Safety of lumbar punctures in patients with thrombocytopenia

BACKGROUND: American Association of Blood Banks guidelines recommend a minimum platelet count of 50 × 10(9) /l for LPs (lumbar puncture), but evidence is lacking. The objective of this study was to describe the range of platelet counts at which LPs are performed, and the rate of traumatic taps and haemorrhagic complications in an adult oncology population. METHODS: A retrospective cohort study of patients receiving LPs over a 2-year period was carried out. Bleeding risk factors captured included anticoagulants, antiplatelets, end-stage renal disease, and other bleeding disorders. Pre-LP platelet counts were those collected ≤24 h from the time of the LP. Traumatic tap was defined as 500 or more red blood cells per high-power field in cerebrospinal fluid. RESULTS: One hundred and thirty-five patients underwent 369 LPs. Twenty-eight (7·6%) LPs were performed at a platelet count ≤ 50 × 10(9) /l; 18 patients received a platelet transfusion prior to the LP, with post-transfusion count available prior to LP in only one patient. Traumatic taps occurred in 16 of 113 (14·2%) LPs in patients with thrombocytopenia (platelet count < 150 × 10(9) /l) compared to 27 of 242 (11·1%) LPs in patients with a normal platelet count (P = 0·48). The presence of bleeding risk factors did not increase the incidence of traumatic taps. There were no haemorrhagic complications. CONCLUSIONS: Among this cohort of adult oncology patients, there were no haemorrhagic complications. Traumatic taps were not increased in patients with thrombocytopenia. The effects of platelet transfusions were rarely assessed prior to LP. Further studies should be pursued to assess whether platelet count thresholds lower than 50 × 10(9) /l are safe for lumbar puncture.