Cardiovascular screening for Antarctic deployment exposed sharp risk-model disagreement
Before deployment to Antarctica, cardiovascular clearance carries unusual weight. A medical problem that might be routine near a tertiary center becomes far more consequential when the patient is thousands of miles from advanced care, evacuation is slow, and weather can make transport impossible. In that setting, pre-deployment screening is not just paperwork; it determines who can safely enter an environment where a single cardiac event can rapidly become a logistics crisis.
This retrospective analysis examined 154 individuals referred for enhanced cardiovascular evaluation before deployment to the U.S. Antarctic Program across three time periods: 2018–2020, 2021–2023, and 2024–2026. The cohort was mostly male, at 79.9%, with a mean age of 44.6 ± 14.6 years. Abnormal electrocardiography was the leading reason for referral, accounting for 70.8% of cases. In other words, the population was already selected for concern before any formal risk score was applied.
PREVENT, FRS, and PCE did not classify the same people the same way
Median 10-year risk varied depending on the calculator used. PREVENT yielded a median risk of 2.7% with an interquartile range of 1.1–6.6, while the Framingham Risk Score was similarly low at 2.3% [0.5–7.1]. The Pooled Cohort Equations pushed the estimate higher, with a median of 6.1% [2.4–9.2].
Among the 72 participants with all scores available, 51% were classified discordantly. That level of disagreement is not trivial in a clearance process. A clinician trying to decide whether an applicant is fit for polar deployment could reach a materially different conclusion depending on which model is used, even when facing the same person, the same history, and the same test results.
Agreement between PREVENT and FRS was strong, with a κ of 0.77. Agreement between PREVENT and PCE was much weaker, with κ ranging from 0.24 to 0.35, while PCE consistently estimated higher risk. The difference likely reflects more than statistical noise. PREVENT is a newer derivation, incorporates renal function, is race-agnostic, extends to heart failure, and can be applied starting at age 30. Those design features matter in a screened workforce that may not resemble the populations used to build older equations.
PCE was calculable in only 52.6% of participants, compared with 70.8% for PREVENT. For a program that needs consistent and usable risk assessment, a model that can be applied to more candidates has immediate operational value. A score that cannot be calculated in nearly half of those referred is a weak fit for a remote deployment pipeline.
Deployment outcomes favored clearance, but evacuation data changed the perspective
Most participants were ultimately cleared. Of the 154 individuals, 132, or 85.7%, were cleared for deployment. Eight participants, 5.2%, were deemed not physically qualified, and 6 of those NPQ decisions were driven by cardiovascular causes.
The evacuation data sharpen the practical stakes. During the study period, 22 cardiac medical evacuations occurred, and 21 involved people who had not undergone prior advanced cardiovascular screening. Only one screened individual required evacuation, and even that case ultimately proved to have no cardiac cause for the symptoms. The pattern does not prove causality, but it strongly suggests that more intensive pre-deployment evaluation identifies higher-risk individuals before they are placed in an environment where rescue is difficult and medical redundancy is limited.
Why renal function and broader endpoints matter in polar medicine
Risk tools built for general outpatient populations may not translate cleanly to an expeditionary workforce. Antarctic personnel face cold exposure, physical exertion, circadian disruption, and prolonged isolation, all of which can stress the cardiovascular system. A screening model that incorporates renal function may be especially relevant because kidney disease and vascular risk often travel together, and renal dysfunction can amplify the burden of hypertension, endothelial dysfunction, and atherosclerotic progression.
PREVENT also extends to heart failure, which is clinically relevant in a setting where exertional tolerance and sudden decompensation both matter. Fitness for deployment depends on more than the probability of a future myocardial infarction. It depends on whether a person can safely sustain work in an austere setting without a foreseeable risk of collapse, arrhythmia, or decompensated pump failure.
The lower-risk estimates from PREVENT and FRS may better reflect the actual screened population, while PCE may overshoot in this context. For occupational medicine teams, overestimation can lead to unnecessary restriction, delayed deployment, or additional downstream testing. Underestimation carries the opposite danger: a worker who looks acceptable on paper but is poorly suited to the Antarctic environment once physiologic stress is layered on top of baseline disease.
Imaging may still have a role for intermediate-risk candidates
The investigators recommend PREVENT as the primary risk-stratification tool and suggest targeted imaging, including coronary artery calcium scoring, for intermediate-risk candidates. That approach fits the logic of modern prevention. When a clinical score lands in a gray zone, anatomy can add clarity by identifying calcified plaque burden that a risk equation may miss.
This is especially relevant in a cohort where many referrals began with an abnormal ECG rather than a history of overt symptoms. ECG abnormalities can signal benign variants, prior silent ischemia, conduction disease, or nonspecific changes that do not map neatly to hard outcomes. Imaging can help separate those possibilities when the consequences of error are high.
The data do not argue for blanket exclusion. They argue for better triage. In a screened population headed to one of the least forgiving work environments on earth, the best model is the one that can be applied consistently, matches older tools reasonably well, and leaves room for imaging when the answer is still uncertain.
A remote setting exposed the limits of familiar calculators
The central lesson is not that one score replaces all clinical judgment. It is that a risk tool built for broad cardiovascular prevention must still earn its place in a highly selected occupational cohort. Here, PREVENT had the strongest overall case: broader applicability, better calculability, and close agreement with FRS.
This remains a preprint, and the study is retrospective, so the findings should not be treated as definitive policy. The sample is modest, and the cohort was already selected for referral, which limits generalizability. Even so, the mismatch between calculators and the evacuation experience raises a concrete question for occupational cardiology: which model best protects workers when access to care is most constrained?