Part 2: The 3P's of AI in Healthcare

Preventative medicine

Image generated using Open AI's DALL·E 3.

Breaking down the shift towards prevention

Imagine a world where diseases can be prevented before they occur or mature. Through advancements in technology over the last century, we have been witnessing this evolving reality. With developments in vaccine technology, deadly diseases such as Polio have been eradicated in almost all countries. We've also seen the field of medical imaging enable the non-invasive early detection of disease since the discovery of the X-ray. The list of these technologies, as well as the number of lives they have already saved, is countless. 

We don’t, however, live in a world of perfect prevention — there are still diseases that we don't understand, illnesses we can't detect early enough, and healthier lifestyles that we struggle to lead.

Countries such as the US are spending up to 80% of healthcare costs on 'treatment and care'. This spending occurs after the patient has become ill, and dramatic savings could occur if we could prevent these illnesses in the first place. The Lancet Public Health estimated the cost of preventable illnesses on the US healthcare system as $730 billion. For example, an estimated 50% of cancer deaths in men are classified as preventable due to lifestyle factors such as smoking, alcohol consumption and obesity.

In light of this, a shift in spending from treatment to preventative care is happening. With this, a wellbeing market is emerging and offering solutions for prevention of diseases. The US spent $700 billion on wellbeing in 2019, and this is set to increase to account for 67% of the total health spending by 2040. This translates to a massive $5.1 trillion, with spending on treatments making up the remaining 33% — almost the total opposite of today's ratios. Globally, governments are indicating their intent to secure the future of this market. For instance, the UK government has placed prevention at the heart of the NHS's Long Term Plan and recently announced the plan to phase out smoking. 

So how is AI advancing preventative medicine and wellbeing? And where are some of the main opportunities?

Diagnostics. Early diagnosis and treatment of disease is often key to preventing it from maturing. In current diagnostics, an examination is conducted and the results are interpreted by clinical staff. AI can help in the analysis of examination results by setting up the problem as a classification task, and has successfully applied to classify disease from images, electrical signals, and even electronic health records. In this way, AI can help make diagnostics more accurate; and save cost by prevention. In the diagnostic space we have start-ups using AI for the detection of heart disease (InVision), stroke (Methinks), skin cancer (Skin Analytics), rare diseases (Mendelian), and even for the early detection of pregnancy complications (Elythea). Whilst other start-ups in this space empower the consumer directly with at-home cancer testing (Cleancard) and general health assessments via whole body scans (Neko) and other indicators (HealthCaters).

Remote Monitoring. Patient monitoring typically involves the use of wearable devices, sensors, mobile apps, and other digital health tools, to track changes in a patient's health condition over time. Monitoring can help clinicians manage patient's outside of the clinic, help make timely interventions, and help to prevent illnesses from maturing. In this way, monitoring can reduce the need for in-person appointments and hospital visits - reducing the burden on healthcare systems. The remote monitoring space was valued at $55 billion in 2022 and is forecast to grow to $506 billion by 2032. AI is a key technology for monitoring, as it enables the analysis of a patient's condition over time. For instance, AI is being used to flag irregularities in data-streams that might indicate health issues, and to alert healthcare providers in real-time to potential problems. Some of the start-ups in this space are building virtual wards for patients for at-home care (Doccla), monitoring cardiac health of patients (Implicitly, Noah Labs, Acorai), monitoring audio (Sensi.AI), managing chronic conditions (HealthSnap), and monitoring Parkinson’s disease severity (Serg Tech).

Wellness. The use of wellness in healthcare involves a focus on preventive measures and lifestyle interventions that can promote health and prevent disease, rather than just treating illness after it has occurred. This space, distinct from diagnostics and remote monitoring, is broader in scope, more accessible to consumers, and aligned with growing trends in preventive healthcare. Some interesting start-ups in this space include monitoring mental health from audio data (Kintsugi), mental health coaching (Plutis), personalised wellbeing companions (VOS, Sensely), nutrition guidance and tracking (Fastic), and a platform for health monitoring and advice that pairs with existing wearable devices (ONVY).

What are some of the key challenges?

Product strategy. Start-ups in the preventative medicine space often have to choose between building integrated hardware and software products, or focusing solely on software-based solutions. For instance, Neko has developed a unique multi-modal imaging device alongside the software for its analysis. Whilst other companies in the space, such as ONVY, focus entirely on a software product offering that pairs with existing hardware. Both strategies have their unique value propositions and market challenges. For companies in the former approach, the key is to demonstrate the superior capabilities of their integrated system, capitalizing on unique devices and datasets to create a more defensible market position. This integrated approach, however, also brings more complex regulatory compliance challenges, as it must meet stringent standards for both hardware and software as medical devices. With the latter category of companies, the focus might be on showcasing the versatility of their software, leveraging lower costs, quicker market entry despite a more competitive software landscape, and potentially simpler regulatory landscape due to its reliance on existing, already-approved hardware. This strategic decision is critical, as it directly influences the company's development, market position, and long-term success in the evolving preventative medicine space.

Adoption. The structure of healthcare organizations inherently limits their ability to rapidly adopt new technologies, primarily due to their segmentation into multiple departments, each with distinct processes, protocols, and systems. This segmentation leads to siloed decision-making and difficulty integrating products across the organization. In this hierarchical structure, top-down pressure is essential for effective adoption. Leaders at the highest levels need to champion adoption, ensuring that the initiative is prioritized across all departments. But convincing leaders to become champions is challenging, unless the situation is urgent. For example, the urgency of the COVID-19 pandemic catalyzed the adoption of AI for remote monitoring, diagnostics and assessment. We saw Huma quickly partner with the NHS to identify patients deteriorating with COVID remotely, and Doccla rapidly integrate their virtual ward products into new hospitals. The urgent need for effective pandemic management and the continuation of healthcare services made the risk of not adopting new solutions far greater than the risks associated with their rapid implementation. With the threat of the pandemic subsiding and urgency diminishing, adoption is again becoming a key challenge. To alleviate this, finding and partnering with internal champions who can apply top-down pressure is vital.

The integration of AI into preventive medicine marks a significant shift towards proactive healthcare. By enhancing diagnostics, remote monitoring, and wellness, AI is reshaping how we approach disease prevention and health management. Although challenges such as strategic product development, navigating complex regulatory landscapes, and adoption remain, AI is set to be a key technology in this space. Stay tuned for our next article where we dive into AI for personalized medicine; the third and final of our 3P's.

At Hummingbird, we’ve had the privilege to back entrepreneurs across four continents building in health such as BillionToOne, Enveda Biosciences, Automata, Basecamp Research, Pristyn Care, Eden and Anima Health, among others. We’re incredibly excited about the next decade of change in healthcare and lifesciences, so if you’re building in this space please get in touch!