Abstract
Introduction: Acute Graft versus Host Disease (aGVHD) is a common complication of allogeneic hematopoietic stem cell transplant (allo-HSCT). Despite advances in prevention and treatment methods, it remains significant cause of morbidity and mortality in this group of patients. Most of the studies assessing risk factors and outcome have yielded inconsistent results. In the current study, we have utilized the national database, which provided us with larger sample size to adequately assess the risk factors for occurrence and mortality of aGVHD in allo HSCT patients.
Methods: aGVHD risk factors for occurrence and mortality were identified in 46652 patients from year 2008 to 2014 utilizing national inpatient sample (NIS) of health care utilization project (HCUP) data. Occurrence was defined as established diagnosis of aGVHD (identified with ICD-9 code of 279.51) and mortality as death prior to hospital discharge. Predictors were defined as baseline patient characteristics that influence incidence and outcome of aGVHD. We performed regression analysis using these baseline patient characteristics as covariates and calculated odd ratio (OR) with 95% confidence interval (CI) in both univariate and multivariate model to assess the risk for occurrence and outcome of aGVHD. In addition we also assessed the burden of the disease on our healthcare system by identifying the difference in length of stay and hospital charges in patients with and without aGVHD.
Results: On multivariate analysis significant risk factors for occurrence of aGVHD were age 0-49 years (OR = 1.70, CI = 1.14-2.53), white race (OR = 1.57, CI = 1.20-2.07), allogeneic cord stem cell transplant (OR = 1.48, CI = 1.07-2.05), conditioning regimen with use of total body irradiation (OR = 1.40, CI = 1.15-1.71), infections including bacterial sepsis (OR = 1.99, CI = 1.67-2.36), viral infections OR = 1.88, CI = 1.53-2.31), fungal infections (OR = 1.54, CI = 1.22-1.98) and c diff colitis (OR = 1.52, CI = 1.21-1.90). Similarly on multivariate analysis significant risk factors for mortality of aGVHD were age >65 years (OR = 3.44, CI = 1.28-9.22), infections including bacterial sepsis (OR = 6.30, CI = 3.83-10.35), viral infections (OR = 1.85, CI = 1.09-3.16), fungal infections (OR = 2.14, CI = 1.25-3.66), presence of comorbidities assessed by charlson comorbidity scale (OR 2.28, CI = 1.34-3.88). These results were replicated on univariate analysis as well with the exception of race (white) for aGVHD occurrence and age (>65 years) for aGVHD mortality, which were only significant on multivariate analysis.The median length of stay (in days) was significantly higher in patients of allo HSCT with aGVHD, 42 (CI= 39-44) as compared to without aGVHD, 26 (CI = 26-27). Similarly the median total charges for hospital stay (in US dollars) were also significantly higher in patients of allo HSCT with aGVHD, 512025 (CI = 462567-561482) as compared to without aGVHD with a median of 298447 (CI = 275865-321029).
Conclusion: Our study confirmed several previously established risk factors for aGVHD. In addition we identified certain new factors, which increases risk for occurrence and mortality of aGVHD. Additionally we identified that aGVHD substantially increases the burden on our healthcare system as measured by increase in both length of stay and cost.
No relevant conflicts of interest to declare.
