From Boardroom to Bedside in Digital Health
How to Overcome Key Pitfalls Faced by Startups for Providers and Patients
By Ricarina Rabinovitz (WG26)
A well-funded startup with an exceptional team developed a groundbreaking tool for individuals and patients with special needs. This solution promised to enhance communication, and completely transform their independence and overall quality of life.
Yet, despite its enormous potential, the solution was abandoned within weeks.
What happened?
The startup failed to integrate successfully into the provider’s system and participants’ lives, causing frustration and disengagement. Initial tests left patients disappointed. Even after the necessary improvements were made to the product, and despite the profound positive impact it could have had on their lives, users were unwilling to give it another chance. They were disillusioned by their initial experience.
Digital health startup companies hold immense potential to transform the lives of patients and their families. Yet, far too often, brilliant companies stumble over some all-too-common pitfalls. Today, we’re exploring key considerations and challenges faced by digital health startups serving providers and patients, and outlining some strategies that can help valuable solutions reach those who need them most.
Solve a problem, don’t tailor one to your solution
A common tendency for digital health startups is to start with a product concept and only then search for its practical applications. An example of this approach can be seen in the wearables market. There are many devices that track the same basic health metrics, yet each wearable may cater to a different health concern or use case.
It sometimes seems as though patients are expected to stack multiple smartwatches and patches on their bodies — one for each condition that may or may not ever affect them.
To mitigate this risk:
Before moving forward with a product idea, it is important to ensure that the product addresses a clear unmet need within a well-defined market. Next, founders should validate that the solution is well-suited to intended users’ lifestyle, preferences, and health needs.
Additionally, it is essential to consider regulatory compliance before getting started, a topic that will be explored further in the next section.
Finally, founders should gauge the end user’s reception through testing and feedback to confirm that the product resonates with and appeals to the target audience, and iteratively implement feedback. By carefully evaluating each of these areas, a concept has a far greater chance of evolving into a viable, impactful product.
2. Develop a product for the bedside, not for the boardroom
Some founders have a tendency to develop a product that’s “perfect” for business stakeholders, but less so for actual end users. Paradoxically, when partnering with a provider for clinical testing, it’s crucial to arrive at the testing phase with a product that is already somewhat refined for client use.
This refinement applies not only to integrations with provider systems, but also to user experience and the accuracy of the product’s outputs. Clinical care teams are already experiencing information overload, and redundant information and notifications only add to that strain.
With the first false alert from a patient-monitoring device, medical teams or caregivers will rush to the patient. By the second, their response may be a bit more hesitant. By the third, they will likely disconnect the device, regardless of its potential benefits. The same goes for any errors or unnecessarily repetitive alerts that a product or software tool may send.
To mitigate this risk:
Founders should develop products with constant feedback from end- users, making user needs a top priority in the development process. Engaging with users – without pride in authorship – may reveal insights that founders might miss due to attachment to their product.
Even in early testing, user experience is critical. If staff, patients, or caregivers lose interest in a product, no amount of management persuasion will bring them back to using it. After a negative first impression, a provider may stop working with an organization even if you implement feedback and improve the product.
3. AI should not be seen as a “black box”, but as a tool at your disposal
Understanding how an AI solution works, effectively communicating its value, and addressing potential challenges are crucial for founders. This builds trust with stakeholders and strengthens their confidence in the solution's viability.
For example, one challenge is the complexity of patient physiology, since critical metrics may fluctuate due to internal and external factors, such as time of day, temperature, and social interaction. Addressing the signal-to-noise in training data and understanding how this may affect the model’s effectiveness, based on existing research and clinical insights, will help to strengthen your model’s credibility and reliability.
Further, training AI models requires accurate data labeling and validation, along with substantial computing power. It’s important for organizations to pay mind to effective resource allocation to avoid unnecessary and ineffective resource use on AI.
To mitigate this risk:
Founders should learn how to explain relevant models and algorithms in simple terms, and research how relevant technology might align and integrate with variability in patient physiology. Founders should also make sure to handle validation of any model outputs on their side, given that providers, caregivers, and patients are often already carrying significant mental loads. However, this process is resource-intensive and demands dedicated personnel to ensure that valuable insights are gathered while minimizing unnecessary noise – which might be challenging for startups to manage.
Ensure integration with provider systems and compliance with patient privacy regulations.
A major frequently-overlooked challenge in digital health is the process of integrating new digital systems into providers’ existing software infrastructure.
A few reasons why integration may be challenging:
Hospital and clinic software infrastructures vary widely and may be outdated.
Data privacy protection systems may prevent full integration.
Connecting via Bluetooth or internet-based solutions may be an issue, especially when the product monitors critical parameters. No provider would risk data leaks, or systems interruptions such as power outages and communication issues, that could prevent critical health information from reaching the care team.
These factors often get in the way of adoption of digital health solutions by providers, or lead to repeated bugs and crashes, ultimately impacting retention.
To mitigate this risk:
Startups must prioritize integration from an early stage. To get a better understanding of existing systems and their hurdles, initiate early discussions around implementation with provider IT teams. Thoroughly consider the necessary investment in systems integration throughout the development process.
5. Data privacy and security is paramount
Overlooking data privacy and security can lead to costly and complex issues down the road, especially given the strict regulations of HIPAA (in the U.S.) and GDPR (in the EU). These regulations have strict requirements for the handling and security of patient data and carry severe penalties for non-compliance.
To mitigate this risk:
As an example, consider a patient-support platform, such as a mobile app or tablet-based tool, that is not integrated with the provider’s system but requires patient data. Early planning can simplify and streamline patient data safety and risk reduction measures, saving both time and resources. Implementing these measures at the outset is much easier and more cost-effective than trying to incorporate them after the system is already established.
Some examples of such measures are:
Using coded identifiers instead of storing identifiable personal information.
Storing patient images as secure links on trusted external HIPAA-compliant storage systems. Local storage of patient images should be limited and only done for the time needed to process the images.
Leveraging platforms such as Google for Startups to benefit from built-in cost-saving features, such as Mobile Device Management (MDM), to support compliance through enhanced device and data security.
Founders should be sure to tailor these considerations to the particular product and its distinct regulatory environment.
It is essential to work closely with qualified healthcare data security experts and engage with the target providers' data teams early on, since providers have varying compliance and storage standards. Consulting experts and comparing data security providers' features can help ensure your platform meets HIPAA and GDPR requirements in a cost-effective manner.
Conclusion
Pitfalls often arise when companies fail to engage closely enough and early enough with customers and end users, whether they are hospitals, clinics, medical teams, caregivers, or patients. Engaging early and maintaining user centricity throughout the product development process can go a long way to saving time and heartache down the road, and will help founders develop excellent products that meaningfully benefit patients.