The concept of lifecycle in HTA

While the concept of evaluating technologies across their LC is not new, with discussion in both regulatory (Reference Baird, Banken and Eichler2) and HTA (Reference Henshall, Mardhani-Bayne, Frønsdal and Klemp3;Reference Gutiérrez-Ibarluzea, Chiumente and Dauben4) contexts, the HTA community is increasingly discussing how to apply LC approaches in HTA (Reference Trowman, Migliore and Ollendorf1;Reference Trowman, Powers and Ollendorf5). HTA is defined as “… a multidisciplinary process that uses explicit methods to determine the value of a health technology at different points in its lifecycle. The purpose is to inform decision-making in order to promote an equitable, efficient, and high-quality health system (Reference O’Rourke, Oortwijn and Schuller6).” The definition indicates the potential need to consider a LC approach during HTA.

There is a growing recognition that HTA systems need to adopt an LC approach to respond to the need for value assessment across the life span of health technologies (Reference Trowman, Migliore and Ollendorf1;Reference Regier, Pollard and McPhail7-Reference Tummers, Kværner and Sampietro-Colom12). For example, concern has been raised that changes in the evidence base or clinical pathway might invalidate initial HTA decisions (Reference Kirwin, Round, Bond and McCabe11) or require updated HTA guidance to inform downstream stakeholders (Reference Trowman, Migliore and Ollendorf1). This concern may be particularly relevant where there is high initial evidentiary uncertainty or decision-making risk, for example, related to reimbursement (Reference Thokala, Srivastava and Smith8;Reference Merlin, Street, Carter and Afzali9). It has also been argued that using HTA to routinely assess technologies across their LCs could increase efficiency and equity in managing health resources (Reference Bryan, Mitton and Donaldson10).

LC-HTA has been proposed as a way to manage evidentiary uncertainty (Reference Trowman, Powers and Ollendorf5); address changes in the evidence base, the design of the technology, or the clinical pathway (Reference Trowman, Migliore and Ollendorf1;Reference Kirwin, Round, Bond and McCabe11); and support iterative decision making (Reference Regier, Pollard and McPhail7;Reference Bryan, Mitton and Donaldson10). Despite different applications, these proposed approaches all include well-established components of standard HTA that are applied to varying phases of the technology LC to address specific decision problems (Reference Trowman, Migliore and Ollendorf1;Reference Regier, Pollard and McPhail7;Reference Kirwin, Round, Bond and McCabe11). An LC approach may represent a prospectively planned systematic sequencing of such components (Reference Kirwin, Round, Bond and McCabe11) and may also include some forms of trigger (Reference Regier, Pollard and McPhail7;Reference Thokala, Srivastava and Smith8;Reference Kirwin, Round, Bond and McCabe11), leading to an HTA reassessment. However, there is no consensus definition of the term “LC-HTA” (Reference Trowman, Migliore and Ollendorf1) or how this term relates to similar concepts such as Health Technology Reassessment (Reference MacKean, Noseworthy and Elshaug13), “Living HTA” (Reference Thokala, Srivastava and Smith8;Reference Merlin, Street, Carter and Afzali9), or “Health Technology Management” (Reference Bryan, Mitton and Donaldson10). As in other areas of HTA, a consensus definition would facilitate communication and collaborative action among the diverse stakeholders concerned with developing LC approaches.

Differences among HTA agencies also raise questions about the feasibility of implementing LC-HTA approaches. The varying remits and capacity of many HTA bodies result in the prioritization of single, comprehensive HTA reviews of new technologies following marketing authorization (Reference Bryan, Mitton and Donaldson10;14). There is also concern by HTA bodies about the feasibility of implementing LC-HTA given the additional resource demands that such an activity would entail (Reference Trowman, Migliore and Ollendorf1). For this reason, the HTAi GPF also recommended a need to identify and define the decision problems where LC-HTA approaches can add meaningful value (Reference Trowman, Migliore and Ollendorf1;Reference Thokala, Srivastava and Smith8;Reference Kirwin, Round, Bond and McCabe11). LC-HTA approaches will likely require selective implementation for most HTA bodies and will need to consider both constraints for the HTA body and the potential of the approach to add meaningful value in addressing the decision problem.

Goals of the position paper

This position paper sets out to:

1. 1. define why the HTA community would want to consider LC-HTA approaches;

2. 2. provide a definition for LC-HTA;

3. 3. describe the health system challenges where LC-HTA might offer the greatest opportunities; and

4. 4. develop a framework to conceptualize how LC-HTA approaches might be structured.

Why the HTA community would want to consider LC-HTA approaches

A key driver for why HTA stakeholders are showing interest in LC-HTA approaches relates to the core purpose of HTA. Despite significant diversity in the remit and application of HTA bodies around the globe, the purpose common to all HTA is to inform decision making through the assessment of the value of a health technology (Reference O’Rourke, Oortwijn and Schuller6). As such, there is an underlying principle that is also common to HTA doers, which is to reduce and manage decision-making uncertainty to enable the timely and evidence-based assessment, appraisal, adoption, utilization, and management of technologies in healthcare. HTA can improve the quality of the information for decision making through its processes and methodologies, and by improving the quality of the evidence base, as HTA bodies seek to do through guidance documents and activities such as providing scientific advice. Therefore, interest in LC-HTA approaches is a natural extension of the core purpose of HTA and is needed to inform decision making about the changing value of a technology at different points in its LC (e.g., Reference Gilardino, Mejía, Guarin, Rey-Ares and Perez15).

To an extent, HTA bodies already have mechanisms that can provide information regarding the changing value of a technology via activities across the LC or through special pathways. Pre- and post-launch activities routinely conducted by HTA bodies or other organizations, such as horizon scanning, early dialogue/scientific advice, managed entry agreements, monitoring implementation, health technology reassessment, optimization, and disinvestment, may be considered potential elements of an LC approach (Reference Trowman, Migliore and Ollendorf1). Part of the motivation for HTA bodies to be interested in a more structured LC-HTA approach is that such activities are not always coordinated and may be undertaken by different organizations or groups within and outside of an organization.

The development of an LC-HTA approach can draw lessons from past experiences and recommendations concerning coordination across a sequence of activities and across stakeholder groups. The 2018 review of global horizon scanning activities by the HTAi GPF (Reference Oortwijn, Sampierto-Como, Habens and Trowman16) recommended an LC approach for the purpose of improving the integration of horizon scanning with downstream decision making. Learnings from the Dutch and UK implementation of managed entry agreements (MEA), through the Conditional Financing (CF) and the Highly Specialized Technology (HST) programs, respectively, suggest that some of the key challenges in MEA can be mitigated through preplanning and coordinated data collection, the need for ongoing stakeholder consultation to align expectations and prevent discrepancies between initial agreements and final data at reassessment, and to ensure a strong mechanism for incorporating new evidence into decision making (Reference Makady, van Veelen and de Boer17-19).

Another motivation for HTA bodies is to consider how to address emerging challenges to existing HTA approaches that are arising from rapid technological advancements. For example, the new German regulatory and reimbursement pathway for digital healthcare applications (DiGA) will require adaptation of their current HTA processes in order to account for emerging evidence both as a consequence of limited evidence at the time of product approval and resulting from ongoing product modification (Reference Brönneke, Herr, Reif and Stern20).

Definition of LC-HTA

LC-HTA is characterized by two features that differentiate this concept from other HTA activities: (i) it explicitly addresses change over time, and (ii) it connects and coordinates several distinct HTA activities. While the definition of HTA states that the value of a health technology is determined at different points in its LC (Reference O’Rourke, Oortwijn and Schuller6), this does not mean that all HTA activities are iterative. In practice, most HTA activities across the technology LC represent snapshots (Reference Thokala, Srivastava and Smith8) taken to inform a decision at a single point in time; this includes activities such as horizon scanning, HTA of a technology for market entry, or health technology reassessment of a technology currently in use. By contrast, the purpose of LC-HTA is to manage change over time, whether that relates to an evolving evidence base or a changing clinical context. LC-HTA activities can begin early in the LC of a technology, for example, to inform decisions about the development of the evidence base, or in later phases, such as determining whether change is sufficiently meaningful to require an HTA reassessment. We note that it will be important to follow a deliberative process (Reference Oortwijn, Husereau and Abelson21) to ensure a common understanding among those involved, for example, in defining a threshold for what would constitute meaningful change. The other differentiator relates to the connection of HTA activities. Although closely linked, many HTA activities are standalone; for example, the use of HTA by an HTA body for an initial reimbursement recommendation is not always reliant on information from horizon scanning. LC-HTA implies interconnected, sequential HTA activities that require prospective and systematic planning. Considering these two differentiating features, we propose the following definition: LC-HTA is a systematic process utilizing sequential HTA activities to inform decision making where the evidence base, the health technology itself, or the context in which it is applied, has a potential to meaningfully change at different points in its LC.

Application of LC-HTA

Our definition of LC-HTA implies that there is a broad range of decision problems facing HTA bodies where such an approach could be applied. We have taken the perspective of the prospective development of an LC-HTA pathway by an HTA body.

Some HTA bodies have established special pathways related to specific decision problems that could be understood as forms of LC-HTA. An example of an LC-HTA approach that occurs early in a technology LC is the Early Value Assessment (EVA) program for medical devices that has been developed by NICE. This program includes a sequence of HTA activities that identify and prioritize key areas of unmet need in the U.K. health system and identifies promising technologies in early development. NICE proactively engages with the technology developers with the intention of providing development guidance, support with data collection, and early access to the health system (22). An example of an LC-HTA approach for more mature technologies is the early access authorization (EAA) program managed by France’s Haute Autorité de Santé (HAS). This program is designed to give patients prompt access to emerging therapies before a regulatory authorization or final reimbursement decision (23).

Scenarios where LC-HTA may be applicable

We consider four key scenarios where applying an LC-HTA approach could yield sufficient value to HTA stakeholders to justify the additional resources required. We conceptualize these scenarios as high-level challenges for HTA bodies that may stem from a variety of different decision problems (see Table 1).

1. 1) Uncertainty relating to limited evidence at the time of review. Although uncertainty related to limited evidence is a relatively common criticism by HTA bodies, there are situations where the extent or context of this evidentiary uncertainty is sufficiently meaningful that an LC-HTA could be warranted. One example is where the initial evidentiary package is limited because of an accelerated regulatory approval of a technology based on promising, early data in situations of high unmet need, such as rare diseases. Another example of limited evidence relates to lengthy time horizons for evidentiary uncertainty to be resolved, such as gene therapies where the intervention’s ongoing impact, safety, and durability are unknown.

2. 2) Technology may be modified over its LC. We consider LC-HTA to have potential utility where the technology itself is not static but can change over time to an extent where there would be a meaningful difference if an HTA reassessment were undertaken. Examples of such change could include medical device “incremental innovation” where the technology product design is periodically upgraded. Another example of a more dynamic form of change relates to changes to diagnostic gene panels through either the inclusion of additional markers or a change in the scientific understanding of existing markers. A more extreme version of such innovation would be technologies that change constantly, such as using Artificial Intelligence in health care.

3. 3) A learning curve related to utilizing technology in practice changes its outcomes. The outcomes delivered by an intervention may change through clinician experience and real-world practice. Effectiveness and safety may change as practitioners gain experience with a complex intervention, such as with a surgical robot. In addition, clinical experience over time may change how medical interventions are used in practice. As real-world utilization provides an increased understanding of how an intervention performs in the context of patient diversity and the local health system, clinicians can optimize their utilization of the technology, for example, changing in dose or timing. This can even extend beyond the original regulatory label. For example, in oncology, new pharmacological interventions are often approved using clinical studies on late-stage patients but, once available to clinicians, may become used in earlier-stage patients.

4. 4) Health service context impacts or is changed by the technology. LC-HTA may also add value where the technology impacts or is affected by changes in the context in which the technology is situated. For example, where a technology causes a significant disruption to existing care pathways, there may be value in a reassessment sometime after implementation to review and evaluate the outcomes of that disruption. Where the context changes independent of the technology, such as a change in the care pathway or policy changes related to HTA methodologies or decision-making parameters (for example, where an HTA methodology guidance changes to allow a form of evidence previously not accepted), then an LC-HTA may be of use to steer the development of technology in anticipation of upcoming policy changes or to assess technologies in a care pathway that has been subject to change. This scenario demonstrates that LC-HTA can have applications beyond individual technology assessment, such as a multi-technology appraisal, disinvestment decision making, or guideline development/update.

Table 1. Scenarios where lifecycle approaches might add greatest value

Note: Table 1 presents the four key scenarios where the TF believed that an LC-HTA approach has the potential to make a meaningful difference in addressing a key challenge for traditional HTA. Example decision problems are provided for each scenario in order to show the range of challenges and decision problems that might arise and to indicate that, even within these scenarios, there is variability in the ability of LC-HTA to resolve the challenges. The column “Application of LC-HTA” represents a qualitative assessment by the TF of the ability of an LC-HTA process to address the decision problem in the specific example. Scoring: Low (limited benefit); medium (some benefit, but other processes required); high (likely to resolve the issue). There may be decision problems that are represented by multiple rows. In such instances, the application of LC-HTA to cross-cutting problems may be equally or more beneficial compared to assessing individual rows separately.

A framework for LC-HTA

The breadth of potential decision problems within the four scenarios demonstrates that LC-HTA has a wide range of applications. This led the Task Force to conclude that rather than a “one-size-fits-all” pathway for LC-HTA, implementation will require tailoring to the decision problem. This observation led the Task Force to develop an LC-HTA framework with three key components that can be used to describe an LC-HTA process.

1. 1. Defining the decision problem: Develop a clear decision problem to be used to guide where and why in the technology LC to apply LC-HTA and for what outcome. A key element is identifying whether addressing the decision problem through the additional activity will be sufficiently meaningful to justify the resources spent.

2. 2. Sequencing of HTA activities: To resolve the decision problem, it will be necessary to determine which HTA activities are required and how they should be connected to ensure appropriate alignment, coordination, and predictability.

3. 3. Developing optimization criteria: Development of clear criteria or guidelines to determine eligibility to an LC-HTA process or when a specific step in LC-HTA should be activated to ensure optimal utilization of different steps in the process. An important implication of utilizing optimization criteria, from an efficiency perspective, is a transparent process to determine when certain HTA activities are worthwhile and when they are not. For example, prospectively planned optimization criteria for an LC-HTA process designed to address changes in surgical robot software could help ensure that HTA reassessment is only activated if a software upgrade changes the technology’s effectiveness or safety profile to a sufficient extent where the original HTA decision might be invalidated.

The LC-HTA framework is intended to be useful for describing real-world implementation of LC-HTA approaches and to help structure the development of new approaches. An important additional aspect for this framework will be to decide which stakeholders to involve and for what components to ensure appropriate alignment, coordination, and predictability. Task Force recommends utilizing deliberative processes (Reference Oortwijn, Husereau and Abelson21) and broad stakeholder involvement (Reference Trowman, Migliore and Ollendorf1) is an important consideration for each of the three components of the framework.

With respect to describing existing approaches, we utilized the framework to characterize the HAS EAA program (Table 2) and the U.K. EVA scheme (Table 3). A standardized approach, such as the LC-HTA framework, will support comparison between potentially diverse applications of LC-HTA, demonstrate which aspects might be missing in existing pathways, and offers a way in which to structure the operationalization of new LC-HTA approaches.

Table 2. How the LC-HTA framework would characterize the HAS Early Access Authorization Scheme

Note: The Early Access to Medicines (AAP) scheme (23) administered by the French HTA agency, Haute Autorité de Santé (HAS), is an example of an LC-HTA approach in action. This scheme has two pathways: prior to regulatory authorization (pre-MA) and prior to reimbursement (post-MA). This example focuses on the pre-MA pathway.

Table 3. How the LC-HTA framework would characterize the NICE Early Value Assessment (EVA) Scheme

Note: UK’s NICE has developed the EVA approach (Reference Pichler, Boysen and Mittmann24) for the purpose of identifying, guiding development, and providing early access to, immature technologies that have been identified as having the potential to address key health system priorities. The approach is applicable to medical devices, diagnostics, and digital products.