It’s that time of year where we’re busy heading into the final quarter of the year. In 2018, we have witnessed a fundamental change in the concept of value, going far beyond pharmacological properties and clinical effects, to not only to include patient value, quality of life and patient management, but also to integrate systems and societal value dimensions more broadly.Where are we in our journey towards the new frontier in the ‘paying for value’ discussion in the US?
Stage 3 Adapt Agreement and Payment Structures to Identified Risks
When drafting the terms of the agreement, multiple considerations are required from a legal, operational, and strategic risk management perspective. The set of legal ambiguities encountered within different countries, and especially in the US (e.g., anti-kickback laws, Medicaid’s “best-price” rule, 340B ceiling price, among others) are certainly challenging but can still be navigated as we have argued in the second part of this report. It is important to also note that many of the operational concerns are second-order challenges, meaning that our willingness to dedicate time and significant resources to address them should be predicated upon the likelihood of success, based on the best estimates of performance under real world conditions. No matter how well we navigate the legal and operational hurdles in the implementation, an OBA that fails to deliver due to miscalculated real world performance is one that can and should be avoided. Thus, our core interest here is in advancing guidance as it relates to the question how to make an agreement perform in the first place. At the outset, it is important to look at the strategic risk management component, which should be the very core of the agreement. Many of the initial
agreements that have been negotiated aimed simply at guaranteeing the same performance in actual practice as was observed in the clinical trial, with partial or full refund when this performance is not reached. While this structure is convincingly simple, it might be inadequate from a risk management or a population health perspective.
For example, a common risk to which manufacturers are exposed with OBAs in chronic diseases is the so called channeling effect. In these conditions, early adoption of newly launched treatments tend to happen disproportionately among patients with more severe disease as they are the ones with the highest switch rates and because these are the ones for which physicians often have no alternative treatment. As a result, in the first year of launch of a new chronic treatment we can often expect to see a high share of severe patients receiving the new treatment (at a higher proportion than in the clinical trial and a higher proportion than other comparator treatments), which will likely decrease progressively over time as the adoption broadens. As observed outcomes tend to be poorer among more severe patients, manufacturers engaging in OBAs for chronic conditions such as multiple sclerosis or cardiovascular diseases face a risk to refund a significant share of their sales in the first years after launch. Anticipating this type of issue is critical to avoid disincentives for manufacturers and potentially provision the agreement to limit risk. For example, a simple cap on the percentage of total refund could be included in the agreements, or patients with specific characteristics could trigger lower refund amounts. Channeling is one among the various uncertainties and risk that need to be managed when embarking on an OBA.
Another interesting example of robust planning and risk management is provided by the two agreements between Amgen and Harvard Pilgrim for Repatha discussed in the previous section. The first agreement, signed in 2015, aimed at guaranteeing the same LDL control in the plan population witnessed in the clinical trial, while in the second agreement signed in 2017 Amgen rebates the cost of Repatha for patients with at least 6 months of treatment who experience a stroke- or MI-related hospitalization. Interestingly, both agreements comprised an exclusion clause of patients that would not meet a minimum level of adherence to treatment, considering that adherence to treatment is in many ways a factor that the health plans do have some ability to influence. This clause was particularly important to secure a fair agreement as one can expect adherence to a cardiovascular treatment to be lower under real world conditions than in the RCTs, which can be problematic as adherence typically has an impact on effectiveness. Hence, the provision in the contracts was appropriate, as the real world adherence is likely to negatively affect the observed real world effectiveness of Repatha but this represents a shortcoming of the health system and not necessarily of the treatment itself. Another possible approach could have been to further enrich the agreements by adding disease management programs that can enhance patient adherence (cf. section 6). Finally, two other interesting features of these agreements are that the first one also included a traditional clause of rebate on volume and the second one includes a refund of out-of-pocket costs. These show that hybrids of financially based and outcomes-based agreements are possible and perhaps prominent, and that the value of the outcomes based agreements can go not only to the payers, but also reach the patients.
In conclusion, whether your OBA is set up as a population-level or patient-level agreement, modeling the expected outcome in the health plan population is a critical step to identify key uncertainties and to develop a robust sales forecast that can guide decisions regarding the contract, such as timing and target level of the outcome, as well as potential provisions regarding specific use of patient population factors. For more on this stage of the contracting process, please have a look at this guided seminar on the gross-to-net implications of prediction-driven outcomes-based agreements here: bit.ly/oba-seminar