[Etiology and prognosis of thrombocytopenia in children: a case analysis in 683 children within 10 years of pediatric intensive care unit].

Objective: To analyze the etiology and prognosis of children with thrombocytopenia (TP) in pediatric intensive care unit (PICU).

Methods: The data of children with TP (exclusion of congenital and unknown TP) admitted to PICU of Anhui Provincial Children's Hospital from January 2008 to December 2017 was analyzed retrospectively. According to the age of onset, the children were divided into infantile group (29 days to less than 1 year), early childhood group (1 to less than 3 years), preschool group (3 to less than 6 years), school age group (6 to less than 10 years) and puberty group (more than 10 years). Moreover, according to the lowest platelet count (PLT), the children were divided into PLT ≤ 20×10/L group, PLT (21-50)×10/L group and PLT > (50-100) ×10/L group. The distribution and mortality of TP were analyzed, and the relationship between age, PLT decrease and prognosis were analyzed by Pearson method.

Results: Among 6 725 children admitted to PICU in our hospital from January 2008 to December 2017, there were 683 children with TP, with the incidence of 10.2%. Among 683 children with TP, there were 387 males and 296 females, with the median age of 2.72 (0.61, 3.08) years, and 92 children died, with a total mortality of 13.5%. Analysis of primary disease showed that TP caused by non-hematological malignant tumor disease accounted for 73.9%, with the mortality of 11.1% (56/505). TP induced by hematological malignant tumor disease accounted for 21.4%, with the mortality of 24.7% (36/146). Pseudothrombocytopenia accounted for 0.6%, with the mortality of 0% (0/4). Other children who gave up treatment accounted for 4.1%. It was shown by further analysis that multiple organ dysfunction syndrome (MODS) caused by TP associated with non-hematological malignant tumor disease accounted for 26.9%, with the mortality of 15.4% (21/136). Sepsis, severe trauma, pneumonia, central nervous system infection and disseminated intravascular coagulation (DIC) accounted for 17.4%, 16.6%, 12.7%, 11.7% and 11.5%; with the mortality of 8.0% (7/88), 2.4% (2/84), 0% (0/64), 20.3% (12/59) and 24.1% (14/58), respectively. The main causes of TP associated with hematological malignant tumor disease were hemophagocytic syndrome [accounting for 27.4%, with the mortality of 32.5% (13/40)] and bone marrow inhibition [accounting for 21.2%, with the mortality of 25.8% (8/31)]. The younger were the children with TP, the higher would be the mortality. The mortality of infantile group was significantly higher than that of early childhood group, preschool group, school age group and puberty group [18.8% (53/282) vs. 14.0% (28/200), 6.7% (7/104), 4.3% (4/92), 0% (0/5), all P < 0.01]. The lower was the PLT, the higher would be the mortality. The mortality of PLT ≤ 20×10/L group was significantly higher than that of PLT (21-50)×10/L group and PLT > (50-100)×10/L group [18.1% (39/215) vs. 13.0% (32/247), 9.5% (21/221), both P < 0.05]. It was shown by correlation analysis that there was no association between age and 28-day death time in children with TP (r = -0.037, P = 0.727), but PLT was positively correlated with 28-day death time in children with TP (r = 0.844, P = 0.010).

Conclusions: MODS, infection, sepsis, severe trauma and DIC are the common causes of TP in PICU. The younger are the children with TP, the lower is the PLT, and the worse would be the prognosis.