How should fever be managed in adults with sickle cell disease? a quality metrics study and mapping of essential elements to improve care delivery

Abstract

Abstract Background: Adults with sickle cell disease (SCD) are at high risk for life-threatening bacterial infections due to functional or anatomic asplenia. Guidelines from the CDC and NHLBI classify fever in this population as a medical emergency. However, no universally validated, adult-specific benchmark exists for assessment and empiric antibiotic initiation. Many centres therefore extrapolate pediatric SCD and febrile neutropenia standards, targeting a door-to-antibiotic time of <30–60 minutes. At our institution, feedback from patients and providers revealed inconsistencies and delays in the management of febrile episodes. Objectives: To evaluate and improve the quality of fever care in adults with SCD at our centre by measuring key process indicators (door-to-antibiotic time, triage code, protocol utilization), identifying the etiologies of febrile episodes, assessing outcomes based on patient characteristics and the care timeline, and identifying actionable quality improvement targets. Methods: We conducted a retrospective study at the Centre Hospitalier de l'Université de Montréal (CHUM), including all individuals with SCD who presented to the emergency department (ED) with fever (≥38.5°C or 101.3°F) between July 2022 and December 2023. Clinical data included genotype, vaccination status, triage level (P2 = to be seen ≤15 min; P3 = ≤30 min), care time intervals, and outcomes (length of stay [LOS], ICU admission, mortality, 30-day ED return rate). Febrile etiologies were classified according to discharge diagnoses. Findings were presented to multidisciplinary stakeholders to guide fever care redesign. Results: 56 febrile episodes in 45 patients were included. Median age was 29 years (range: 17–50); 73% were women. Genotypes included 34 SS/Sβ0 and 22 SC/Sβ+. Fever was the only presenting symptom in 11%. Other symptoms included musculoskeletal pain (32%), upper respiratory symptoms (21%), and dyspnea (11%). Vaccination status was noted in the file in 23%. Process Metrics (median time): ED arrival-to-nurse-triage: 23min (6min–6h33). Triage-to-physician: 34min (0–16h07); 2 patients left without being seen. Triage was P2 in 54% (fever documented in 47%) and P3 in 46% (fever in 23%). Time-to-physician was shorter in P2 (29min) than P3 (52min; p=0.001). Antibiotics were given in 80%. No non-treated patients had bacterial infections. Assessment-to-antibiotic time: 2h46 (P2) vs. 3h39 (P3; p=0.123). Door-to-antibiotic time: 4h34 overall (3h10–22h31), shorter in P2 (3h36) than P3 (6h20; p=0.022). Fever/ACS protocols were used in 31%. Etiologies: Infectious causes were most common: viral infections (21 cases) and suspected pneumonia (8). Blood cultures were positive in 6 cases: Streptococcus (S. anginosus, S. pyogenes, 2), Staph. hominis (1), and E. coli (3). One patient had malaria. No cause was found in 7 cases. When fever was the only symptom, diagnoses included URI (4), dental abscess (1), and unknown (1). Outcomes: 23 were discharged home, while 31 required hospitalizations (4 acute chest syndrome (ACS), 13 vaso-occlusive crises (VOC), 9 infections, 5 non-hematological conditions), with a median LOS of 2.17 days (0.13–13.12). ACS was diagnosed in 9/31 hospitalized cases, 5 were transfused. One patient required ICU admission; no deaths were reported. The 30-day ED return rate was 11% (1 pleural effusion, 1 pulmonary embolism, 1 VOC, 3 reinfections). Time to antibiotic administration was not significantly associated with these outcomes. LOS did not vary significantly by genotype, age, sex, or presence of fever at triage but was significantly longer in ACS (5.07 vs. 1.27 days; p < 0.001). Quality Improvement Targets: more consistent triage classification of febrile SCD visits and better use and faster initiation of standardized fever/ACS protocols. Limitations: Small sample size and incomplete vaccination data limited ability to determine predictors of LOS. Milder febrile episodes managed outside the ED and ED avoidance were not captured. Conclusion: This study revealed delays and process gaps in ED fever care for adults with SCD at our centre. Universally assigning all febrile visits to P2 may help but would be insufficient to achieve timelines. This should be paired with additional quality improvement strategies. Nurse-activated protocols and risk-adapted strategies, similarly to neutropenic fever management, are under development. These interventions aim to ensure timely, high-quality care while reducing ED burden.