Evidence inventory: A patient-centered framework for Centers for Medicare & Medicaid Services to assess the clinical benefit of drugs to inform its maximum fair price negotiation

Abstract

The Inflation Reduction Act passed in August 2022 empowers the Centers for Medicare & Medicaid Services (CMS) to negotiate maximum fair prices for certain expensive drugs. In the process of determining maximum fair prices, CMS gathers evidence on the clinical benefits of these drugs as compared with their therapeutic alternative(s) from various sources. As patients are the primary beneficiaries of the treatment, we recommend that CMS should embrace an approach that prioritizes patient experience when evaluating such diverse sources of evidence in the assessment of the clinical benefit. Thus, we propose to draw on several existing frameworks to support the concept of “evidence inventory,” a patient-centered approach to systematically evaluate benefits of drugs under consideration. This 4-step process to develop an evidence inventory includes the following: (1) Formulate the research question—in a PICO(T) (P = population, I = Intervention or exposure, C = comparator, O = outcome, and T = time frame) format, (2) Synthesize evidence, (3) Evaluate the evidence—using evidence inventory, and (4) Reevaluate evidence as new information become available. Patients and other relevant stakeholders play a critical role in each of these 4 steps. The proposed evidence inventory holds the potential to provide a structured and transparent patient-centered framework for evaluating the clinical benefits of drugs as compared with their therapeutic alternative(s) informing CMS maximum fair price negotiation.

Plain language summary

The Inflation Reduction Act became law in August 2022 and allows the Centers for Medicare & Medicaid Services (CMS) to negotiate prices for expensive medicines. We suggest a simple patient-centered 4-step process called an “evidence inventory” to evaluate the benefits of medicines.

Implications for managed care pharmacy

Our proposed framework provides a structured and transparent way to evaluate evidence and incorporate patient perspective in informing CMS maximum fair price negotiation. Changes in drug pricing policies can influence patient outcomes.

The Centers for Medicare & Medicaid Services (CMS) is the federal organization that offers health coverage to more than 160 million people in the United States.¹ Medicare is a health care insurance program managed by the federal government and overseen by CMS.² More than 63 million beneficiaries are enrolled in 1 or more parts of the Medicare program: Medicare Part A (hospital service), Medicare Part B (physician service), and Medicare Part D (prescription drug).^2,3 In 2021, approximately 49 million (77%) Medicare beneficiaries were enrolled in Part D and the total Part D drug expenses exceeded $200 billion.⁴

The Inflation Reduction Act (IRA) was signed into law on August 16, 2022, to strengthen the Medicare program by reducing Medicare drug expenditures and provide meaningful financial relief for millions of people.⁵ In line with this goal, the “Price negotiation program to lower prices for certain high-priced single source drugs,” section 1191 of the IRA, allows CMS to negotiate maximum fair prices with the manufacturer for certain older drugs lacking generic or biosimilar competition.⁶ The initial batch of 10 drugs subjected to price negotiation includes apixaban (Eliquis); etanercept (Enbrel); sacubitril/valsartan (Entresto); dapagliflozin (Farxiga); insulin aspart (Fiasp); insulin aspart (Fiasp FlexTouch); insulin aspart (Fiasp PenFill); insulin aspart (NovoLog); insulin aspart (NovoLog FlexPen); insulin aspart (NovoLog PenFill); ibrutinib (Imbruvica); sitagliptin (Januvia); empagliflozin (Jardiance); ustekinumab (Stelara); and rivaroxaban (Xarelto).⁷ From June 1, 2022, to May 31, 2023, these drugs collectively cost Medicare $50 billion—equivalent to roughly 20% of Part D prescription drug expenses.⁷

CMS considers various criteria to negotiate for the “maximum fair price” of the selected drug, including Part D net price and/or the Part B average sales price of therapeutic alternative(s) as an starting point of an offer, evidence related to benefit of the drug as compared with the therapeutic alternative(s), costs of research and development, and cost of production and distribution.⁸ The initial offer undergoes adjustment based on the clinical benefit of the selected drugs as compared with their therapeutic alternative(s) generated from various sources. Although the list is not exhaustive, these sources can range from clinical guidelines to patient opinion. As patients are the ultimate users of these drugs and possess unique insights into practical aspects of living with the disease and its treatment, we hold the view that such evidence should be evaluated based on patients’ experience.^9,10 Consequently, there is a need for a patient-centered, clear, transparent, and rigorous framework to harmonize various sources of evidence. IRA Section 1194, negotiation and renegotiation process subsection (b) negotiation process requirements (1) methodology and process, also mandates that “the Secretary shall develop and use a consistent methodology and process, in accordance with paragraph (2), for negotiations under subsection (a) that aims to achieve the lowest maximum fair price for each selected drug.”⁵ Thus, we propose drawing from several existing frameworks to develop a patient-centered evidence inventory to assess the clinical benefits of the drug as compared with its therapeutic alternative(s) to inform CMS maximum fair price negotiation.

Evidence Inventory

The evidence inventory proposes to create a structured table that comprehensively and systematically summarizes and weighs various sources of evidence for clinical benefits of drugs under consideration.^11-13 Some methodological concepts for evaluation of clinical benefit are adopted from the US Food and Drug Administration (FDA), ie, “Benefit-Risk Assessment for New Drug and Biological Products Guidance for Industry”¹⁴ and Cochrane handbook for systematic reviews of Intervention.”¹⁵ Using our proposed evidence inventory framework, we then discuss the approaches to capture opinions and experiences of patients, caregivers, and experts (researchers and clinicians). It is our expectation that this inventory serves as a clear, transparent, and rigorous patient-centered tool to appraise current evidence on clinical benefits of drugs under consideration systematically and quantitatively with the ultimate intention to guide CMS’ maximum fair price negotiation. Figure 1 presents a general process of evidence evaluation.

FIGURE 1.

Systematic Process of Evaluation of the Evidence

Formulate research question—The first step toward building an evidence inventory is to develop a well-formulated research question that defines the target population (P), intervention(s) or exposure (I), comparators (if applicable) (C), outcomes(s) (O), and time frame (T).^16,17 PICOT framework structures the research question by identifying a specific and focused therapeutic context for which the drug is indicated. A well-defined research question is crucial for assessing whether the current drug(s) under consideration fulfills the specific unmet needs of the population—patients who do not respond to available treatment.¹⁴ Additionally, to ensure outcomes important to patients are incorporated in the decision-making process, patients experienced in the condition need to be involved in formulation of the research question and in identification of patient-relevant endpoints.¹⁸ For this purpose, a committee comprising the relevant stakeholders, such as patient(s), caregiver(s), researcher(s), and clinician(s), along with CMS, can be formed and given the responsibility to cocreate a draft research question. Existing frameworks for evaluating patient-centered impacts, value elements, and experience will aid the process.^19-21 The finalized research question is then circulated to larger patient and clinician communities with experience in the condition for review and comments.

Synthesize evidence—Once the research question has been identified, in addition to the manufacturers’ submission, current evidence on the clinical benefits can be extracted from the systematic literature search process.^14,17,22 The systematic literature search process entails searching for academic and scholarly materials, such as journal articles, books, reports, and other sources relevant to the research question. A literature review team can be formed, consisting of patient(s), caregiver(s), researcher(s), clinician(s), and CMS, which will then use scientifically rigorous methodology (eg, Cochrane review) in the literature search and evidence synthesis.¹⁵ There are a range of important roles in the literature review process, some of which require research experience or technical expertise. Numerous resources are available for training purposes as needed.^10,18 Patient(s) and caregiver(s) have an important role in shaping search strategies, defining inclusion and exclusion criteria, extracting evidence to capture outcomes drawn from their real-life experiences, and interpreting the findings.¹⁸ Evidence exists about the positive benefits of including patients in this process.²³ Brief steps for evidence synthesis include (1) developing a clear search strategy (eg, defining search databases, time period, and search terms), (2) defining the criteria for study selection (eg, therapeutic indication, target population, study design, comparators, language, and settings), (3) searching for evidence/literature, (4) extracting data (eg, study characteristics, methodologies, key findings, uncertainty of evidence, and heterogeneity), (5) assessing the risk of bias for all included studies (eg, AMSTAR2 criteria for systematic reviews, financial and professional conflict of interest for clinical guidelines), and (6) synthesizing evidence if deemed necessary (eg, meta-analysis).

As shown in Supplementary Figure 1 ^{(194.9KB, pdf)} (available in online article), evidence can come from various sources—for example, meta-analyses, randomized controlled trials (RCTs), real-world evidence such as observational studies, comparative effectiveness research, patient-centered outcome research, expert opinion, patient input, and other sources such as the FDA, online resources, clinical practice guidelines, and so on. CMS has suggested that they might also consider cost-effectiveness studies if allowed by law, as long as they do not discriminate against extending the lives of senior, disabled, or terminally ill individuals compared with younger, nondisabled, or non–terminally ill individuals.²⁴ Thus cost-effectiveness studies can also be considered as an evidence source for evaluating drug benefits. To summarize individual study findings, an evidence table/spreadsheet can be crafted reporting heterogeneity, biases, uncertainties, and other information relevant to the research question. Diverse study types possess varying degrees of strength in contributing to evidence. For instance, a meta-analysis of RCTs may yield the most robust evidence of clinical benefits compared with observational studies. Thus, sources of evidence can be grouped by study types (eg, RCTs vs observational studies).

Evaluate evidence—The next critical and most challenging step is to evaluate and weigh the evidence from various sources to assess the clinical benefit of the drug under consideration. An evidence review committee consisting of patient(s), caregiver(s), researcher(s), clinician(s), and CMS can be formed to evaluate such benefits. The committee can rate the certainty of the evidence for each study using a standardized version of an established evidence assessment criteria such as AMSTAR2²⁵ for the systematic reviews and GRADE²⁶ criteria for the RCTs. Both methods rate the quality of the evidence using 4 levels (high, moderate, low, and very low). Similar to this approach, the National Heart, Lung, and Blood Institute rates the quality of the controlled intervention studies, systematic reviews and meta-analyses, observational cohort and cross-sectional studies, case-control studies, pre-post studies, and case series studies on a rating scale: good, fair, or poor.²⁷ Using one of these methods, CMS can assess the confidence level of all its information sources.

Next, develop an evidence inventory (see Table 1), a structured table that consist of 6 distinct components/sections: (1) target population (PICOT), (2) all sources of evidence, (3) relative importance weights, (4) quality of the study—previously determined by the evidence review committee—based on predefined levels (1 = good, 0.5 = fair, and 0 = poor), (5) clinical benefits as compared with its therapeutic alternative(s) treatment, and (6) comments. Three of these sections—target population, all sources of evidence, and quality of the study—are described in previous sections of this article. Relative importance weight is characterized as the importance score assigned to an individual source of evidence. Several methods can be leveraged to generate relative importance score/weight from diverse stakeholder perspectives including patient(s), caregiver(s), researcher(s), clinician(s), and CMS, for example, approaches such Best Worst Scaling²⁸ or discrete choice experiments.²⁹ Such practice will increase consistency and transparency in generating evidence weights. Patient involvement in the process to generate evidence weight enhances patient-centered assessment of clinical benefits and concurrently establishes the correspondence and credibility of the findings, ie, “respondent validation.”³⁰ Informed by the evidence review committee, the clinical benefit section of the evidence inventory can categorize whether the retained source of evidence indicates the presence of clear clinical benefit of the drug under consideration over its therapeutic alternative(s)—coded as “yes” = 1 for clear benefits, “unclear” = 1 for uncertainty, and “no” = 1 for no superior benefits. Finally, the comments section provides additional context and clarification.

TABLE 1.

Example for Evidence Inventory

Now, let’s imagine we are trying to figure out how well a drug provides clinical benefit as compared with its therapeutic alternative in reducing mortality. We can now collect this information through an evidence inventory using 3 distinct columns, each emphasized by a black box in Table 1— the “relative importance weight” in section 3, “quality” in section 4, and “clinical benefits” in the reduction of mortality represented by column “Yes” in section 5. Now we do simple math—multiply these 3 columns. For example, for manufacturers submission, relative importance weight (0.15) is multiplied by the quality (1) of evidence and yes (1) for clear evidence of benefit, else yes = 0. The result is 0.15. Now, repeat this process for each source of evidence (for each individual row) and add all the results. In our example, the resulting sum of product is 0.7 (see Table 1 for detailed calculation). This indicates that, after accounting for the quality of the evidence source, there is 70% confidence that the drug under evaluation is superior in reducing mortality as compared with its therapeutic alternative(s) treatment. The total possible score for yes, unclear, and no will add up to 100% for each treatment outcome if all included sources of evidence are determined to be of good quality. Additionally, when comparing multiple therapeutic alternatives, one can generate separate evidence inventory for each comparator alternative and then summarize the findings (Supplementary Table 1 ^{(194.9KB, pdf)} ). Such information regarding the clinical benefit can inform CMS’ maximum fair price negotiation.

Reevaluate evidence—The evaluation process should be iterative as new evidence becomes available. Thus, mechanisms such as reconvening the evidence review group to reevaluate the available evidence using evidence inventory should be a part of the reevaluation process. The duration for ongoing monitoring can be established in accordance with legal requirements and determined to be suitable by both the evaluation committee and the pharmaceutical manufacturer.

Implication for the Stakeholders

The concept of the patient-centered evidence inventory provides a structured and transparent methodology for evaluating the clinical benefits of drugs and improving the decision-making process and has several implications for patients, CMS, clinicians, and manufacturers. The involvement of patients in the evidence evaluation process ensures that patients’ perspectives and needs are considered, resulting in more patient-centered decision-making. Patients will have access to clearer information about the clinical benefits of specific drugs, aiding them in making informed health care choices. Consistency and transparency in determining maximum fair prices are enhanced, aligning with the goal of reducing drug expenses for CMS while ensuring patient access to affordable treatments.

Clinicians may find it easier to discuss treatment options with patients when there is clear and transparent information about the clinical benefits of drugs. Additionally, their expertise is valued in the evidence evaluation process, allowing them to contribute insights into the clinical benefits of drugs based on their practical experience. Drug manufacturers will face negotiations with CMS to establish maximum fair prices, which may impact their revenue from certain high-priced drugs. This framework for evaluating clinical benefits outlines precise criteria for assessing the robustness of evidence, granting drug manufacturers the opportunity to bolster the weight of evidence in favor of their pricing.

Nevertheless, it is crucial to acknowledge that the proposed framework has not undergone empirical testing. Consequently, CMS may face several challenges in its implementation. Key considerations include, first, owing to a lack of information or biases in the selection and inclusion criteria, the proposed inventory might not encompass all relevant information, leading to an incomplete understanding of the clinical benefit. To address this, it is recommended that diverse stakeholders verify the completeness of such information. Second, although certain evidence types, like RCTs, have well-established quality assessment criteria, others, such as opinions from patient advocacy groups, may lack established assessment criteria. As a result, CMS may need to rely on self-reported conflicts of interest in determining the quality of such evidence. Predefined inclusion/exclusion criteria for inclusion of advocacy groups and the information might reduce such biases. Third, the relative importance weight of evidence can be influenced by the preferences and composition of stakeholders, potentially introducing bias in making comparisons across its therapeutic alternative(s) treatment. Such biases might be reduced by diverse representation, conflict of interest disclosure, rotating membership, training and standardization, and public engagement. Fourth, proprietary information from manufacturers may not be openly discussed among all stakeholders, posing a challenge in developing relative importance weight for various sources of evidence. In such cases, CMS may rely on their unbiased professional judgment. Lastly, resources and time constraints might be a challenge in the implementation of such a framework. Outsourcing, seeking collaboration, and prioritization can help make the most of available resources.

Conclusions

The evidence inventory can serve as a patient-centered tool, providing a structured and transparent method for evaluating various sources of evidence on clinical benefit informing CMS fair price negotiations. CMS should consider potential sources of biases that could impact the evidence inventory framework for impartially assessing diverse sources of evidence.