EHR-Based Biological Aging Tracks Surgical Risk After Cardiothoracic Procedures
Biological age, not chronological age, may be the more useful marker when the body is asked to withstand major surgery. In a cohort of 4,313 cardiothoracic-surgery patients at a large US hospital, biological aging estimated from standard laboratory data embedded in the electronic health record was associated with recovery after surgery, including adverse outcomes, intensive-care-unit stay, and in-hospital death.
The study centers on a familiar clinical problem. Two patients can arrive for cardiothoracic surgery with the same birthday on the chart and very different physiologic reserve at the bedside. One may recover quickly from the hemodynamic stress, inflammation, and tissue injury of the operation; the other may spiral into prolonged critical illness. A laboratory-derived biological age score aims to capture that hidden reserve using routine data rather than specialized research testing.
Routine Laboratory Data as a Measure of Physiologic Reserve
Biological aging reflects the gradual decline in resilience that makes patients more vulnerable to disease and death as they get older. The appeal of this approach in surgery is straightforward: standard laboratory panels already exist in most electronic health records, so the measure can be generated without new blood draws, new workflows, or exotic infrastructure. In principle, that makes it attractive for precision risk stratification in busy hospital systems.
The preprint tested whether electronic health record laboratory data could do more than merely describe age-related decline. It asked whether the resulting measure would identify patients who were biologically older before surgery and whether those patients would fare worse afterward. It also examined postoperative biological-aging trajectories, which may reflect the body’s response to surgical stress, inflammatory burden, and critical illness.
Worse Outcomes Concentrated in the Biologically Older Patients
Patients classified as biologically older before surgery more often died or experienced adverse outcomes than patients measured to be biologically younger. In practical terms, the measure separated patients with greater physiologic fragility from those with more reserve, even within a population already exposed to major cardiothoracic procedures.
One sentence captures the bedside implication: biological age identified a higher-risk surgical phenotype before the operation began.
That pattern is clinically plausible. Cardiothoracic surgery places intense demands on cardiovascular, renal, hematologic, and inflammatory systems at once. A patient whose laboratory profile already signals advanced biological aging may have less ability to buffer oxidative stress, inflammatory activation, endothelial dysfunction, and postoperative metabolic instability. Those mechanisms can translate into poor wound healing, organ dysfunction, prolonged ventilation, or hemodynamic lability, even when conventional preoperative assessment looks acceptable on paper.
The abstract does not provide the exact effect sizes, absolute event rates, or model specifications, so the safest reading is directional rather than quantitative. Even so, the association was strong enough to support the central hypothesis that routine EHR data can carry prognostic information about surgical recovery.
Postoperative Trajectories Mirrored ICU Burden and Mortality
The postoperative signal was especially important because recovery after cardiothoracic surgery is not determined solely by preoperative status. The body’s response to surgery can accelerate or amplify biological aging trajectories, and those trajectories were associated with length of ICU stay and in-hospital death.
A second blunt clinical reality emerges here: postoperative biological-aging changes tracked the cost of critical illness.
This matters because ICU length of stay is more than an administrative metric. It often reflects prolonged respiratory support, recurrent instability, organ support requirements, or inability to mobilize and clear the inflammatory aftermath of surgery. A biological-aging measure that shifts in parallel with ICU burden suggests it may be sensitive to the physiologic toll of the operation itself, not just the patient’s baseline status. That makes it potentially useful for postoperative surveillance, early triage, and identifying patients whose recovery is diverging from expectation.
Why an EHR-Embedded Aging Metric Could Matter in Cardiac Surgery
The practical appeal of this approach lies in scale. Many hospitals already possess the laboratory data needed to estimate biological aging, which means the measure could be deployed without adding a new test to the perioperative workflow. In a specialty where risk assessment already combines age, comorbidity, frailty, hemodynamics, and operative complexity, an EHR-based aging metric could add a deeper physiologic layer.
For surgeons, anesthesiologists, intensivists, and perioperative teams, such a tool could sharpen conversations about operative risk, postoperative monitoring intensity, and likely recovery trajectory. It could also help distinguish patients who are chronologically similar but biologically quite different, a gap that conventional risk scores often struggle to capture. If validated prospectively, the measure might eventually support more individualized decisions about resource allocation, rehabilitation planning, and early escalation of care.
The biological rationale is also attractive. Standard clinical laboratories can reflect the cumulative effects of inflammation, organ dysfunction, metabolic dysregulation, and tissue injury. Those are the same processes that influence how well a patient tolerates cardiothoracic surgery and how quickly they recover afterward. A composite measure built from those data may therefore function as a proxy for resilience, not just age.
What This Preprint Can and Cannot Support
The paper is a preprint and has not undergone peer review, so its findings should be treated as provisional. The abstract establishes association, not clinical utility, and it does not show that using the measure would improve outcomes or change management. It also does not disclose the detailed modeling strategy, calibration, subgroup behavior, or how the biological-age metric performed against existing perioperative risk tools.
Even so, the signal is biologically coherent and clinically relevant. Laboratory-based aging measures may offer a scalable way to quantify physiologic reserve in patients heading into high-risk surgery and to monitor the postoperative stress response once they reach the ICU. The next questions are narrower and more demanding: whether the measure performs consistently across other surgical populations, whether it adds value beyond established risk scores, and whether it can guide interventions that change recovery rather than merely predict it.