforming the billing and claims data into information relevant to the clinicians, who are the only ones who can safely change the way the care is delivered and, therefore, improve the quality and reduce the cost. New risk stratification algorithms will be tested especially in the more flexible commercial partnerships. Better risk stratification has the potential to allow broader engagement of doctors and a willingness on their part to assume some level of risk for decision-making they control.
Forward-thinking providers will shift their thinking from inpatient to clinic. Much of what we are doing to prevent things like heart failure readmissions is very inpatient centric. We are closing the barn door after the horse has already left. At some level, admission of a heart failure patient to a hospital in the first place is often a failure of chronic disease management. We should have known sooner that the heart failure patient was gaining weight, that his/her symptoms of shortness of breath, etc., were worsening so we could bring clinic/outpatient resources (e.g., care management) to bear before it was necessary to admit the patient. This pertains to the provider side of the equation, but it also pertains to the patient side. We will begin to see organizations improve the patient portal part of the system, so that we improve the ability of the system to receive patient-reported data, which is a very important part of a distant-early-warning system for acute exacerbations of chronic diseases.
Some of the “super high risk” patients need to be steered toward hospice care rather than continuing to employ a tour-de-force approach to them. Their protoplasm is so shot that nothing we do in the course of expending tremendous resources is going to make any real difference. We are, in fact, prolonging death, not life.
Once sponsors of at-risk contracting are awakened rudely by the reality of the cost of leakage of patients outside their shared-accountability network of hospitals and clinicians, they begin to examine their network for gaps in coverage. Most sponsors of these clinically integrated networks find themselves needing to augment their owned network of providers with non-owned, affiliated hospital facilities and/or physicians. Expansion of these virtual networks will bring with it a host of problems ranging from issues about governance of the sponsoring entity and assumption of risk for financial performance to the need to integrate data from disparate EMRs with the sponsors data systems, which will drive the need for data warehousing to make analysis of network performance feasible.
As I mentioned previously, during the coming year, some shared accountability sponsors will begin to grow weary of bearing all the risk and will begin to look to participating physicians and advanced practice clinicians to begin to assume some downside risk along with the hospital or integrated delivery network sponsor. Heretofore, in order to get up and running the sponsor often assumed full responsibility for the risk and set up a schema that buffered the clinicians from any measure of risk, fearing that if they tried to commit the docs to downside risk, no one would participate. In larger, more competitive markets, some of these sponsoring organizations will have a run at a transition from treating physicians as customers to treating them as partners.
Another area of focus, especially in the more collaborative commercial partnerships, will be to test and refine attribution algorithms used to assign patients to providers. The CMS attribution algorithms that assign Medicare beneficiaries to physicians and other providers are badly flawed. Surveys done by one ACO sponsor have demonstrated that Medicare beneficiaries who were initially attributed to certain doctors, but were then attributed to different doctors in subsequent contract years, in fact still identify the original physicians as “their” physicians.
It is important that the clinicians feel ownership and are able to fingerprint the way the system is designed. They need to be comfortable that it is clinically relevant and that reasonable measures have been taken to stratify the risk (i.e., the physician gun of “my patients are sicker” needs to be unloaded). They need some time to become comfortable with the idea that a credible link can be created between the financial (billing and claims system) and the care delivery system. Once they are engaged in what they are reasonably assured is a fair and balanced system, they then need to be given information and tools to help them gain confidence that they can succeed – that there really is an upside for them and the patient. They have to have confidence that those administering the plan understand that not all variation can be eliminated; i.e., that a well-tested clinical guideline will likely only cover about 80 percent of patients. Physicians need to feel they will be free to vary randomly along with their colleagues around the other 20 percent being empowered to adapt to the specific circumstance of their patients.
Engaging physicians in a partnership takes time and considerable transparency and willingness to admit that all the variation is not due to clinical practice differences (assignable variation). We need to admit that a fair share of the variation is due to data system problems – data quality assurance, data governance issues, et al. And we need to not just pay lip service to the data quality issues. When physicians point out data quality problems, we need to fix those problems so that when the next report comes out, physicians see that we listened when they said that a patient was dead or had moved. That is one way trust is established in the partnership.
Quality and Safety
We will continue to make progress in measuring and improving clinical effectiveness and safety. The contractual engagement of NQF by AHRQ has contributed significantly to improving clinical relevance of the CMS measures. Efforts to reduce the burden of chart abstraction to report required measures will continue. Commercial payers, including commercial exchanges will reinforce the importance of maintaining or improving quality while reducing cost.
The availability of flexible analytic software will increase in the marketplace as will evidence-based best practice knowledge asset starter sets, which can jump start engagement of clinicians. It is always easier for clinicians to react to a straw dog than to have to come up with best practice guidelines from whole cloth. As such, best practice starter sets becomes more prevalent, they will act like accelerators allowing organizations to manage an order of magnitude more improvement initiatives than when they must re-invent the knowledge wheel with every new project.
Patient injury prevention will begin to shift from incident reporting to regarding it as a process failure. This will require that patient injury prevention be laid out as a process flow so that process failure tools can be applied (e.g., fish-bone diagrams and other cause-and-effect tools).
Driven by the need to improve cost structure in order to be financially viable under at-risk contracts, organizations will begin to dissect waste into its three component parts: 1) ordering waste; 2) workflow waste (inefficiency); and 3) defect waste (complications incurred in delivery the care).
Catalyzed by financial incentives, providers will begin to triage variation into desirable and undesirable. We cited the fact that well-written guidelines apply to the care of about 80 percent of patients with a given condition. Variation and adaptation by providers of the care for the other 20 percent of patients is desirable. However, variation from the guideline for care for patients to whom the guideline did apply is wasteful variation, which needs to be driven out.
Providers will begin to discover that some tools used to wring out undesirable variation are the same for the three types of waste (e.g., outcome improvement packets with their goals and aim statements), but other tools are different (e.g., care improvement maps work better for reducing ordering waste, whereas Lean value stream maps work better for reducing workflow waste).
As sponsoring organizations get further and further into improvement initiatives during the coming year, it will become more and more apparent that the complexities of the competencies outlined above require a robust electronic infrastructure to provide the analytics and information needed to manage such a complex system. Many sponsors will discover that their old point solution applications have created a spaghetti maintenance nightmare for their IT departments and will go to HIMSS seeking more robust systems that can be installed quickly and are flexible enough to deal with the ever-changing landscape and complexity of patient care
Summary and Conclusion
Times of turbulence and change are times of opportunity and 2015 appears to me to provide great opportunity for those who are prudently adventurous. It is a great time to be working in healthcare as the shift from volume-based to value-based care takes place. The future belongs to those who are able to think outside the old box.
Would you like to use or share these concepts? Download this presentation highlighting the key main points.