Pharmaceutical companies interested in joining the global digital health revolution are facing difficult decisions on how to use the innovative technology. Is it better to provide a low-risk digital device, such as a diagnostic tool to improve the patient’s experience or to make long-term investments in digital technology, such as developing diagnostic tools? While the answer depends largely on patient needs and business model, here are some key factors to success in the digital testing space.

1. What is at stake?

Software functions1 Devices that complement the definition of the device or software such as SaMD can be deployed on mobile platforms, other general purpose computing platforms, or hardware device functionality or control. Digital technologies that fall under the Food and Drug Administration (FDA) regulation include:2

  • Mobile applications that use built-in features such as light, vibration, camera, or other similar sources (such as mobile medical applications used to diagnose or treat disease).
  • Software functions that control the operation or function of an implanted medical device (e.g., change settings).
  • Software functions used in active patient monitoring to analyze patient-based medical device data.

The FDA is risk-averse, so companies often believe that the best bet is to build a Part 1 or Part III digital device to avoid going through Part II or Part III approval and approval processes. However, going Part 1 means launching a digital device with limited functionality to address medical purposes or to address critical health care needs. Also, consumers can already choose from hundreds of thousands of health-related apps, so the chances of identification are limited. In contrast, investing in certified clinical services for high-risk digital diagnostics can provide clinicians with a more effective tool and a way for pharmaceutical companies to generate revenue.

2. What is your use case?

Basically, digital involves following one of the two main revenue models:

  1. To mobilize indirect income for complementary medicine
  2. To create an independent direct revenue stream

Many companies take the first option, but often unknowingly make both mistakes. This can happen when there is a lack of transparency in the business or there is a material change in the associated therapeutic (FDA approval, clinical information, etc.). In such cases, many companies have allowed the digital product to move forward without re-evaluating the investment strategy and business model. Establishing an independent revenue stream for digital production requires a unique investment study, risk level and time frame. In this case, the investment and risk ratio may indicate the need for a different strategy – a partnership or idea to put the product out of license.

When building a digital product, pharmaceutical companies should consider direct and indirect revenue streams at the beginning of the product development process, especially when entering new markets or seeking an FDA license for a friend treatment. While regulatory clearance is more effective than traditional medicine, digital products may face post-market adoption barriers, as well as trade and return risks. Switching from an indirect to a direct revenue model can be challenging and a complete restart on product design and characteristics, control strategy and clinical planning may be required. By mapping both direct and indirect revenue streams, companies may be better able to evaluate product development transactions and make strategic decisions with a clear end goal.

To fully understand your best options, conduct a strategic analysis of each route, taking into account the market opportunities, position, assets and risk tolerance for your company. Think about the market demand and what kind of digital diagnostics will best meet that demand, and if there is a way you can make a big profit on that investment, for example by tracking a paid diagnostic tool. Also, keep in mind that the results of strategic analysis may indicate that you are not continuing. Time, cost of opportunity, investment (clinical and other) and risk – if there is no income guarantee – can be too much.

3. Can you simplify the work of clinics?

Some companies want to use artificial intelligence (AI) technologies to automate clinical labor. As physicians are usually the last customers, they are more likely to automate their work, and may introduce additional adoption barriers, business barriers, and legal issues. In contrast, digital healthcare, which increases clinical decision-making, brings improved services and outcomes, increases (and increases the chances of recovery). When designing your digital diagnostic method, consider how the product enhances physician practice and / or improves clinical outcomes.

4. Is your growth process FDA-ready?

Traditionally, software developers are accustomed to working efficiently with development sprints, knowing that at least MVP can be marketed as quickly as possible with the intention of making and adjusting, correcting errors and implementing improvements in future versions as needed. When developers typically verify and validate these products, the required strength is usually much lower than for high-risk Class II or III medical products.

Making your way through the FDA’s strict regulatory approval process requires a deep understanding of compliance and personal patience. This means following regulatory design controls, keeping detailed documentation on the growth path, as well as robust testing, evaluation, and verification processes throughout the life cycle of the product. And, although products may be updated and modified after approval, the first FDA-approved edition must be certified to meet its clinical purpose and at a unique level.

In addition, while all ML model developers are performing some model verification, there is a very high level of commitment to medical products. If the confirmation is done incorrectly, the results of the clinical trial show that it does not work reliably in the real world setting, and they return to research and development. Technically, this means using the best practices, such as nest cross-certification, that are appropriate in terms of sample size, patient and disease characteristics. It also means avoiding “target leakage” during the verification process, which means that hidden verification errors that make the model appear effective in the development process will fail when it is distributed in the real world. This is crucial for the results of the clinical trial to be presented as expected.

Setting yourself up for digital health success

All of this comes from aligning your digital healthcare strategy with your business and marketing opportunities. For many companies, this means creating more and more systematic digital diagnostics for the medical property. But without doing your best to ensure your best option, you may miss out on digital health innovation opportunities, or at least you can secure an option through the product development cycle. It is important to consider and understand the needs of the target market and the therapeutic landscape in order to promote valuable digital health technology. This includes understanding the doctor’s perspective, incentives, and barriers to adoption. It also means building a truly cost-effective digital diagnosis for patients and clinics that rely on you for treatment options – even with a more rigorous development process and licensing.

Bill Weed is the Associate Director in Health; Jim Williams is an associate director in life sciences; And Jacob Graham it is. Partner in the practice of life sciences, all by Guide.