Thomas R. Graf, MD
WE ARE IN a unique time in healthcare. Although there are periodic crises of cost in medicine, this time we have a convergence on quality that has produced a consensus around 3 things: quality measures, the ability to measure quality in a granular fashion, and the desire for patient-oriented outcomes. This convergence has created a sharper focus on accountability for quality. Combining these new foci has resulted in a drive to improve the value proposition of medicine in America.
Currently we spend more than $3 trillion annually1
on care that is highly variable in terms of quality, clinical outcomes, and cost; diferences in spending can be 10-fold or greater.2
Given that we are paying a signifcant premium for healthcare relative to other industrialized nations, there should be a more consistently excellent product.3
So, as healthcare providers and healthcare payers are under intense scrutiny to reliably deliver measurably better health, they are looking to all elements of the health supply chain to deliver that value.
At the same time, the cost to bring a therapy through the FDA-approval process keeps rising, and pressure from generics and alternative medications is growing. Increased sophistication of the mechanisms of action of these new drugs, along with expectations from venture capital frms and shareholders, has pushed the overall cost of pharmaceuticals past what we spend on physicians in the United States. Today, pharmaceutical spending is second only to the spend on hospitals.4
This has led to potential innovation in the pricing and positioning of new million-dollar therapies, as the drive to deliver true value for patients/plan members intensifes.5
For many years, payers and, to some extent, providers have been shifting from traditional fee-for-service care toward models that reward results. Tentative steps such as paying for process metric achievement for primary care physicians to more quality gated gainsharing models have given way to far more performance risk sharing, where providers are focused on total medical expense budgets or partial capitation agreements. Some states, such as California and Maryland, have even more advanced models. Despite these advances, the share of healthcare spending tied to these new models remains small and the pressure to develop ways to drive quality improvement as a cost-saving measure continues to mount.6
Additionally, as new and very expensive therapies are developed, the desire to ensure efcacy at these prices is even greater. The need to show that a medication can change a disease trajectory and afect an outcome that patients, providers, payers, and employers (or whoever the ultimate payers is) care about is paramount.
This begs the question: what does each group care about? The simple answer is that each stakeholder cares about many things and many concerns overlap7
- Patients care about the impact on their lives and functional status. They care what a drug costs them out of pocket, and they care somewhat about the overall cost.
- Employers have similar concerns, but they focus more on cost than individual responsibility.
- Providers care about tho same outcomes as patients and employers, but with a greater focus on medical quality in addition to impact, and are interested in intermediate outcome measures far more than patients.
- Payers care about medical quality and increasingly about total cost, but they also worry about the customer service experience of members and how drug costs afect employers.8
The other important consideration is the time horizon of the impact. For patients, it is their lifetime; for employers, the duration of employment; for payers, the period of coverage; and for providers, the length of time they spend caring for each patient. These diferent time horizons create additional challenges in thinking through how to create a successful value equation.9
The ideal is a program that improves medical quality in a way that is visible and important to patients and lowers the total cost of care in a year or less (for Medicaid, perhaps in a month or less). This is very hard to create. Readmission reductions programs are a great example: They have a big impact and quickly and defnitively impact cost. Medical quality is often more about the reliable delivery of the best care.10
This creates a real challenge for pharma. While most drugs take considerable time to afect outcomes that matter to patients and translate into cost reductions, the impact on medical quality may come quickly. In diabetes, for example, although many medications have been shown to reduce glycated hemoglobin levels within several months, which makes doctors happy, they take far longer to impact heart attack rates, which is a measure that patients can appreciate. Insurers, meanwhile, often don’t see value until an avoided heart attack translates into lower costs for patients—who are not admitted for the heart attacks they don’t have. And the employer won’t see the impact until enough employees avoid heart attacks to actually bring down future rate increases. This is a signifcant issue now that some drugs cost hundreds, thousands, or tens of thousands of dollars per month.
For patients with diabetes, thus far, the proprotein convertase subtilisin/kexin type 9 (PCSK9) inhibitor, evolocumab (Repatha) has been shown to reduce cardiovascular (CV) events (eg, hospitalization, heart attack and stroke), which is an incredibly important outcome for a patient. This happens in a clinically relevant time horizon (median 26 months), which is important for patients and their employers and from a public policy standpoint. Politicians have chosen to weigh in on mandating coverage of even marginally efective treatments (Ornish intensive cardiac rehab for Medicare for instance) and a case could be made that the evidence is powerful. However, insurers will be challenged by creating appropriate rules so that this medication is used only for those patients who will truly beneft because of implementation issues.11
Clearly, reducing CV events will be important for both quality and cost. However, since many patients are currently well treated with much less expensive medications, and many more could be if they were maximally managed, there is a risk that patients will receive these $14,000-per-year drugs when they really do not need them.12
If patients are exposed to side efects and complications of PCSK9 inhibitors, this would impact the cost savings they create. Additionally, all cause morbidities and mortality and cost need to be considered. Reducing CV complications alone is not enough.
An analysis of the total medical expense would be helpful and key to the real value equation. Understanding the quality improvements, both on CV outcomes and overall health, is the frst step. We need to know mortality impact, morbidity impact, and utilization impact to measure the positive effects. We also need to know the full cost, not just of the medication, but of managing the complications, the natural course of progression, and the cost of monitoring. From an employer and patient perspec tive, the lost work time and other “life” efects are very important but largely invisible to the insurer and to the physician.
There’s also the challenge of defning a realistic comparator. Total medical expense, a likely best candidate, is often harder to determine than it would first appear—and absent a solid comparator—the relative impact is impossible to judge. Pharma has ofered several interesting ideas for how this might be managed:
- The challenge of rebates. We are now well past the era of using volume discounts to help control costs. The goal today is controlling the total cost of care, especially pharmaceutical costs, which have outpaced the overall medical spend. Some have looked at rebates or refunds if certain complications occur, such as heart attacks.13 The challenge is that these practices do not necessarily support the optimal use of the medication, that they encourage widespread marketing to the lowest-risk patients to ensure complications do not occur too frequently. This also does not get at the underlying issue: the therapy may simply cost too much. Offering a 30% reduction for certain patients on a drug that is 50% overvalued does not control costs.
- Pay-for-performance. The idea of not charging patients who do not respond to a treatment is another approach. This model is uniquely suited to the high-cost specifc-use medications being introduced in cancer care; it would fail for other drug classes where determining response is more difficult. To truly develop comprehensive accountability, a program that includes annual total cost of care and trend impact is needed. The challenge here is that there are many factors outside of a specifc disease and certainly beyond a drug’s impact that afect the total cost of care. Many pharmaceutical companies have developed or partnered with others to provide wrap-around services to improve adherence and modify lifestyle—offering stress reduction treatments or help with exercise and diet to support improved outcomes and reduced healthcare utilization. However, these services may interfere with similar programs at the health plan or provider level. Also, these programs are either not explicitly tied to cost or only focus on the cost of the specifc diagnosis or disease state. This makes perfect sense from a pharmaceutical and provider perspective. However, from the payer, employer, and patient perspectives, programs that reduce CV costs and utilization but raise costs in other areas are not helpful. Proving causality or even an indirect relationship between the two is even harder.
- Making partnerships scalable. For this approach, a collaboration between physicians, other providers, and drug manufacturers would help with the all-encompassing nature of total medical expenses. However, this only becomes practical for companies that have medications for multiple disease states and systems that have large numbers of patients. For insurers, this works only if the same pharma–provider coalitions care for a signifcant number of their members. The practical application of these global innovations is challenging.
What can we really do? How can we move forward? It is clear that there is no silver bullet to improving quality while driving down cost in healthcare; in the pharmaceutical arena, as well. A tailored approach blending all the ideas will be necessary. We also need to look for related areas in which we can eliminate non–value-added costs from the system. Examples include site-of-service issues, where the infusion location often translates into a 2- to 10-fold cost difference, or cost-plus-percentage markups for medications that work well when medications are priced in the hundreds of dollars but fail when prices reach $500,000. These are simple changes that can impact cost that do not really create accountability but do rationalize the overall pharmaceutical spend.
Creating a direct link between the impact of the proper use of the medication and the corresponding price paid is critical to the long-term success of healthcare. The keys are:
- Reaching agreement among payers, providers, and pharma on a process for measuring the direct and long-term impacts, quality, utilization, and cost of new therapies.
- Agreeing on a process to ensure selection of the optimal patients for each new therapy, along with a mechanism to create accountability for patient selection.
For many, the idea of a rebate for treatment failures or complications makes sense; for others, the idea of overall cost of care for a specifc disease category on an annual basis works; for still others, By working through these areas and tailoring the methodology to the disease state and specifc medication profle, we can best make lasting progress to drive quality improvement to reduce total cost of care rather than just cutting a few dollars today.