When the complication rates drop even further than promised, the provider stands to gain financially; but when the number of exacerbations are too high, they are liable for the additional costs as shown in NHS Oldham Primary Care Trust.
The same principle holds for elective, maternity and acute care, where payment is not 'per year' but per episode. Here too, complications and unnecessary interventions would be incurred at the providers' expense, as would be the case in any industry. Stroke care, care for acute myocardial infarction or hip fracture can, for example, be contracted as a 90-day episode, starting at the admission to the hospital. It would thus include both the acute episode and the direct rehabilitation efforts where the main contracted patient outcomes would be mortality rate and functional status.
For maternity care (which should deliver optimal outcomes for both mother and infant, and an optimal patient experience during the care process) caesarean section rates and hospital admissions that trend above expectations would be incurred at the providers' expense. This would mean that the better the care around the pregnancy and delivery (including a smart redistribution of tasks amongst the medical specialists, midwifes or nurses involved), the better the outcomes and the lower the costs.
In elective care, interventions that do not add value to the patient should also receive no reimbursement as shown in Geisinger's ProvenCare. This not only implies that reinterventions or complications are handled at the provider's expense, but also that the intervention was called for, and that the same results could not have been achieved through non-surgical treatment. This is particularly important when contracting care per episode to prevent non-value added interventions. For cancer care, the best contracting method would probably be a mix of acute (episodic) and chronic (on a yearly basis) periods.
Geisinger's Proven Care
Geisinger's first Proven Care program ensured that patients undergoing an elective Percutaneous coronary intervention (PCI, or percutaneous angioplasty) would always receive optimal, evidence-based care, in which all of the essential steps are taken (unless, of course, clear contraindications apply). The promise Geisinger gives their patients and payers is that one bundled payment will encompass all of the care required for the PCI and its follow-up for 90 days. This includes potential complications, the cost of which will be borne by Geisinger. The introduction of this program has shown markedly improved outcomes such as reducing readmissions by 44 percent, for example. The ProvenCare concept is now also used for coronary artery bypass surgery (CABG), hip replacement, cataract surgery and other elective surgeries. In Stockholm County, a similar model is used for hip or 2 year guarantee on additional surgery to the joint, and a 5 year guarantee for post-operative wound infections
Back to top
Prometheus: episode based payment
Prometheus is an episode-based payment method which aims to optimally incentivize the provider for delivering high quality, low-cost care. It does so by allocating an 'evidence-informed case rate' (ECR) to a patient episode, which includes "all covered services related to the care of a single illness or condition, bundled across all providers". The Heart Failure ECR, for example, would include all healthcare activities (specialist contacts, medication, nursing activities, and so forth) that would together constitute 'evidence based heart failure care'.
The ECR is risk-adjusted, to account for the severity and complexity of the patient's condition: when a patient is in poorer overall health, the ECR is adjusted upwards. In addition, the ECR includes an allowance for potentially avoidable complications (exacerbations, infections and so forth) that could have been avoided by more optimal care. Finally, the ECR includes a negotiated basic margin for the provider(s).
To date, ECRs have been developed for 21 conditions, representing approximately 35 percent of a US commercial payer's total healthcare spend. The categorization used by Prometheus (chronic, acute medical, inpatient procedural, and outpatient procedural) can be easily translated in the categories used here.
Prometheus' price-setting ECR model is rather elaborate (too elaborate for some) due to the deliberate effort to separate 'insurance risk' from 'technical risk'. The former cannot be mitigated by providers (sicker or older patients will usually be more costly to care for), and should be carried by the insurance company; while the latter (preventing potentially avoidable complications and inefficiencies through operational excellence) is exactly what should be carried by the provider. This sets it apart from less refined models where risks are more integrally transferred to the provider, such as population-based capitated payment models.
For all these types of care, payment would be provided on a per-patient basis, with entitlement dependent on having a specific condition (having Parkinson's disease, being pregnant, and so forth). This is different for basic medical care where thresholds for eligibility cannot be defined in any feasible way and where providers are expected to primarily focus on keeping people healthy as much as possible. For such types of care, paying a population-based fee rather than a case-based fee makes more sense. While this is rather similar to the way UK and Dutch GPs are already being paid, it differs in ensuring that a much more direct link should be forged between the outcomes of the care and the payment received. Because of the nature of this care, true 'outcomes' may be harder to define and, depending on the scope of services bundled under this care, more population-based proxy-outcome measures such as vaccination levels, unplanned hospital admissions and appropriate referral measures may be used.
In a system based on these principles, competition between providers on value promises to unleash innovative forces that are rarely seen in most Western healthcare systems, but that international best practices show are possible: elective care without waiting times, acute care with minimized mortality and morbidity rates, diabetic patients with life expectancies equaling people without diabetes and with equal quality of life. The reduction in cost just by avoiding non-value added care are enormous; the additional savings from increased efficiency in the delivery of value-added care are equally large.
Back to top
NHS Oldham Primary Care Trust
In the NHS Oldham Primary Care Trust region, a few General Practitioners realized that they could both increase quality and significantly reduce costs in the area of musculoskeletal disease (MSK): rheumatology and orthopedics. They have set up a service that handles all MSK GP referrals in the region, referring to hospitals only those patients that require clinical admission or surgery. Run by GPs and a specialized nurse, both rheumatology and orthopedic specialist expertise is available as well.
Emphasizing self-management, shared decision making, and steering away from low-value adding activities (such as unwarranted practice variation), they have already succeeded in reducing the number of hospital admissions by half. In addition, they have completely shifted rheumatology care into the community, admitting only eight rheumatology patients into acute trust care in 2011.
With overall MSK expenditure at approximately GBP23 million, the service's current value-based contract promises to reduce expenditures by 14 percent over 2 years while simultaneously increasing health gain (measured by patient reported outcome measures, amongst others). "Vital in achieving integrated care", Dr. Nye (one of the service's founders) says, "is establishing the role of the pathway co-ordinator, a clinical provider, whose role it is to performance manage and integrate care. The payment system should reward this rather than the multiple episodes of often disjointed care."
Back to top
Payment: case-based or capitation?
Internationally, more and more examples of episode-based payment for elective care (including 'guarantees') and year-based payment for integrated chronic care are emerging. Similarly, paying primary care professionals a fixed sum per enrolled patient to cover the basic medical needs for that population (including, for example, basic acute care) is common practice in Europe.
In general, case-based payment makes sense when a condition, the care required for that condition and the desired outcomes can be feasibly circumscribed, as is the case with most of the care being delivered 'downstream' in the value chains of healthcare. The advantage of this model of contracting is optimal transparency, and an optimal incentive for efficiency and productivity. The disadvantage is that by paying 'per case', the contracts may stimulate the volume of unwarranted elective interventions, and will require that care for multiple conditions includes clear allocation of costs to individual episodes of care.
Population based-payment (also called 'capitation') is the better choice for most 'upstream' care, where prevention and coordination are key and the patient's problem may not be easily categorized as a specific, clear-cut 'condition'. The advantage of population-based payment is its simplicity and (from the payers' perspective) the predictability of healthcare expenditures. Yet the potential disadvantage of capitation is that insurance risk is shifted to the provider, since payment is per capita, not per patient. When limited to basic medical care, the amount of insurance risk that is transferred to the provider is limited. Home care, ambulatory care for the elderly and basic acute care (24/7) can also be feasibly incorporated into such a model. The extent to which providers can take on full capitation risk including high cost care such as cancers is open to debate. Asking providers to carry insurance risk, as is the case in ACOs and other fully capitated models, may mean taking on risks which may be outside their control. Groups of over 150,000 population can generally contain most of these risks, others may need reinsurance. Alternatively these patients need to be excluded and case-based payment for these types of care can be used to prevent this from occurring, while also allowing for restricting non-value added (over-) production.
Back to top
Pricing: fixed or negotiated?
In contracting outcomes in this way, healthcare systems can either opt for set prices, or opt for negotiation between the provider and the payer. There is ample evidence to show that when government bodies attempt to set meaningful prices for healthcare services, it often becomes a costly, time-consuming and ultimately hopeless affair. In the Netherlands, for example, (a country with a reputation of precise and meticulous regulatory institutions), laboratory and radiological examinations are priced as they were decades ago, corrected for inflation but not for the technological advances that have taken place since then.12 Setting prices for integrated care products will not be a lesser challenge for regulators. Although, on the one hand, the scale of the overall task will become smaller (by reducing the large number of prices to be set for individual activities to the core integrated services to be contracted), the individual activities will still all have to be figured into the overall price, and the importance of adequate compensation for case-mix becomes greater.
Where competition is a feasible option, the practice of allowing prices to be set in negotiation is a more practical and ultimately more value-producing approach. Again, taking case-mix into account is essential, but those providers with the best quality/price ratio should win out, thereby driving innovation, reducing waste and increasing overall value. Where competition is more difficult to arrange (as is the case in rural areas or for certain high-cost and scarce resources such as acute cardiac surgery services, for example) tendering the care services is an excellent method to drive down costs while simultaneously ensuring high outcomes.
Price competition is not seen as a unequivocally good thing within healthcare: providers and policy makers are often worried that pressure on prices will drive providers to reduce quality. Indeed, when quality outcomes are not measured, this is a real risk. Also, the track record of negotiated pricing in healthcare is mixed: some good examples (driving down the cost of drugs through reference pricing, for example) are juxtaposed with the plethora of bad examples (real and imagined) where prices go up because of the (real or imagined) imbalance between providers and payers at the negotiating table. When outcomes are part of the equation, however, price competition becomes a much more attractive option.
More importantly, in the model outlined here, the product to be priced is meaningful, comparable and measurable. What is paid for is not activities but results, and that shifts the paradigm completely. Given the inverse cost-quality relation that still often exists in healthcare, providers will constantly push the outcomes frontier to achieve both higher outcomes and lower costs simultaneously, thus pushing the relative price-level further down.
Back to top