Harnessing routine blood pressure to predict survival in sickle cell disease: construction of a BP-SCD mortality risk score

Background

Adults with sickle cell disease (SCD) face excess long-term mortality, notably driven by cardiovascular complications. Blood pressure (BP), a powerful cardiovascular risk indicator in the general population, remains insufficiently integrated into current risk-stratification strategies in SCD, despite its physiological relevance to vascular dysfunction and end-organ damage. A pragmatic prognostic tool based on routinely collected markers could enhance the early identification of high-risk patients.

Objective

To develop a prognostic score based on a restricted set of routine, clinically meaningful variables to predict long-term mortality in adults with SCD.

Methods

We analyzed the French prospective Etendard cohort, including 383 adults with SCD enrolled in stable clinical condition across five national referral centers. Individuals with severe renal impairment, chronic liver disease, or restrictive lung disease were excluded. Standardized systolic and diastolic BP measurements and routine laboratory testing were performed locally, and LDH values were normalized using laboratory-specific upper-limit-of-normal ratios.

Survival analyses were conducted using Cox proportional hazards models. Candidate variables were selected based on clinical relevance and physiological significance. Variables retaining significance in univariate analyses were entered into multivariable analyses to identify independent predictors of mortality. Continuous variables were categorized using optimal cutoffs derived from maximized log-rank statistics. Score thresholds (0–10) and risk categories were algorithmically optimized for discrimination, monotonicity, and population balance.

Results

Over a median follow-up of 17.9 years, five routinely measured variables remained independently associated with mortality in multivariate analysis: age (HR 1.069, 95% CI 1.045–1.093, p < 0.0001), LDH (HR 1.543, 95% CI 1.198–1.986, p = 0.0008), hemoglobin (HR 0.752, 95% CI 0.638–0.885, p = 0.0006), systolic BP (HR 1.024, 95% CI 1.009–1.039, p = 0.0018), and diastolic BP (HR 1.026, 95% CI 1.004–1.049, p = 0.0229). These variables were assigned 0, 1, or 2 points according to statistically optimized thresholds, allowing construction of a BP-SCD Score that stratified patients into three categories with distinct 15-year mortality rates: low risk (0–3 points, 11.1%), intermediate risk (4–6 points, 19.3%), and high risk (7–10 points, 48.6%).

Conclusion

By integrating systolic and diastolic BP with age, hemoglobin levels, and hemolysis intensity, this BP-SCD Score provides a simple, physiologically grounded, five-marker tool enabling effective long-term mortality risk stratification in adults with SCD. Highlighting BP measurements as a strong predictor of long-term risk supports more personalized surveillance strategies, ranging from standard monitoring to intensified cardiopulmonary evaluation in high-risk patients.BP SCD MORTALITY RISK SCORE