Effect of capillary refill time-guided resuscitation on early fluid administration and fluid balance in septic shock: a single-center randomized controlled trial

Septic shock is a common critical illness in the intensive care unit (ICU). Capillary refill time (CRT) and lactate are used to guide perfusion assessment in septic shock. Their effects on additional intravenous fluid exposure during hemodynamic optimization after initial resuscitation remain uncertain. This was a single-center, prospective, randomized, open-label trial. A total of 85 patients with septic shock were randomized to capillary refill time-targeted resuscitation (CRT-T; n = 44) or lactate-targeted resuscitation (Lac-T; n = 41) for 6 h during hemodynamic optimization after initial resuscitation. The CRT-T target was CRT ≤ 3 s, whereas the Lac-T target was arterial lactate ≤ 2 mmol/L or a decrease of > 10% per hour. Six-hour total intravenous fluid input was prespecified as a secondary endpoint in the written protocol and was subsequently prioritized as the main early resuscitation outcome of this report after the trial was under-recruited relative to the mortality-based sample-size target; accordingly, these analyses are interpreted as exploratory. Ethics/scientific review and a written protocol were completed before recruitment. Public registration was delayed because of synchronization and review delays on the registry platform and was ultimately completed retrospectively after enrollment had ended; the retrospective registry covers the broader SBRI/CRT program and does not establish prospective primacy of the 6-hour fluid-volume endpoint. The 6-hour total intravenous fluid input was 1135 (824–1480) mL in the CRT-T group and 1350 (990–1850) mL in the Lac-T group. The Hodges-Lehmann location-shift estimate for CRT-T minus Lac-T was − 250 mL (95% CI, -500 to -20 mL; P = 0.035). Positive fluid balance at 6 and 24 h was also lower in the CRT-T group (nominal P < 0.05). No significant between-group differences were observed in CRT, arterial lactate, or norepinephrine dose at 6–24 h. Twenty-eight-day mortality was 22.7% (10/44) versus 29.3% (12/41), corresponding to a risk ratio of 0.78 (95% CI, 0.38–1.60); the trial was not powered for mortality. Ventilator-free days within 28 days were 14 (6.75–21.25) versus 12 (0–12) days, with a Hodges-Lehmann estimate of + 4.0 days (95% CI, 1.0 to 8.5; P = 0.005). Renal replacement therapy-free days and vasopressor-free days did not differ significantly between groups. All downstream clinical outcomes are considered exploratory. In a post hoc sensitivity analysis excluding patients with baseline CRT ≤ 3 s, the location-shift estimate was − 240 mL (95% CI, -500 to 0 mL; P = 0.052; n = 77). The CRT-targeted strategy was associated with modestly lower 6-hour total intravenous fluid input during hemodynamic optimization after initial resuscitation. The clinical importance and fluid-component contribution of this difference remain uncertain. The study compares overall target-driven treatment strategies and does not isolate the effect of CRT versus lactate. All findings are exploratory and do not establish improved clinical outcomes. Clinical trial registration The trial was retrospectively registered in the Chinese Clinical Trial Registry on August 5, 2025 (ChiCTR2500107145), after participant enrollment had been completed. Ethics and scientific review and a written protocol were in place before recruitment. Public registration was delayed because of synchronization and review delays on the registry platform; by the time the registration process was completed, enrollment had ended and the entry was therefore retrospective. The retrospective registry covers the broader combined SBRI/CRT research program and lists SVRI, SBRI, and CRT as primary indicators; it does not prospectively establish 6-hour total intravenous fluid input as the primary endpoint of the randomized comparison. Retrospective registration is therefore acknowledged as an important reporting limitation and cannot establish prospective specification of the outcome hierarchy.