A Personalized, Rate-of-Rise (ROR)-Driven Model to Optimize Red Cell Exchange Interval in Chronic Sickle Cell...

Chronic automated red blood cell exchange (RCE) is a cornerstone therapy for preventing sickle cell disease (SCD)-related complications, particularly stroke and recurrent vaso-occlusive crises. Current scheduling commonly relies on uniform empirical intervals of 3-6 weeks and does not account for substantial inter-patient variability in hemoglobin S (HbS) rebound kinetics, potentially resulting in undertreatment of rapid rebounders and unnecessary procedures in slower rebounders. We hypothesized that each patient's personalized rate of rise (ROR) of HbS could be used to generate an individualized, risk-adjusted RCE interval. We retrospectively analyzed adult patients with SCD receiving chronic outpatient RCE. The primary cohort comprised 18 patients with HbSS, while a single patient with HbSC was analyzed separately. HbS values from consecutive outpatient procedures were used to calculate an individual ROR (%HbS/day). A personalized scheduling formula was derived: t(interval) = S(risk) × (HbS(threshold) - HbS(post))/ROR(personal), where S(risk) represents an adjustable safety factor for patients at higher clinical risk. Individualized intervals differed markedly from uniform scheduling. Slow rebounders (< 0.5%/day) could potentially extend beyond 5 weeks, whereas rapid rebounders (> 1.0%/day) reached prespecified HbS thresholds within approximately 2-3 weeks. For most patients, the model recommended shorter intervals than those used, suggesting that under current scheduling many patients may exceed their HbS threshold between procedures. Because most patients would require shorter rather than longer intervals to maintain HbS below 30%, strict implementation of the model could increase program-level blood utilization; therefore, its principal value lies in precision and resource reallocation rather than guaranteed net savings. A personalized ROR-based scheduling framework integrates objective HbS kinetics with clinical risk stratification and may be most applicable to patients with stable, reproducible rebound kinetics. Prospective validation and integration into electronic health record systems are warranted to determine whether model-guided scheduling improves clinical outcomes.