The Forgotten Telos: Why Public Health Cannot Be Reduced to Dataism

In modern civic infrastructure, glowing data streams have become the primary language of clinical care.

    Walk into any modern hospital or public health department, and you are immediately greeted by the hum of dashboards. We track bed turnover rates down to the minute, calculate readmission percentages to the decimal, and rely on algorithmic risk scores to dictate community funding, but in the 2020s, data isn't just a tool for medicine. It has become its language. But as our systems grow hyper-optimized for numbers, a quiet crisis of purpose is unfolding across our civic infrastructure. We have perfected the art of measurement, but in the process have we forgotten what, and WHO, we are actually trying to heal?

The Rise of Healthcare Dataism

    In his analysis of modern technological culture, philosopher Yuval Noah Harari popularized the term Dataism, which is the belief that the universe is ultimately composed of data flows, and that the value of any phenomenon is measured strictly by its ability to be processed, quantified, and modeled. In other words, Dataism views data flows as the supreme value of the universe, where all living things and even human societal structures are biological algorithms made to process this data. Isn't that what we want in medicine? For example, if I get my blood drawn for a Basic Metabolic Panel because I am concerned that I have diabetes, I would probably want my physicians to objectively give me a verdict and treatment plan based on my glucose levels. In that scenario, is the function of my pancreas really the only reality that matters?

    When applied to medicine and public health, Dataism can be deeply alluring. It promises an objective, frictionless future where algorithms can anticipate clinical decline long before standard symptoms surface, or optimize county resource distribution through predictive risk scoring. 

    Yet, when numbers become the sole authority rather than an observational lens, a subtle distortion occurs. The patient at the center of the spreadsheet is gradually reduced to a high-dimensional collection of feature vectors. We begin to optimize for the data point while overlooking the person behind it. When aggregate metrics look flawless, systemic failures such as diagnostic sensitivity gaps or overlooked clinical nuances in underrepresented groups are easily masked behind high average performance scores. When a model’s success is defined strictly by high aggregate accuracy scores, it creates a false sense of security. A system can report 95% overall accuracy on paper while simultaneously masking severe diagnostic sensitivity gaps in female or minority populations. In practice, this means an algorithm might repeatedly miss early warning signs for a specific subgroup, delaying life-saving interventions simply because those errors are statistically buried beneath a high overall average. When Dataism serves as the lens through which we view medicine, we tend to ignore the human realities that exist outside of a standard dataset. The ultimate danger of Dataism isn't just that it makes medicine impersonal, it’s that it automates institutional bias, giving systemic inequities the illusion of mathematical objectivity. There is a vast difference between using metrics to observe a system and using metrics to rule it.

    The important thing to keep in mind, though, is that Dataism isn't born from cruelty. Its born from exhaustion and the need for efficiency at scale. 

Aristotle (384–322 BCE), whose concept of telos defines purpose through ultimate ends

Rediscovering Telos in the Age of Algorithms and Data

    To reconcile the immense analytical power of healthcare data with the lived reality of human suffering, we must re-anchor our clinical and technological tools in classical philosophy, specifically, Aristotle’s concept of telos. In Nicomachean Ethics, Aristotle defined telos as the ultimate end, intrinsic purpose, or supreme good toward which a practice aims. For example, the telos of a knife is to cut sharply. The telos of an archer is to strike the target. For medicine and public health, the telos has historically been unmistakable: the restoration of health, the alleviation of suffering, and the promotion of human flourishing (eudaimonia). Healthcare is not simply a business, and it's certainly not meant to disseminate its service unequally and neglectfully. One reason humankind has made it so far is because of our sense of community and cohesion, which is what healthcare is meant to act off of. It is meant to serve the human above all, not just the data that lies within them.

    However, when an institution relies on data as its primary governing logic, a subtle teleological shift occurs, what organizational theorists call metric displacement. The intrinsic goal of human restoration is quietly supplanted by instrumentally measurable proxies: reducing bed turnover times to under 30 minutes, minimizing 30-day readmission penalties, or maximizing algorithmic throughput.

    When metric compliance becomes the primary telos, the healthcare worker is placed in a moral conflict. A physician whose electronic health record (EHR) system forces them to spend more time typing structured fields than looking into a patient's eyes is no longer practicing medicine in its fullest sense. They are serving as a human data transducer for a predictive billing engine.

    Furthermore, this teleological corruption actively worsens systemic inequality. When predictive risk models are trained on historical billing data or past utilization rates without interrogating the telos of those datasets, they mistake past institutional access for actual health need. An algorithm designed to allocate care management resources based on projected cost rather than physical sickness will naturally divert support toward wealthier populations who historically spent more money on care, while under-allocating resources to marginalized patients who lacked access in the first place. By confusing the proxy (healthcare expenditure) with the true telos (human illness), automates and codifies institutional neglect under the guise of mathematical neutrality. 

    Most importantly, Dataism poses a risk to the human side of healthcare. When people are treated as biological algorithms and the host of data rather than a living and feeling being, the system take on a deeply technical and dehumanizing ruse, leaving behind the deeper and more human purpose behind medicine. As we inevitably integrate artificial intelligence deeper into clinical functions, preserving the telos of medicine must remain our central imperative, ensuring that technological progress never comes at the cost of human dignity.

Toward a Socio-Technical Framework: Re-aligning System with Purpose

    Acknowledging the dangers of Healthcare Dataism does not mean abandoning quantitative measurements or digital tools. Data remains an indispensable asset, allowing public health departments to track epidemics, hospitals to manage crowded emergency rooms, and researchers to spot population-level trends. The goal is not to eliminate measurement, but to ensure that metrics serve the true telos of care rather than dictating it.

    To restore this balance, health systems must transition from a purely bureaucratic framework to a socio-technical framework that evaluates healthcare tools, policies, and software by their impact on human healing and equity, not just administrative efficiency. Re-anchoring our current system requires three fundamental shifts:

1. Reclaiming Time for the Clinical Encounter

    Long before predictive algorithms entered the hospital, Electronic Health Record (EHR) platforms transformed clinicians into data-entry clerks. When a physician spends more time clicking through mandatory billing templates than establishing eye contact with a patient, the system prioritizes documentation over care. Real-world reforms must focus on reducing administrative drag, redesigning EHR interfaces, eliminating redundant compliance metrics, and using ambient voice technologies to handle note-taking, so that time and attention are returned to the doctor-patient relationship.

2. Auditing Current Metrics for Proxy Bias

    Dataism’s flaws are already embedded in today’s operational targets. For instance, using past healthcare costs as a proxy to determine who receives intensive care management naturally starves under-resourced communities of support, not because they are healthier, but because they historically lacked access to expensive care. Healthcare leaders must audit their existing administrative formulas, resource-allocation scores, and billing algorithms to ensure they measure actual health need, rather than financial expenditure or historical access.

3. Subordinating Technology to Human Agency

    Whether dealing with a standardized insurance checklist today or an AI risk model tomorrow, technology should inform human judgment, never replace it. Automated outputs, diagnostic recommendations, and risk scores must be treated as secondary cues rather than binding directives. Institutional policies must actively protect clinical autonomy, empowering doctors, nurses, and social workers to override algorithmic or administrative suggestions when the qualitative reality of the patient standing in front of them demands a different path.

Conclusion: Reframing the Servant and the Master

    As digital tools and artificial intelligence become further woven into public health policy and clinical routines, our defining challenge will remain moral rather than technical. Data can reveal how a healthcare system operates, but only human empathy, clinical judgment, and moral reasoning can guide why it operates.

    Metrics and software make extraordinary servants, but disastrous masters. By re-anchoring our modern health infrastructure in the Aristotelian telos of healing, we can ensure that our tools amplify our capacity to restore health without reducing the human being in the exam room to a collection of data points.