Thrombocytopenia in Hospitalized Non-ICU Patients

Thrombocytopenia is common in patients admitted to the intensive care unit, with a reported prevalence of 8-68% and incidence of 13-44% in published series. Those who develop thrombocytopenia in an ICU have a statistically increased risk of ICU mortality, duration of mechanical ventilation, and platelet, RBC, and FFP transfusions. Few studies have examined the occurrence and risk factors for thrombocytopenia in non-ICU hospitalized patients. The only systematic investigation of thrombocytopenia in the non-ICU hospitalized medical population dates to 1989, predating modern diagnostic assays for a variety of infectious and drug-induced causes of thrombocytopenia. We, therefore, undertook this study to characterize the incidence and causes of thrombocytopenia in the general medical, non-ICU patient population at a tertiary care hospital. For this study, we performed a single-institutional retrospective analysis of patients admitted to a general medical ward at a tertiary care medical center (Duke University Hospital). Inclusion criteria included all adult patients (>18 years) admitted from the emergency department to the general medicine floors during the calendar year defined as 01/2014-01/2015. Exclusion criteria included pre-existing thrombocytopenia, ICU admission, or patients undergoing chemotherapy. Primary endpoints included patients with incident thrombocytopenia, defined as a platelet count on admission greater than 150 x 10^9/L with subsequent platelet counts decreasing to < 150 x 10^9/L. Patients meeting these criteria were evaluated for admission diagnosis, pertinent past medical history, cause of thrombocytopenia (infection, splenic sequestration, drugs, surgery, liver disease), cost of admission, and prognosis. Preliminary evaluation of patients admitted to the general medical service in 2014 reveals ~711 patients who meet inclusion criteria. Of patients developing absolute thrombocytopenia, 56% experienced a platelet count fall of greater than 30%. Those with incident thrombocytopenia greater than 30% had a statistically significant increase in mortality (n=36/399) compared to those with milder thrombocytopenia (n=13/312), (9.0% versus 4.2%, p<0.05), and had an increased length of stay (12.6 days versus 6.8 days, p<0.01). In patients who developed absolute thrombocytopenia, review of diagnosis codes associated with thrombocytopenia reveal a predominance of infectious etiologies (43%; sepsis, severe sepsis, septic shock, bacteremia, urinary tract infection/pyelonephritis, pneumonia, Clostridium difficile colitis). These results are similar to prior published data which suggests infection as a leading cause of incident thrombocytopenia in non-ICU hospitalized patients. Additional studies are underway to delineate non-infectious causes of thrombocytopenia.