**Background:** Anemia affects up to 66% of patients at ICU admission and nearly all after 72 hours. Iron deficiency (ID) impairs hemoglobin synthesis and immune function, both critical in sepsis. Standard iron tests (ferritin, transferrin) are unreliable in sepsis due to acute-phase derangements. Reticulocyte hemoglobin equivalent (Ret-He) reflects functional iron availability over the preceding 3–4 days and may be useful for diagnosing ID and monitoring IV iron therapy in septic patients. This study aimed to assess the impact of intravenous iron on Ret-He and reticulocyte subpopulations in iron-deficient sepsis patients.
**Methods:** This prospective study enrolled consecutive sepsis/septic shock patients (per Third International Consensus definitions) with PCT > 0.5 ng/mL in a 10-bed mixed ICU from September 2021 to June 2022. Exclusion criteria included bleeding, recent iron or RBC transfusion, thalassemia suspicion (Mentzer index < 13), macrocytosis (MCV > 96 fL), absence of ID/IDA, and contraindications to IV iron. ID was defined as normal Hb with Ret-He < 30.2 pg; IDA as Hb < 120 g/L (women) or < 130 g/L (men) with Ret-He < 30.2 pg. Enrolled patients received iron dextran 0.2 g IV three times weekly (days 1, 3, 5, 8, 10, 13) with a 25 mg test dose. Patients with AKI or CKD also received epoetin alpha 50 units/kg IV on the same days. Ret-He was measured at baseline, day 4 (after 2 doses), and day 9 (after 4 doses). Reticulocyte subpopulations (IRF, LFR, MFR, HFR) were assessed. Daily iatrogenic blood loss was recorded.
**Key Results:** 35 patients were enrolled (median age 69 years, IQR 60–73; 43% women, 57% men; median ICU stay 11 days, IQR 7–21). Severity of illness was high: median SOFA 9 (IQR 7–12), APACHE II 23 (IQR 16–28), SAPS II 53 (IQR 37–68). 37% had AKI, 20% received RRT. Baseline median Ret-He was 27.2 pg (IQR 24.6–28.8), below the reference range (30.2–36.2 pg). After two doses of iron dextran (day 4), Ret-He increased significantly to 30.8 pg (IQR 28.2–32.1; p < 0.0001). The median increase between successive determinations was 3.0 pg (IQR 1.9–6.1). The change from first to second determination was significant (p < 0.01), but from second to third was not (p = 0.09). Median time to Ret-He normalization was 4 days (IQR 3–5). All reticulocyte subpopulations changed significantly after two doses: IRF increased from 18.3% to 29.3% (p < 0.001), LFR decreased from 81.7% to 70.7% (p < 0.001), MFR increased from 12.7% to 16.0% (p < 0.01), HFR increased from 3.5% to 13.0% (p < 0.001). Hb concentration did not change significantly (96 g/L at both time points; p = 0.05). No significant between-group differences were found between patients receiving iron dextran alone (n=20) vs. iron dextran plus epoetin alpha (n=15). Observed in-hospital mortality was 23% (8/35), within the predicted range of 20–50%. No adverse reactions to IV iron were reported.
**Clinical Implications:** Divided-dose IV iron dextran rapidly normalizes Ret-He in iron-deficient sepsis patients, with significant changes in reticulocyte subpopulations indicating increased erythropoietic activity. This approach may reduce reliance on allogeneic RBC transfusion and its associated complications (e.g., TRALI, microcirculatory injury). The study supports using Ret-He for diagnosing and monitoring ID in sepsis, avoiding unreliable standard iron tests. Limitations include the small sample size (though a posteriori power analysis confirmed adequacy with 32 pairs needed at alpha < 0.0001, beta 0.20), lack of bone marrow confirmation (impractical in critically ill patients), and short follow-up insufficient to detect Hb changes. Further research is needed on safety, infection risk, and long-term outcomes.