Capillary refill time guidance trimmed early fluid exposure in septic shock
In a single-center randomized trial of septic shock, bedside capillary refill time (CRT) tracking led to a smaller 6-hour intravenous fluid burden than lactate-guided resuscitation during the hemodynamic optimization phase after initial resuscitation. The comparison was modest in absolute terms but clinically relevant because fluid overload can worsen pulmonary edema, delay liberation from mechanical ventilation, and compound tissue congestion in already unstable patients.
The trial randomized 85 adults to either CRT-targeted resuscitation, with a goal of CRT ≤ 3 seconds, or lactate-targeted resuscitation, with a goal of arterial lactate ≤ 2 mmol/L or a decline of more than 10% per hour. Over the first 6 hours, total intravenous fluid input reached 1135 mL in the CRT-targeted group and 1350 mL in the lactate-targeted group. The Hodges-Lehmann estimate for the between-group difference was −250 mL, with a 95% confidence interval from −500 mL to −20 mL and a nominal P value of 0.035.
That difference is not large enough to rewrite septic shock care on its own.
A bedside perfusion target can pull fluid practice in a different direction
CRT and lactate reflect different pieces of the perfusion puzzle. CRT is a rapid bedside marker of peripheral circulatory recovery and microvascular refill, while lactate can remain elevated because of stress metabolism, impaired clearance, adrenergic drive, and ongoing oxygen debt. A lactate-based target may therefore keep clinicians giving fluid longer, especially when the number declines slowly despite improving circulation. A CRT-based target can signal that the peripheral circulation has recovered before the biochemical marker normalizes, potentially curbing additional fluid exposure once the patient’s macrocirculatory state is already acceptable.
Positive fluid balance was also lower at 6 and 24 hours in the CRT-targeted group. Those are exploratory findings, but they fit the physiologic logic of a strategy that asks clinicians to stop escalating volume once capillary refill has normalized. In septic shock, where endothelial leak and venous capacitance already promote interstitial edema, even moderate reductions in fluid accumulation may matter at the bedside.
No meaningful between-group difference emerged in CRT, arterial lactate, or norepinephrine dose at 6 to 24 hours. The hemodynamic signal was therefore narrow: less fluid, but not a clear separation in measured perfusion variables or vasopressor exposure. The study compares overall target-driven strategies rather than isolating CRT itself, so the result cannot be interpreted as a pure test of the biomarker.
Ventilator-free days improved, but the mortality signal remains unsettled
Ventilator-free days within 28 days favored the CRT-targeted arm, with a median of 14 days versus 12 days and a Hodges-Lehmann estimate of +4.0 days, 95% CI 1.0 to 8.5, P = 0.005. In a syndrome where excess fluid can impair lung mechanics and prolong dependence on respiratory support, a smaller early fluid load offers a plausible explanation for better ventilator liberation. Still, this was not a trial powered to settle hard outcome questions, and the downstream clinical findings remain exploratory.
Twenty-eight-day mortality was 22.7% in the CRT group and 29.3% in the lactate group, corresponding to a risk ratio of 0.78 with a 95% confidence interval from 0.38 to 1.60. The interval is wide enough to include both a meaningful benefit and no effect. Renal replacement therapy-free days and vasopressor-free days did not differ significantly between groups, which tempers any claim that the smaller fluid burden translated into broad organ support savings within this sample.
Trial conduct and registration limit how far the data can be pushed
The trial was prospective, randomized, open-label, and performed at a single center, but enrollment stopped well short of the mortality-based sample-size target. Six-hour total intravenous fluid input had been prespecified as a secondary endpoint in the written protocol and was later prioritized as the main early resuscitation outcome after the study was under-recruited. Those facts matter because they frame the data as exploratory rather than definitive.
Registration was retrospective, completed after enrollment had ended, even though ethics and scientific review and a written protocol were in place before recruitment. The retrospective registry covered the broader SBRI/CRT research program and listed SVRI, SBRI, and CRT as primary indicators; it did not prospectively establish 6-hour total intravenous fluid input as the primary endpoint of the randomized comparison. In practical terms, the fluid signal is real but should be read with caution because the reporting hierarchy was not locked in prospectively.
A post hoc sensitivity analysis excluding patients with baseline CRT ≤ 3 seconds produced a similar direction of effect, with a location-shift estimate of −240 mL and a P value of 0.052 in 77 patients. The consistency is reassuring, but the near-threshold result also shows how fragile the estimate remains once the sample is narrowed.
What the fluid signal means at the bedside
For clinicians managing septic shock, this trial suggests that CRT-guided resuscitation may reduce early fluid administration without obvious loss of hemodynamic control. That matters in a syndrome where every additional bolus can carry a tradeoff: more preload may help a subset of patients, yet excess volume can worsen pulmonary edema, tissue swelling, and venous congestion. CRT offers a visually immediate readout of peripheral perfusion that may allow earlier stopping rules than lactate alone.
The biologic plausibility is straightforward. When microcirculatory refill improves, additional crystalloid may add little perfusion benefit but still accumulate in the interstitial space. Lactate, by contrast, can remain elevated for reasons that do not necessarily demand more fluid. The two targets therefore do not behave like interchangeable surrogates, and this trial shows how target choice can change treatment intensity even when measured hemodynamics look similar.
The central question now is not whether CRT can guide fluid restraint in septic shock on a single-center basis. The unresolved issue is whether a CRT-centered strategy can reliably improve patient-important outcomes across broader settings, with prospective registration, larger enrollment, and a trial design strong enough to separate biomarker choice from the rest of the resuscitation bundle.