Platelet transfusion practices in immune thrombocytopenia related hospitalizations

BACKGROUND: The role of platelet transfusions in management of Immune Thrombocytopenia (ITP) remains controversial. Current guidelines recommend that platelet transfusions in ITP be reserved for catastrophic hemorrhage or invasive surgical procedures. This study assesses the nationwide platelet transfusion practices in hospitalized children and adults with ITP. STUDY DESIGN AND METHODS: We studied hospitalizations with ITP as the primary admitting diagnosis from 2010-2014 in National Inpatient Sample (NIS), the largest all-payer inpatient database. Univariate and multivariable logistic regression analyses were used to determine factors predicting platelet transfusions. Sampling weights were applied to generate nationally representative estimates. Propensity score matching was used to perform sensitivity analyses. RESULTS: From 2010 to 2014, there were 78,376 admissions with ITP as the primary admission diagnosis (mean +/- SD age: 45 +/- 27 years; females 56%, children [age < 18 years] 22%) and 282,285 with ITP as one of all the admission diagnoses. Overall, 27% admissions with ITP as primary (children 4%) and 15% admissions with ITP as one of all the diagnoses documented at least one platelet transfusion. On multivariable adjustment the following factors were associated with worsening disease severity and a higher odds of platelet transfusion, adult age (adjOR = 9.03, 95% CI = 7.40-11.02), male gender (adjOR = 1.21, 95% CI = 1.11-1.31), bleeding occurrence (intracranial/gastrointestinal/genitourinary/epistaxis) (adjOR = 1.78, 95% CI = 1.61-1.96), admission to rural non-teaching hospital (adjOR = 1.85, 95% CI = 1.52-2.22), and small bed-size hospital (adjOR = 1.23, 95% CI = 1.05-1.45). Of admissions reporting platelet transfusions, only 26% reported a bleeding complication, and 11% had a major operating-room surgery/procedure. Overall, 65% of transfused patients had neither bleeding nor a major operative procedure during the hospitalization. Admissions with platelet transfusions had a significantly longer mean length of hospitalization 2.2 days (95% CI = 1.96-2.41, p < 0.001), and accrued higher mean total hospital charges; $31,150 USD (95% CI = 27,644-34,656, p < 0.001). However, platelet transfusions were not associated with in-hospital mortality (adjOR = 1.02, 95% CI = 0.73-1.45, p = 0.892). CONCLUSION: Platelets are administered to a small fraction of the hospitalized ITP patients. In a majority of these cases however, platelet usage does not appear to be concordant with the current guidelines or associated with improvement in clinical outcomes.