Care-Related Complications Drive Hospital Utilization in Adults With Sickle Cell Disease: A Statewide Administrative Analysis
Yue M Zhang, Shirley Johnson, Daniel M SopAbstract
Background
Adults living with sickle cell disease (SCD) are frequently hospitalized, with wide variation in outcomes and cost. Serious care-related complications such as inflammatory syndromes, device-related events, and postoperative complications may be important and potentially preventable contributors to excess utilization. Large administrative datasets like Virginia Health Information (VHI) provide an opportunity to examine these patterns at scale, but they lack the clinical detail needed to capture many SCD-specific comorbidities and disease-severity markers (for example, genotype, transfusion history, vaso-occlusive phenotype, or disease-modifying therapy exposure). We therefore used VHI data to establish a population-level baseline of care-related complications among hospitalized adults with SCD.
Methods
We performed a retrospective analysis of 8,109 adult SCD hospital admissions from 2012–2015 using administrative data linked to ZIP-level socioeconomic indicators. ICD-defined care-related complications were modeled as a binary outcome using multivariable logistic regression, adjusting for age, sex, admission type, payer, socioeconomic context, diagnostic category, and comorbidity burden measured with the Elixhauser Comorbidity Index. Length of stay was modeled with negative binomial regression, and hospital charges were modeled using generalized linear models with a Gamma distribution and log link. Model performance was assessed using likelihood ratio tests, c-statistics, scaled deviance, and error metrics.
Results
Care-related complications occurred in 4.3% of admissions (513 of 11,847 encounters). Patients with complications had ∼81% longer LOS (rate ratio = 1.81; 95% CI 1.67–1.96; p < 0.0001), corresponding to an adjusted mean LOS of ∼7.7 days vs ∼4.3 days without complications, and ∼55% higher hospital charges compared with admissions without complications. Greater comorbidity burden independently increased risk: each 1-point increase in the Elixhauser Index was associated with ∼6.5% higher odds of complications (OR ≈ 1.07 per point; p < 0.0001) and 3–4% higher charges. Selected comorbidities including cardiac arrhythmias, chronic renal disease, and pulmonary disease were independently associated with inflammatory complications. Charges showed modest correlation with age (ρ = 0.105) and stronger correlation with comorbidity burden (ρ = 0.175), while admission type, payer status, socioeconomic context, and diagnostic category also significantly influenced LOS and costs.
Conclusions
In this statewide administrative dataset, ICD-defined care-related complications were a major independent driver of prolonged hospitalization and higher charges among adults with SCD. Comorbidity burden and selected high-risk conditions further amplified this risk. However, administrative datasets cannot capture important SCD-specific disease characteristics, so these results should be interpreted as a system-level baseline rather than a mechanistic explanation. Ongoing analyses using clinically enriched registry and imaging-linked datasets aim to identify SCD-specific drivers of complications and guide targeted prevention strategies.