Build, Prove, Launch: Facilitating better digital health evidence generation

• Applying a strategic approach to research supports effective product design as well as efficient evidence generation for new digital health solutions
• Approaching research as an iterative, three-phased process, saves time and resources by helping digital health startups avoid common mistakes
• To develop a successful research road map, startups need to identify the key questions of the research phase with the primary audience(s) for the findings and align those to the research methods employed

The number of digital health startups has grown rapidly over the last decade, with clinical evidence to support their claims lagging far behind.¹ Reasons for this gap are complex and range from a lack of research expertise to funding constraints. Indeed, startup founders often don’t know where to start when it comes to digital health research. When asked what word or words first come to mind when they hear the word research, the vast majority of founders will respond with some mix of “complicated, expensive, time-consuming” and “hoop” through which to jump. It’s understandable why many startups perceive or experience research in this way.^1,2 When approached without a cohesive overall plan, research can be exactly that: overly complex, unnecessarily time intensive, and a barrier rather than a facilitator to adoption.³

The need for evidence is clear, and thoughtful research is the path to saving time and resources.⁴ What if, from the very beginning, we in the digital health innovation ecosystem leveraged research as the useful, essential tool it’s meant to be? What would that look like? What could it do for us?

In this article, we share a few of the common, costly mistakes startups make with regard to research, and outline a three-phase framework for making smart investments when it comes to research.

Research can prevent a variety of costly mistakes if you know when and how to use it effectively. The path from idea to solution is anything but a straight one- nor should it be. We know from decades of design research that the best solutions arise from teams that embrace agile, iterative cycles of development and testing, and an openness to needed pivots. Design in healthcare is no different. While a research path is rarely a straight one either, a strategic research plan serves as a map guiding us toward engaging, evidence-based solutions, and around the pitfalls that commonly sideline digital health startups.

What makes research strategic? Strategy, put simply, is a plan of action to achieve a larger goal. When digital health companies undertake research piecemeal, it’s clear why research has a reputation as time-consuming and expensive. Anyone who has designed or executed a research study knows that research studies require decision-making, planning, and coordination. The most costly mistakes companies often face are a direct result of a failure to incorporate research early and often according to a cohesive, forward-looking plan. Startups commonly focus exclusively on demonstrating impact on health outcomes. Failing to consider other factors, like usability, in a research study may mean having to design and conduct an entirely separate project to satisfy the research requirements of regulators or potential investors.

Any digital health research project is well served by starting with a general research roadmap. We recommend this roadmap feature three crucial phases: BUILD, PROVE, and LAUNCH. While generally happening in sequence, a strategic research approach lends itself to revisiting each phase as needed to establish the appropriate level of evidence to push a solution forward from product development (BUILD), to showing safety and effectiveness (PROVE), and finally establishing marketability and scalability (LAUNCH). This helps startups by preventing them from undertaking a later stage, large-scale study before they are ready while using earlier-stage research endeavors to ramp up to larger studies so that when the time comes, they are well prepared for success and maximize the use of money and resources.

Critical to each phase are key research questions for the purposes of either exploration of unmet needs and user journeys, or for confirmation of hypotheses and assumptions driving solution design. Consumers of the information generated in each phase may be primarily internal, external to the company, or a mix of both. Additionally, each phase centers around the type of research questions of interest, and whom that evidence is meant to serve.

The BUILD phase is centered on exploration and ideation related to product development. Key research questions include: What is the user journey for this patient or clinician? Where are the unmet needs and priorities? What factors influence the patient’s behavior or decisions? In this phase, we begin to refine our understanding of the problems we aim to solve, uncover potential opportunities for impact, and detail important context to inform solution design. Build phase research is typically smaller scale (e.g. smaller sample) and lighter weight, intended for use by internal teams to ideate and make informed design decisions. Examples of types of research methods and approaches used in the Build phase include in-depth interviews, focus groups, and user testing with low-fidelity prototypes (shown in Table 1).

Even though Build phase research tends to be less formal and more exploratory, it is crucial for successful solution design. For example, a company may rush to get their product built, only to find at launch that it lacks a central feature users want. A quick, low-cost round of user testing with a low-fidelity prototype could have averted this issue. Low-fidelity prototypes, from mocked-up screenshots on paper to the earliest versions of a mobile app, are high-value, low-cost research tools to help startups save time and money by avoiding costly design missteps.

The PROVE phase is centered around validating the solution, meaning research activities are aimed at answering questions related to effectiveness and feasibility. In the PROVE stage, the first aim is to confirm, prior to the LAUNCH phase, that the solution in development is in fact the right solution for the target problem. Once there is evidence that this is the case, the next step is to investigate the question of whether that solution is built correctly- with the appropriate features and functionality that users need to be successful.

The PROVE phase is often rushed and incomplete, resulting in missed opportunities to target the precise solution that customers want and identify tweaks or even a full pivot to address the core problem. Research in this phase begins with smaller, lighter-weight studies to demonstrate initial evidence to indicate efficacy and feasibility. For example, high-fidelity prototyping and user testing provide early evidence of the feasibility of a solution in real-world contexts. Research later on in the PROVE phase typically shifts, as needed, toward larger-scale efficacy testing and potentially a randomized control trial, the most rigorous of approaches in scientific research. Research conducted early in the PROVE phase also provides invaluable data to inform later research design decisions from measurement to logistics. It is the chance to specify and test research measures and indicators in smaller-scale studies prior to pursuing heavier research undertakings or heading to the LAUNCH phase. This is the time to bring a statistician into the mix, along with some seasoned research guidance to help your team explore and test endpoints and measures and uncover conceptual and logistical issues that could throw your larger-scale study off track.

There are many complex decisions to make when designing a larger-scale research study: Should we randomize? Do we need to collect new data or can we use data already collected in the past? Are there strategies we can use to make the study simpler logistically such as decentralizing data collection (e.g. in the home or local site) or utilizing synthetic (algorithmically manufactured) data? Incorporating research expertise and learnings from early, smaller-scale research efforts means that by the time of a larger, more complex study the “kinks” have been worked out, making the design of the research protocol more robust.4 Additionally, these earlier experiences will identify opportunities to incorporate strategies for reducing time and effort and increase the likelihood of ending up with sound, useful results.

In the PROVE phase, we begin to generate evidence for external parties (e.g. investors, scientific community), however, research in this phase is also ripe with information for internal product development teams. Because of this, the PROVE phase may also involve revisiting BUILD phase activities as this phase offers the research tools to explore previously unidentified challenges or aspects of the healthcare problem we hope to solve.

Research in the LAUNCH phase revolves around answering questions related to implementation and market adoption for a range of audiences from the internal product launch team to potential customers and regulatory bodies. This happens through implementation-focused interviews with pilot users, key stakeholders, and structured observation of barriers and facilitators to effective use of the product “in the wild.” Slow uptake or adoption failures are often a result of unidentified implementation issues versus a flaw in the solution itself. These sometimes fatal implementation issues are often highly preventable, oftentimes with relatively simple fixes, such as tailoring the timing and format of information delivery for a clinical decision support solution or adjusting the locations where a user can access the tool based on a site’s specific workflows and routines. In this way, LAUNCH phase research leverages findings from prior phases to fine-tune solution design in early implementations to facilitate user adoption in larger-scale deployments, supporting wider market traction.

When approached thoughtfully, each new implementation can be treated as a research opportunity. Building research-related capabilities upfront into a solutions backend can exponentially increase a company’s capability to conduct ongoing research efficiently in a low-cost, low-budget way. The thought and effort put into establishing useful, workable research measures in the PROVE phase research pays off in the LAUNCH phase, providing the experience and insight needed to incorporate invaluable research infrastructure for the LAUNCH phase and beyond. Results provide information for internal teams to guide specific implementations, as well as gain an evergreen picture of the potential impact of a solution in the wider market.

The need to generate evidence, whether for regulators, investors or potential customers can be an expensive, burdensome hurdle for digital health startups.  When approached with a comprehensive, phased strategy, research can go from being an obstacle to a powerful and accessible tool.

With experienced guidance and a well-defined plan of action through the BUILD, PROVE, and LAUNCH phases, startups can successfully navigate their digital health research roadmap, saving time and resources while ensuring the development of effective and impactful solutions. A thoughtful research strategy can contribute to solving the evidence gap for new digital health solutions in a way that bolsters rather than stresses resource-constrained startups.