Precision Healthcare and Life Sciences: What Held Up and What Boards Should Learn

In December 2022, I listed ten trends that would shape precision healthcare and life sciences. Some have arrived, some have not, and the difference holds a lesson for any board approving an AI strategy today.

In Brief

The Problem. Boards are asked to approve strategies built on technology trends, and the forecasts that support those trends are often vendor market projections that do not survive contact with patients, regulators, or payers.

Why It Persists. Market-size projections are easy to cite and hard to test, while the evidence that matters (controlled trials, regulatory acceptance, and validation in the organization's own setting) arrives later and less dramatically.

What Boards Should Do. Judge each trend, and each AI proposal, by the evidence path behind it rather than by the size of the market, and ask management to show which initiatives rest on proof and which rest on promotion.

A Field That Moved Faster Than Its Forecasts

Precision healthcare rests on a simple idea: treat each patient according to his or her own biology, history, and circumstances. Advances in genomics, artificial intelligence (AI), digital health, and the shift toward value-based care each pushed the idea forward. In 2022, I argued that companies that built these trends into strategic planning would be positioned to compete. That argument was right in direction and wrong in method. The original version of this article measured each trend by a market forecast. Almost all of those forecasts came from 2021 vendor reports, and I have removed them.

The useful exercise for a director is not to forecast market size. It is to ask which advances were proven in practice, which were promoted ahead of the evidence, and what separated the two. Reviewing the ten trends four years later, the pattern is consistent. The advances that lasted traveled a visible evidence path: controlled studies, regulatory review, payment, and follow-up. The ones that faded relied on projections. That is the test I now apply to AI proposals.

Genomics

The genomic basis of disease is driving targeted therapies and diagnostics, and the falling cost of sequencing made that possible. The National Human Genome Research Institute (NHGRI) reports that the cost of a high-quality draft human genome sequence had dropped to about USD 14 million by 2006, and that a whole-exome sequence generally cost below USD 1,000 by the time of its 2021 update (NHGRI, 2021).

Gene editing moved from the laboratory to approved therapy within a decade. On December 8, 2023, the Food and Drug Administration (FDA) approved two gene therapies for sickle cell disease in patients 12 and older. One of them, Casgevy, was the first approved treatment to use CRISPR gene editing (Berkowitz & Simkin, 2023). The path from discovery to approval ran through controlled trials, regulatory review, and long-term follow-up. That path is the standard against which other claims should be measured.

AI and Machine Learning

AI and machine learning (ML) are changing drug discovery, diagnostics, and patient care. Spending has followed. Healthcare AI spending reached about USD 1.4 billion in 2025, nearly triple the prior year, and healthcare organizations were deploying AI at 2.2 times the rate of the broader economy (Yap et al., 2025).

The two examples in my original list have since moved in opposite directions. DeepMind's AlphaFold predicts protein structures with accuracy that its developers reported as competitive with experimental structures in a majority of cases in a blind assessment (Jumper et al., 2021). On October 9, 2024, the Nobel Prize in Chemistry went half to Demis Hassabis and John Jumper for protein structure prediction and half to David Baker for computational protein design (Al Jazeera, 2024).

IBM's Watson for Oncology, which I listed in 2022 as a tool that recommended personalized treatment plans, should not have been on the list as a success. In July 2018, STAT reported that internal IBM documents showed it had been trained on synthetic cases rather than real patient data, and had recommended unsafe and incorrect treatments (Ross & Swetlitz, 2018). In January 2022, IBM announced the sale of Watson Health's data and analytics assets to the private equity firm Francisco Partners (Ridley, 2022). The lesson is not that clinical AI fails. It is that a system which shapes care must be validated in the setting where it will be used, and that marketing claims are not evidence.

Digital Health

Telemedicine, remote patient monitoring, and mobile health give patients more control over their health and give providers more ways to deliver care. Investment remains strong but selective. United States digital health ventures raised USD 7.4 billion across 244 deals in the first half of 2026, and Rock Health no longer treats AI as a distinguishing feature because it has become common (Olsen, 2026).

My two 2022 examples both changed hands or changed shape. Google completed its USD 2.1 billion acquisition of Fitbit on January 14, 2021, and committed not to use Fitbit users' health and wellness data for advertising (Drees, 2021). Teladoc reported a 2022 net loss of USD 13.7 billion, driven by a USD 13.4 billion goodwill impairment related to its acquisition of Livongo (Pifer Parduhn, 2023). Telehealth itself has endured: Congress extended the Medicare telehealth flexibilities through December 31, 2027 (American Academy of Sleep Medicine, 2026).

Digital health has also found a large market in documentation. Ambient clinical documentation drafts clinical notes from the conversation between patient and clinician, and it accounted for about USD 600 million of healthcare AI spending in 2025 (Yap et al., 2025). These tools change the medical record itself. Their accuracy, and the clinician's review of each note, are matters for oversight. In a blinded study at the University of California, San Francisco, 22 attending physicians found that AI-drafted discharge narratives contained more omissions and inaccuracies than physician-written ones, although overall harm scores were low; the authors concluded that AI drafts reviewed by physicians may be a viable option (Williams et al., 2025).

Value-Based Care

The shift toward value-based care rewards delivery models that improve outcomes and reduce costs, and it raises the importance of predictive analytics and population health management. The shift is real but partial. In 2023, 28.5 percent of United States health care payments flowed through alternative payment models with downside risk or population-based payment (Health Care Payment Learning and Action Network, 2025).

My 2022 example needs a correction. I wrote that Geisinger's MyCode Community Health Initiative provides personalized care plans. The study I cited reports something narrower and still important: in exome sequences from 50,726 participants in the DiscovEHR collaboration, about 3.5 percent of individuals carried deleterious variants in 76 clinically actionable genes (Dewey et al., 2016). That finding is what makes population genomic screening a value-based care question: a health system can identify risk years before disease.

Real-World Data and Real-World Evidence

Real-world data (RWD) and real-world evidence (RWE) inform clinical decisions, support the development of new therapies, and support the adoption of clinical decision support (CDS) tools and software as a medical device (SaMD). FDA leaders set out what RWE is and what it can show in 2016 (Sherman et al., 2016). In December 2018, the FDA published its framework for evaluating whether such data can support effectiveness decisions, under authority added by the 21st Century Cures Act (FDA, 2018). Flatiron Health, which builds RWD from electronic health records for cancer research, was acquired by Roche in 2018 (Roche, 2018).

AI is now increasingly used to generate that evidence. In January 2025, the FDA issued draft guidance proposing a risk-based framework for establishing the credibility of an AI model for a particular context of use in drug and biological product decisions (FDA, 2025). As of October 21, 2026, that guidance remains in draft. A board reviewing an in-licensed asset should ask who has examined the AI-generated evidence behind it.

Personalized Medicine

Advances in genomics, proteomics, and metabolomics identify biomarkers and drug targets that make therapies and diagnostics specific to the individual. Foundation Medicine's comprehensive genomic profiling test remains the clearest commercial example. In November 2017, the FDA approved FoundationOne CDx, which detects alterations in 324 genes and serves as a companion diagnostic across five tumor types, while the Centers for Medicare and Medicaid Services (CMS) proposed national coverage through a parallel review (Harris, 2017). Approval and payment arrived together, which is why the test reached practice.

Collaborative Research and Strategic Partnerships

Collaboration with academic institutions, biotechnology companies, providers, and technology companies shares resources, expertise, and data, and it helps a company earn a return on its research and development (R&D) investment. Capital for the sector recovered in 2024: overall biopharma funding reached USD 102 billion, up from USD 71 billion in 2023 (IQVIA Institute, 2025). The partnership I cited in 2022 still illustrates the model. On October 1, 2019, Novartis and Microsoft announced a multiyear collaboration in which Novartis founded an AI innovation lab and chose Microsoft as its strategic AI and data science partner across discovery, development, and commercialization (Microsoft, 2019).

Regulatory Innovation

Engaging with regulators early, following current guidance, and addressing challenges before they become objections remain essential. Expedited pathways are now routine. The FDA's Center for Drug Evaluation and Research approved 46 novel drugs in 2025, and about half used expedited programs such as fast track, breakthrough designation, or priority review (Eglovitch, 2026). Regulators have also become more receptive to RWE. Aetion, whose evidence platform I cited in 2022, provides RWE to biopharmaceutical companies, payers, and regulators for uses that include regulatory submissions, and Datavant agreed to acquire it in May 2025 (Datavant, 2025).

Microbiome Research

The human microbiome, the bacteria, fungi, and viruses that live in and on the body, plays roles in digestion and immune function. In 2022, I cited a market forecast for the field. The more useful milestone is regulatory. The FDA approved Rebyota in November 2022 and Vowst, the first orally administered fecal microbiota product, on April 26, 2023, both for recurrent Clostridioides difficile infection (FitzGerald & Young, 2023). Partnerships among pharmaceutical companies, biotechnology firms, and academic institutions continue to drive the field.

Blockchain

In 2022, I wrote that blockchain could improve data security, traceability, and interoperability in healthcare and clinical trials. In my view, four years later, it has not become a board-level strategy for health data, and the market forecast I cited no longer deserves a place in this article. The interoperability problems it was meant to solve are being addressed through other means. The national Trusted Exchange Framework and Common Agreement (TEFCA) went live on December 12, 2023, and about 464 million documents had been exchanged through it by the end of 2025 (Assistant Secretary for Technology Policy, 2026).

What Held Up and What Did Not

Trend	Where it stands in October 2026	Board question
Genomics	CRISPR therapy approved (December 2023)	Which genomic programs have a payment path?
AI in discovery	Recognized by the 2024 Nobel Prize in Chemistry	Where does AI shorten our research timelines, and how do we know?
AI in clinical decisions	Watson for Oncology failed; validation now expected	Who validated each clinical AI system in our setting?
Digital health	Investment selective; ambient documentation about USD 600 million of 2025 AI spending	Who reviews AI-drafted notes?
Value-based care	28.5 percent of payments in downside-risk or population models (2023)	Which AI investments depend on risk-based contracts?
RWD and RWE	FDA framework (2018); AI guidance still draft	How was AI-generated evidence diligenced?
Personalized medicine	Approval paired with coverage	Does each program have both?
Partnerships	Funding recovered in 2024	What do partners get to do with our data?
Regulatory innovation	About half of 2025 novel approvals expedited	Which submissions rely on AI-generated analyses?
Microbiome	First approvals in 2022 and 2023	What evidence supports any further expansion?
Blockchain	Not a board-level strategy	None needed

The Strongest Objection

The strongest objection is that hindsight is easy. Every trend list looks uneven four years later, and a board that waited for regulatory approval and payment before investing would always be late. That is fair, and early investment is often the right decision. My point is narrower. The board should know which kind of bet it is making. An early investment in an unproven capability can be sound if it is sized as an experiment, staged, and reviewed against evidence. It becomes a governance failure when a projection is presented as proof and the board approves it on that basis.

For the Board

• Which of our precision medicine or AI initiatives rest on evidence from our own setting, and which rest on vendor or market claims?
• For any AI that shapes clinical decisions, who validated it locally, and how is its performance monitored after go-live?
• For AI-generated evidence in a regulatory submission or an in-licensed asset, how has management established the model's credibility for its context of use?
• How does management review the accuracy of AI-drafted clinical documentation?
• Which strategic initiatives are sized and staged as experiments, and which are funded as if the evidence were already in?

Find practice briefs on AI-generated evidence and clinical AI in the Avancer AI Governance Navigator.

References

Al Jazeera. (2024, October 9). Baker, Jumper, Hassabis win Nobel Prize in chemistry for work on proteins. https://www.aljazeera.com/news/2024/10/9/three-scientists-win-nobel-prize-in-chemistry-for-work-on-proteins

American Academy of Sleep Medicine. (2026, February 6). Congress ends partial government shutdown and extends telehealth flexibilities and the work GPCI floor through 2027. https://aasm.org/congress-ends-partial-government-shutdown-and-extends-telehealth-flexibilities-and-the-work-gpci-floor-through-2027/

Assistant Secretary for Technology Policy. (2026, February 11). Data liquidity, affordability, and access: The history and growth of TEFCA. U.S. Department of Health and Human Services. Retrieved October 21, 2026, from https://healthit.gov/resources/data-liquidity-affordability-and-access-the-history-growth-of-tefca/

Berkowitz, B. A., & Simkin, M. M. (2023, December 8). FDA approves two gene therapies to treat sickle cell disease including a CRISPR-based therapy. Foley and Lardner. https://www.foley.com/insights/publications/2023/12/fda-gene-therapies-sickle-cell-disease/

Datavant. (2025, May 15). Datavant to acquire Aetion, empowering healthcare and life sciences organizations to generate high-quality, scalable real-world evidence from connected, privacy-protected data [Press release]. Business Wire. https://www.businesswire.com/news/home/20250515882812/en/Datavant-to-Acquire-Aetion-Empowering-Healthcare-and-Life-Sciences-Organizations-to-Generate-High-Quality-Scalable-Real-World-Evidence-from-Connected-Privacy-Protected-Data

Dewey, F. E., Murray, M. F., Overton, J. D., Habegger, L., Leader, J. B., Fetterolf, S. N., O'Dushlaine, C., Van Hout, C. V., Staples, J., Gonzaga-Jauregui, C., Metpally, R., Pendergrass, S. A., Giovanni, M. A., Kirchner, H. L., Balasubramanian, S., Abul-Husn, N. S., Hartzel, D. N., Lavage, D. R., Kost, K. A., . . . Carey, D. J. (2016). Distribution and clinical impact of functional variants in 50,726 whole-exome sequences from the DiscovEHR study. Science, 354(6319), Article aaf6814. https://doi.org/10.1126/science.aaf6814

Drees, J. (2021, January 14). Google finalizes $2.1B Fitbit acquisition: 5 details. Becker's Hospital Review. https://www.beckershospitalreview.com/digital-health/google-finalizes-2-1b-fitbit-acquisition-5-details/

Eglovitch, J. S. (2026, January 13). CDER approved 46 novel drugs in 2025, half for rare diseases. Regulatory Focus. Regulatory Affairs Professionals Society. https://www.raps.org/resource/cder-approved-46-novel-drugs-in-2025-half-for-rar.html

FitzGerald, M. J., & Young, A. (2023, May 15). Gut check: The FDA approves microbiome-based therapies, with future approvals expected. Nixon Peabody. https://www.nixonpeabody.com/insights/alerts/2023/05/15/fda-approves-microbiome-based-therapies

Food and Drug Administration. (2018). Framework for FDA's real-world evidence program. U.S. Department of Health and Human Services. https://fda.gov/media/120060/download

Food and Drug Administration. (2025, January 7). Considerations for the use of artificial intelligence to support regulatory decision-making for drug and biological products [Draft guidance]. U.S. Department of Health and Human Services. Retrieved October 21, 2026, from https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological

Harris, J. (2017, November 30). FDA approves FoundationOne CDx, CMS agrees to cover. OncLive. https://www.onclive.com/view/fda-approves-foundationone-cdx-cms-agrees-to-cover-for-medicare-beneficiaries

Health Care Payment Learning and Action Network. (2025). 2024 measurement effort [Infographic]. https://hcp-lan.org/wp-content/uploads/2025/10/2024-APM-Infographic.pdf

IQVIA Institute for Human Data Science. (2025, April 15). Stabilization and improvement seen in multiple key biopharma R&D metrics, says IQVIA Institute report [Press release]. https://www.iqvia.com/newsroom/2025/04/stabilization-and-improvement-seen-in-multiple-key-biopharma

Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., Bridgland, A., Meyer, C., Kohl, S. A. A., Ballard, A. J., Cowie, A., Romera-Paredes, B., Nikolov, S., Jain, R., Adler, J., . . . Hassabis, D. (2021). Highly accurate protein structure prediction with AlphaFold. Nature, 596(7873), 583–589. https://doi.org/10.1038/s41586-021-03819-2

Microsoft. (2019, October 1). Novartis and Microsoft announce collaboration to transform medicine with artificial intelligence [Press release]. https://news.microsoft.com/source/2019/10/01/novartis-and-microsoft-announce-collaboration-to-transform-medicine-with-artificial-intelligence/

National Human Genome Research Institute. (2021, November 1). The cost of sequencing a human genome. Retrieved October 21, 2026, from https://www.genome.gov/about-genomics/fact-sheets/sequencing-human-genome-cost

Olsen, E. (2026, July 17). Large funding rounds help boost digital health investment in H1. MedTech Dive. https://medtechdive.com/news/digital-health-funding-first-half-2026-rock-health/825414

Pifer Parduhn, R. (2023, February 23). Teladoc reports historic net loss of $13.7B in 2022. Healthcare Dive. https://www.healthcaredive.com/news/teladoc-historic-net-loss-137b-2022/643213/

Ridley, E. L. (2022, January 20). IBM to sell Watson Health assets to private equity firm. AuntMinnie. https://www.auntminnie.com/imaging-informatics/artificial-intelligence/article/15630230/ibm-to-sell-watson-health-assets-to-private-equity-firm

Roche. (2018, February 15). Roche to acquire Flatiron Health to accelerate industry-wide development and delivery of breakthrough medicines for patients with cancer [Press release]. https://www.roche.com/media/releases/med-cor-2018-02-15

Ross, C., & Swetlitz, I. (2018, July 25). IBM's Watson supercomputer recommended "unsafe and incorrect" cancer treatments, internal documents show. STAT. https://www.statnews.com/2018/07/25/ibm-watson-recommended-unsafe-incorrect-treatments/

Sherman, R. E., Anderson, S. A., Dal Pan, G. J., Gray, G. W., Gross, T., Hunter, N. L., LaVange, L., Marinac-Dabic, D., Marks, P. W., Robb, M. A., Shuren, J., Temple, R., Woodcock, J., Yue, L. Q., & Califf, R. M. (2016). Real-world evidence: What is it and what can it tell us? New England Journal of Medicine, 375(23), 2293–2297. https://doi.org/10.1056/NEJMsb1609216

Williams, C. Y. K., Raghu Subramanian, C., Ali, S. S., Apolinario, M., Askin, E., Barish, P., Cheng, M., Deardorff, W. J., Donthi, N., Ganeshan, S., Huang, O., Kantor, M. A., Lai, A. R., Manchanda, A., Moore, K. A., Muniyappa, A. N., Nair, G., Patel, P. P., Santhosh, L., . . . Rosner, B. I. (2025). Physician- and large language model-generated hospital discharge summaries. JAMA Internal Medicine, 185(7), 818–825. https://doi.org/10.1001/jamainternmed.2025.0821

Yap, G., Xiao, D., Hu, J., Sanday, J., & Beatty, C. (2025, October 21). 2025: The state of AI in healthcare. Menlo Ventures. https://menlovc.com/perspective/2025-the-state-of-ai-in-healthcare/