Inpatient Utilization Among Children and Adolescents with Sickle Cell Disease: Analysis of a Large Populated-Based Retrospective Cohort

Background:  Sickle cell disease (SCD) is characterized by marked heterogeneity in clinical outcomes, severity and utilization of health care services.  This heterogeneity is particularly evident with regards to utilization of inpatient hospital services.  Some children with SCD are frequently admitted to the hospital while others are rarely or never admitted.  In addition, rates of readmission within 30 days of hospital discharge are high in SCD. The causes of this variable utilization and high rates of readmission are not well understood. 

Objectives:  We sought to determine rates and primary causes of SCD-related hospital utilization among children and adolescents with SCD and to determine whether rates of hospitalization and of 7-, 14- and 30-day readmissions varied by age, SCD genotype, gender and cause of hospitalization.  

Methods:  Children who lived in the 28-county greater metro Atlanta area with a confirmed SCD diagnosis and who received care at the Children's Healthcare of Atlanta (CHOA) SCD program between January 1, 2010 through December 31, 2014 were included.  The earliest and latest encounter during the five-year period was used to determine total length of observation for each patient.   To ensure a substantial period of observation, individuals with two consecutive encounters >18 months apart or with < 2 yr of observation were excluded.  SCD genotype (SS, SC, and Sβ⁰ and Sβ⁺thalassemia) was confirmed for each patient by review of hematologic and clinical data, including results of diagnostic hemoglobin electrophoresis; those with rare SCD genotypes (e.g. SD, SE) were excluded.  The primary cause for each admission was determined through medical chart review and classified into 4 mutually exclusive hierarchical categories in the following descending order: acute chest syndrome (ACS), pain crisis, fever/infection, and other complications of SCD (i.e. an admission for both pain and ACS was categorized as ACS).  Scheduled hospitalizations for elective procedures (e.g. splenectomy, cholecystectomy, venous catheter placement) were excluded.  A readmission was defined as a hospitalization occurring within 7, 14, or 30 days of a previous hospital discharge. 

Results:  The analysis included 1,331 individuals with 5,362 person-years of observation; 68% had SS/Sβ⁰ thal genotypes, 49% were male.   Age at the time of the earliest encounter ranged from 2 months to 19 years; average length of observation was 4.02 years (min, max= 2, 5). Total n of hospital admissions was 5,317; 19.4% were never hospitalized, and 44.8% were hospitalized less than once per year.  The most common primary cause for hospital admission was pain (53.1% of admissions).  Overall and cause-specific hospitalization rates varied by age and genotype (Figure) with overall hospitalization rates lowest among children 4-9 years old compared to other ages (0.68 vs. 1.00 admissions per person-year) and highest among those with SS or Sβ⁰ thal  compared to those with SC or Sβ⁺thal (0.97 vs. 0.68 admissions per person-year).  Hospitalization rates associated with pain increased with age while rates for fever/infection decreased with age.  Of the 1,073 patients who were admitted, 356 were readmitted within 30 days of a previous admission at least once.  All-cause 7-, 14- and 30-day readmission rates were 5.6%, 10.0% and 18.2%, respectively.   30-day readmission rates were lower among younger age groups (15.7% for age 1-3 yr, 15.3% for 4-6 yr, 15.8% for 7-9 yr) but higher in older patients (18.3% for 10-12yr, 19.9% for 13-15yr, and 23.3% for 16-18yr).  Readmission rates were highest following index admissions for pain (20.4%) and lowest for ACS (11.3%).  Gender was not associated with readmission or cause of readmission.

Conclusion:   SCD-related hospitalization rates were highest in early life and in later adolescence with admissions for fever/infection most common in younger children and admissions for pain crises in older children.  Rates of readmission strongly correlated with age and were highest following admissions for pain.  These data highlight to need to develop and implement improved strategies prevention and treatment of pain in children and adolescents with SCD.  Further studies should also explore individual and/or family factors that may contribute to the high rates of hospital utilization and 30-day readmissions that occur among a relatively small percentage of patients.

Figure. Hospitalization rate per person-year by type of admission and age