Background/Purpose:  Disparities in care among adults with SLE are well documented.  We investigated associations of demographic factors and volume of annual inpatient hospital admissions with poor outcomes, including intensive care unit admission, renal failure and in-hospital mortality, among hospitalized patients with pediatric SLE.  

Methods: The Pediatric Health Information System (PHIS) is an administrative database contributed to by >40 freestanding U.S. pediatric hospitals.  We queried PHIS regarding all discharges for patients aged 3-21 years with at least one ICD-9 code for SLE from Jan 2006-Sept 2011. Patient demographics, medical insurance, hospital and ICU admissions, lengths of stay, renal failure (based on ICD-9 coding), and deaths were recorded. We classified hospitals according to the volume of inpatient admissions per year, in quartiles.  We used summary statistics and univariable analyses to examine demographics of hospital admissions, readmissions, and lengths of stay. We employed multivariable logistic regression analyses, controlling for patient age, sex, race, ethnicity, insurance type (private, governmental, self-pay, or none), U.S. region (Northeast, South, Midwest and West), and hospital volume to examine risk factors for adverse outcomes, including ICU admission, renal failure and in-hospital mortality. 

Results: A total of 3,389 patients with 14,631 admissions were identified in the study period.  2,781 patients (82%) were female and the median age at the time of the index admission was 16 years (IQR=14–18).  White and African American race each comprised nearly 37% of the patients (n=1,250 and 1,252), while 5% (n=172) were Asian.  Over a quarter of patients (n=888, 26%) were Hispanic or Latino.  87% of patients (n=2,953) had insurance, with over half supported by governmental insurance (n=1876, 55%).  More than a third of patients had renal disease (n=1273, 38%); however, only 0.7% (n=24) had renal failure and 0.2% (n=8) required dialysis therapy.  The high volume hospitals had shorter length of stay as compared to low volume hospitals (median 2 days, IQR=1-5, vs. 3 days, IQR=1-6, p<0.001), although readmissions per patient were more frequent in the higher volume hospitals (median 2, IQR= 1-6, vs. 2, IQR=1-3, p<0.001).  Ten percent of admissions included an ICU stay.  Overall in-hospital mortality was low at 0.4% (n=57).  In multivariable models, ICU stays were associated with age <10 years at first admission (OR=1.61 [1.16 to 2.25], p<0.05).   Poor outcome, defined as renal failure and death, was strongly associated with governmental insurance (OR=2.51 [1.47 to 4.29], p<0.05). 

Conclusion: In our cohort of hospitalized children with SLE, hospital volume affected length of stay and number of readmissions, but not in-hospital mortality.  Governmental insurance in this group of patients was associated with renal failure and in-hospital mortality.  Further studies are needed to understand the relationships of medical insurance type and hospital volume with poor outcomes, in order to address modifiable barriers to care in pediatric SLE.

ACR Meeting Abstracts - https://acrabstracts.org/abstract/risk-factors-for-poor-outcomes-in-hospitalized-patients-with-pediatric-systemic-lupus-erythematosus/