Vasopressor Test Reveals Uneven Capillary Refill Responses in Septic Shock

Raising blood pressure does not reliably restore perfusion to the tissues of patients in septic shock, and a new physiological pilot study helps explain why. In a prospective multicenter study conducted at two university-affiliated hospitals in Chile between April 2022 and July 2023, researchers subjected 32 septic shock patients to a standardized vasopressor test, transiently increasing mean arterial pressure from 65 to 85 mmHg by adjusting norepinephrine infusions. Only eleven patients, or 34 percent, showed a clinically meaningful improvement in capillary refill time, defined in advance as a shortening of at least one second. Three patients, about 9 percent, actually worsened by that margin, while the remaining eighteen, 57 percent, showed no meaningful change at all.

The findings, published as an open-access research article, underscore the phenomenon known as macro-microcirculatory uncoupling. Current sepsis guidelines recommend maintaining mean arterial pressure at 65 mmHg or above with vasopressors such as norepinephrine, yet reaching that target does not consistently restore tissue perfusion. Increasing arterial pressure or cardiac output often fails to improve microvascular flow or organ function, indicating that systemic hemodynamic optimization alone may not ensure adequate oxygen delivery to tissues. The new study adds a quantitative portrait of just how variable that uncoupling can be from one patient to the next.

Capillary refill time, measured by pressing a fingertip until it blanches for ten seconds and timing the return of color with a stopwatch, is a simple bedside marker of peripheral perfusion. It has gained clinical standing through the ANDROMEDA-SHOCK trial, in which a resuscitation strategy targeting capillary refill time improved relevant outcomes compared with lactate-guided therapy. The new Chilean cohort falls at the lower end of that range, reinforcing that responses are far from uniform.

The study population was deliberately narrow. Patients were enrolled within 12 hours of hospitalization if they had septic shock by Sepsis-3 criteria, a history of chronic hypertension, and preload unresponsiveness, confirmed by at least two negative dynamic predictors such as stroke volume variation below 12 percent, pulse pressure variation below 13 percent, or an inferior vena cava distensibility index below 18 percent. Persistent hypoperfusion was defined as capillary refill time of at least three seconds or lactate above 2 mmol/L. All patients were sedated, mechanically ventilated, and had already undergone initial fluid resuscitation, with norepinephrine titrated to hold mean arterial pressure at 65 mmHg for one hour before baseline measurements.

After an hour of stability at the elevated 85 mmHg target, the team repeated echocardiography, hemodynamic measurements, visceral Doppler ultrasound, and tissue perfusion assessments. A key baseline difference emerged between responders and non-responders: mean systemic filling pressure, an estimate of the upstream pressure driving venous return, was significantly lower in responders, at 13 mmHg versus 18 mmHg. In multivariable logistic regression, baseline mean systemic filling pressure remained independently associated with capillary refill improvement. Stroke volume, by contrast, was not an independent predictor, and the proportion of patients achieving refill improvement did not differ significantly between those whose stroke volume rose more than 10 percent during the test and those whose did not.

Still, flow mattered at the individual level. A linear mixed-effects model accounting for repeated measurements within each patient revealed a significant inverse association between the change in stroke volume and the change in capillary refill time, meaning that within a given patient, increases in stroke volume tended to accompany shortening of refill time. That association persisted after adjustment for baseline mean systemic filling pressure. The authors interpret this apparent discrepancy, no population-level correlation between cardiac output changes and refill changes, as evidence of substantial interindividual variability in how macroscopic flow translates into microvascular recruitment.

Baseline perfusion status strongly shaped who responded. Seventeen patients, 53 percent, had prolonged capillary refill time of three seconds or more at baseline, and 28 patients, 88 percent, had elevated lactate. Patients with prolonged baseline refill time were far more likely to improve during the test, ten of seventeen versus one of fifteen, a difference the authors say suggests the vasopressor test functions less as a therapy and more as a diagnostic probe, revealing perfusion recruitability in patients whose baseline microcirculation is impaired. Notably, patients whose refill time was already normal but whose lactate was elevated rarely improved, hinting that raising pressure may add little information when peripheral perfusion is already preserved.

The regional vascular findings were equally instructive. Using Doppler ultrasound, the team measured resistive indices in the kidneys, liver, and spleen, dimensionless measures of blood flow velocity variability that rise with vascular tone and downstream resistance. These indices did not change significantly during the vasopressor test, a pattern consistent across both responder groupings, and changes in capillary refill time did not correlate with changes in any splanchnic index. This dissociation supports the concept of territorial heterogeneity: skin, a highly reactive vascular bed under strong sympathetic control, behaves differently from splanchnic territories governed by distinct regulatory mechanisms. It also diverges from earlier work by Brunauer and colleagues, who reported associations between peripheral perfusion markers and visceral Doppler pulsatility indices, a difference the authors attribute to variations in patient populations, illness severity, timing, and vasopressor exposure, including vasopressin use in a substantial share of the cohort.

Safety signals were reassuring. No arrhythmias or sinus tachycardia occurred during the tests, echocardiography revealed no moderate or severe myocardial depression at the higher pressure target, no cases of digital ischemia were observed, and the three patients whose refill time worsened returned to baseline within an hour. The absence of deterioration in visceral Doppler indices suggests that transient pressure augmentation did not adversely affect regional vascular resistance in this setting, though vasopressor therapy is known to carry risks of ischemia and tachyarrhythmias at high doses or with prolonged exposure.

The authors are careful to frame the work as hypothesis-generating rather than practice-changing. Capillary refill time was measured visually without formal assessment of interobserver agreement, the sample was small and restricted to already-resuscitated preload non-responders with chronic hypertension, no formal sample size calculation was performed, and residual confounding or model overfitting cannot be excluded. The study assessed physiological endpoints, not clinical outcomes, so the prognostic or therapeutic value of refill changes during a vasopressor test remains unestablished. Even so, the consistent link between lower baseline mean systemic filling pressure and refill responsiveness points to venous return physiology, a dimension rarely considered in current resuscitation paradigms, as a potential determinant of which patients benefit when clinicians push arterial pressure higher.

Subject of Research: Capillary refill time responses to a standardized vasopressor test in septic shock patients

Article Title: Dynamic changes in capillary refill time during a vasopressor test in septic shock patients: a physiological pilot study

Article References: Valenzuela, E. D., Mercado, P., Oviedo, V., Amthauer, M., Contreras, R., Carpio, D., González, A., Gabrielli, L., Slama, M., Bakker, J., Castro, R., & Dubin, A. (2026). Dynamic changes in capillary refill time during a vasopressor test in septic shock patients: a physiological pilot study. Critical Care: Sepsis and Severe Infection, 1(1), Article 5. https://doi.org/10.1186/s44541-026-00006-6

Image Credits: AI Generated

DOI: 10.1186/s44541-026-00006-6

Keywords: septic shock, capillary refill time, vasopressor test, mean arterial pressure, mean systemic filling pressure, microcirculation, visceral Doppler, norepinephrine, perfusion recruitability, pilot study, hemodynamics, resuscitation