Digital Health Technologies: Key Considerations for Reimbursement and Access
In the first plenary session, we heard about the Digitale Gesundheitsanwendungen (DiGA) pathway in Germany, a fast-track process for gaining reimbursement for digital health applications that are considered to be low-risk technologies (Figure 1).1 This process involves provision of initial evidence of benefit leading to short-term reimbursement by sickness funds (which cover approximately 70 million people), followed by a year–long data collection period over which benefit must be proven. This pathway seems particularly appealing for small, start-up companies, as the barriers to reimbursement seem apparently lower and less costly than if a large scale RCT was required up-front. However, whilst conceptually appealing, the speakers highlighted that, in practice, the process has not always worked well. Decision makers are becoming increasingly stringent in terms of the initial evidence that they are willing to accept, meaning that small scale RCTs may be required, followed by a larger RCT at one-year; this is despite the fact that only technologies judged as low-risk are being considered for this pathway. These evidence requirements may be prohibitive for companies from a cost perspective. Some participants have even dropped out of the process given a lack of agreement on what evidence would be considered sufficient for reimbursement. A key topic of discussion was the generalisability of DiGA-type schemes in other countries, with one speaker highlighting the challenges of applying such a process in the US market given the multitude of different payers in the healthcare system.
In another session, a speaker from NICE presented the latest iteration of the NICE evidence standards framework for digital health technologies.2 This framework, like DiGA, aims to streamline the process for reimbursement of and access to medical technologies by providing information on the minimum requirements for technologies at different risk levels. It will be interesting to see how effectively this framework is utilized, and whether this will influence similar processes in other geographies and make the reimbursement pathway more accessible to MedTech companies that may not have extensive, or even any, HTA experience.
Important considerations with regards to data privacy were highlighted in the context of leveraging digital data to ensure patient-focused decision-making.3 One session highlighted potential issues with regards to consent bias (the notion that those patients who are willing to share their data may be systematically different to those who are unwilling), which may have significant implications when drawing conclusions from digitally derived data sets. This issue is not likely to be easily solved; on the one hand, if you bombard patients with information as to what their data will and will not be used for, with the aim of improving transparency and thus likelihood of consent, this may actually have the opposite effect if patients become paralysed by this information. On the other hand, a lack of information may either lead to patients consenting by default without fully understanding the implications or refusing to consent given privacy concerns. A patient representative highlighted that patients may be more likely to consent to the use of their data if they know that it would be used to benefit them in the long-term, emphasising that closing the loop and feeding back to patients about the usefulness of their data is a vital step. Whilst consent and consent bias may not be fully ‘solvable’, alternatives were proposed such as federated analysis, where data are analysed in separate datasets and then combined based on aggregated parameters (meaning that individual patient data do not need to be gathered in a single location), as well as the creation of non-identifiable patient data.