The Year of AI: Solving Dichotomies in Healthcare

AI this, AI that—AI is in everything!

And healthcare is no exception.

As we covered in our previous post, AI is being applied to administrative tasks, but we see limitations to that approach. At The Healthcare Syndicate, we believe that for AI to drive large-scale transformational change, it must focus on improving patient outcomes.

There are many persistent and dichotomous challenges that prevent us from achieving a healthier future, and we believe AI has the potential to bridge those gaps. This article provides a high-level overview of key areas where AI can play a pivotal role. In future posts, we will explore each of these concepts in greater detail.

Healthy care vs sick care:

In healthcare, "sick care" refers to the traditional, reactive approach that focuses on diagnosing and treating illnesses as they occur. In contrast, "healthy care" (or "well-care") emphasizes proactive measures aimed at maintaining wellness and preventing diseases before they manifest. The sick care model dominates today’s system, but as healthcare shifts toward value-based care, a hybrid model that integrates both sick and healthy care is essential.

Why? Chronic disease management is inherently longitudinal—it’s not just about responding to acute flare-ups but also maintaining a steady baseline of health to prevent those flare-ups in the first place. This creates a new and growing demand for healthy care solutions, which aim to optimize health and minimize the frequency and severity of disease-related episodes. The future of healthcare will require seamless transitions between these two modes of care, ensuring that patients not only recover from acute events but also stay as healthy as possible between them.

Example: Asthma and the Shift Between Sick and Healthy Care

Consider asthma, a chronic respiratory condition where the treatment approach shifts dramatically between sick care and healthy care:

• Sick Care Approach (During an Attack)

  • Patients use rescue inhalers (e.g., albuterol) to quickly open airways.

  • They are advised to avoid physical exertion and stay in controlled environments to minimize triggers.

  • In severe cases, steroid medications or emergency medical care may be needed.

• Healthy Care Approach (Prevention and Maintenance)

  • Outside of attacks, controlled exercise and lung-strengthening activities (e.g., swimming, breathing exercises) can improve respiratory function.

  • Patients follow daily maintenance therapy, such as inhaled corticosteroids, to reduce inflammation and prevent flare-ups.

  • Identifying and managing environmental triggers (e.g., allergens, pollutants) helps maintain long-term stability.

Asthma management illustrates a striking contrast: during an attack, patients must avoid physical exertion, whereas long-term maintenance requires consistent physical activity to strengthen lung function and reduce flare-ups.

Other examples of this dynamic include:

• MSK Disease (e.g., Osteoarthritis, Chronic Knee Pain): During flare-ups, patients are advised to reduce activity and focus on pain management, but once stabilized, they must increase physical activity to strengthen muscles and prevent further deterioration.

• Type 2 Diabetes: During hypoglycemic episodes, quick sugar intake is necessary, but outside of emergencies, the focus shifts to maintaining stable blood sugar levels through diet and exercise.

The challenge in these conditions is ensuring patients transition seamlessly between sick care and healthy care. AI can help in two ways:

Closing the Continuum of Care Loop: Helping Patients with Care Navigation

For patients with chronic conditions, knowing when and how to transition between different care states is often confusing. AI-powered virtual care navigators can:

• Provide personalized recommendations on when to rest vs. when to resume activity, leveraging AI-powered decision support systems that can modify treatment plans in real-time based on the patient’s condition.

• Coordinate care between specialists, ensuring seamless transitions between treatment phases.

• Offer adaptive health coaching, modifying lifestyle interventions based on the patient’s real-time health status.

Has the Patient Returned to Their Healthy Baseline?

One of the biggest challenges in integrating sick and healthy care is determining when a patient is truly “back to normal”. AI can assist by:

• Establishing Personalized Health Baselines: Using long-term health tracking data, AI can create dynamic baselines for each individual.

• Monitoring Recovery Trends and Predicting Risk Flare-Ups: AI can assess whether key biomarkers, movement patterns, or glucose levels are returning to pre-flare-up levels. Machine learning models can analyze patient history and real-time biomarkers to anticipate when a patient is about to transition from stable to acute.

• Predicting Risk Flare-Ups and Detecting Early Warning Signs: Machine learning models can analyze patient history and real-time biomarkers to anticipate when a patient is about to transition from stable to acute. AI can also flag deviations from normal health patterns, prompting early interventions before the next flare-up occurs.

By integrating AI into healthcare, we can seamlessly transition patients between sick care and healthy care, ensuring they receive the right interventions at the right time. This shift not only improves outcomes but also aligns with the broader movement toward preventive, value-based healthcare.

Trust and Profit: Bridging the Divide in Healthcare

In healthcare, trust and profit often exist in tension because key stakeholders—patients, providers, and payers—can have adversarial interests. Specifically, payors and providers are often at odds with each other, and patients end up paying the price. The result is a system where inefficiencies erode trust, and financial incentives don’t always align with patient outcomes.

The Administrative Burden Problem: A Summary

AI has been positioned as a solution to healthcare’s excessive administrative workload, but in reality, it will likely not have the desired effect in aggregate. The most common use of AI is in automating manual and repetitive tasks, but this does not inherently make trust easier to establish. Instead, automation often leads to a cycle of escalating barriers as different stakeholders attempt to maintain control. Just as Captchas have evolved to counteract automation, healthcare payers and providers respond to automation by introducing new manual steps designed to be resistant to AI-driven processes. Rather than facilitating trust, automation in administrative workflows creates an adversarial dynamic where each side seeks to counteract the other’s efficiencies. The result is a system where little real knowledge is exchanged, and the difficulty of building trust only increases in an age of AI.

The Limits of Automation in Establishing Trust

Unlike blockchain and cryptocurrency systems, where proof of work (PoW) ensures transparent and decentralized transaction verification, healthcare operates within a highly regulated and opaque framework where trust is derived from human oversight, not computational effort. In Web3, PoW builds trust by making fraudulent activity prohibitively expensive, ensuring that transactions are verifiable and immutable. However, healthcare's challenges stem not from a lack of computational verification, but from misaligned incentives and competing interests between stakeholders.

Rethinking Trust in Healthcare

Given that automation alone cannot establish trust, healthcare needs mechanisms that reinforce accountability rather than just efficiency. Potential strategies include:

• Helping Clinicians Align Financial and Clinical Outcomes with Patients: While payor-provider conflicts may seem like a bleak situation, we believe AI has the opportunity to bring healthcare back to its core by enabling clinician-patient alignment without being bogged down by infrastructure. From outcomes tracking to decision support, AI-driven systems can help clinicians focus on patient care while ensuring financial incentives are tied directly to patient health rather than administrative processes.

• Facilitating Value-Based Care with Performance Tracking: Value-based care is the future, but achieving it is extremely challenging. Every aspect, from contracting to performance tracking and payout, must be executed flawlessly—any gap in implementation risks failure. A core challenge is that healthcare outcomes depend on both quantitative factors (such as biomarkers like blood sugar levels) and qualitative factors (such as patient-reported well-being), which traditional rule-based methods struggle to incorporate effectively into contract. AI has the potential to bridge this gap by integrating and interpreteing both types of data, enabling a more holistic approach to performance tracking and ensuring that value-based care models function as intended.

By recognizing the limits of AI-driven automation in trust-building and focusing instead on clinician-patient alignment and performance-based value care, healthcare can bridge the trust gap without creating an endless cycle of counterproductive automation.

Personalized vs. Population Health

Clinical practice has long been guided by "gold standard" guidelines—expert-driven consensus on clinical evidence designed to be broadly applicable to populations. However, individual health conditions are highly variable, and edge cases abound. How else can we explain the lifelong smoker who never develops lung cancer or the young, non-smoker diagnosed with stage 4 lung cancer?

A landmark study in Nature analyzing over 20 years of blood test data found that individuals maintain highly specific and stable "setpoints" for various biomarkers that remain consistent over decades. Traditionally, blood tests are evaluated against a "normal range" derived from population data, but this research suggests that comparing lab results to an individual’s historical baseline rather than broad population averages could improve diagnostic accuracy and treatment personalization.

Bridging Personalized and Population-Based Treatment

To make personalized medicine actionable while maintaining the gold standard of evidence-based treatment, we need AI to:

• Aggregate Patient Baseline Data: Collect longitudinal health data to establish personalized baselines for each patient. The proliferation of wearables, health apps, and digital medical devices will generate the data needed to create these personalized health profiles, allowing for a more precise and individualized approach to treatment. AI can analyze these vast data sets to find hidden connections, identifying patterns that might not be immediately apparent and enabling a deeper understanding of individual health trajectories.

• Develop and Popularize N-of-1 Trials: In order to truly leverage individual baseline data, we need to understand how they deviate from the baseline in different situations. Digital apps and wearables will be critical in capturing these data across different times and settings. AI can formulate N-of-1 experiments from these data in a non-intrusive manner, ensuring that patient behavior is not explicitly structured around artificial test conditions. Instead, AI can passively observe natural variations, identifying meaningful deviations and responses to treatment. This approach makes personalization scalable while remaining representative of real-world patient behavior.

• Relate Individual Data to Population Data: Use AI models to compare an individual’s trends with broader epidemiological data to personalize treatment plans. While personalized data provides crucial insights into individual health, epidemiological data remains essential as a reference point. It helps establish population-wide trends, identify risk factors, and validate personalized interventions against established evidence. AI can bridge the gap by contextualizing individual deviations within larger health trends, ensuring that personalized treatments remain evidence-based and broadly applicable.

Despite being listed last, personalized healthcare data is actually the first step in achieving the full potential of AI in healthcare. By integrating these AI-driven approaches, healthcare can close the loop back to healthy care vs. sick care, ensuring that transitions of care require tighter integration between clinicians and patients. AI has the potential to facilitate the shift from reactive population-based medicine to proactive, individualized healthcare that optimizes both preventive and acute care interventions. Furthermore, personalized data enhances trust in the healthcare system, as discussed in the previous section on trust. When patients see that their care is based on their unique health patterns rather than generic population metrics, they are more likely to trust and engage with the healthcare system, ultimately leading to better adherence and outcomes.

What does this mean for investors?

Investors looking to support transformative AI solutions in healthcare should prioritize innovations that enhance patient outcomes rather than just drive administrative efficiency. The most impactful AI applications will be those that can be directly measured against improvements in patient health, ensuring tangible benefits beyond cost savings. Even in administrative applications, AI should be developed to facilitate better patient outcomes, reinforcing healthcare's core mission rather than simply optimizing financial efficiencies.

At a high level, The Healthcare Syndicate sees these insights as our evolving thesis on AI in healthcare:

• Investors should focus on companies leveraging AI to bridge the gap between sick care and healthy care.

• AI-driven solutions that improve care navigation and personalize treatment based on individual baselines will be critical.

• The next wave of healthcare innovation will likely be driven by AI’s ability to integrate personalized and population-based health data.

• AI-enabled healthcare models should aim to be more effective, trusted, and scalable.

Companies and investors that align their AI strategies with these principles will not only achieve long-term adoption and regulatory support but will also help shape the future of value-based, patient-centric healthcare.

Please subscribe to our newsletter if you haven’t, and share our newsletter with a friend. Stay tuned to our newsletter for more insights into healthcare innovation!

Join us at The Healthcare Syndicate as we back the most ambitious founders 10Xing the standard of healthcare!