Transfusion Evidence Synopsis: Platelet transfusion before CVC placement in patients with thrombocytopenia (PACER trial)

Clinical question: In haematology and intensive care (ICU) patients with a platelet count of 10–50 × 109/L, does withholding platelet transfusion prior to central venous catheter (CVC) insertion increase the incidence of bleeding in comparison to transfusing one unit of platelets? Evidence from study: Withholding platelet transfusion did not meet the criteria for non-inferiority and resulted in more CVC-related bleeding. Study design: Single-blinded, randomised controlled non-inferiority trial. Study years: 2016–2022. Countries: Netherlands. Setting: ICU and haematology wards. No. of patients: 411 CVC placements randomised. 373 included in the per-protocol analysis. Baseline characteristics: 56%–57% haematology patients. Otherwise broadly similar – age, gender, median platelet count, INR, APTT, CVC type. Inclusion criteria: Exclusion criteria: Use of therapeutic anticoagulation, history of acquired/congenital bleeding disorder, INR >3 (changed from INR >1.5 mid-trial), <24 h from previous CVC placement. Comparison: 1:1 randomisation – receive a one-unit platelet transfusion before CVC placement or no platelet transfusion. Any type of CVC was permitted. The non-inferiority margin was set at 2.5% in the no transfusion group. Primary outcome: Grade 2–4 catheter-related bleeding within 24 h of CVC placement assessed according to this bleeding scale. Secondary outcomes: Grade 1 or >Grade 3 bleeding, Platelet/red cell transfusions within 24 h, Hb/platelet count at 1 + 24 h, allergic transfusion reactions, TRALI, length of ICU/hospital stay, inpatient mortality, financial costs. Grade 2–4 bleeding occurred in 4.8% of the transfusion group, and 11.9% of the non-transfusion group. Most bleeding was Grade 2, with no Grade 4 events. In subgroup analysis, the bleeding risk was higher for haematology inpatients than on ICU, and for subclavian CVCs (Table 1). This is the first RCT of platelet transfusions prior to invasive procedures. The authors suggest further research is needed to explore whether we should consider giving multiple platelet transfusions in patients with a platelet count <20 × 109/L. An alternative study design, such as the ongoing T4P study (IRAS ID: 312405) may be required to explore what a safe platelet count is for invasive procedures. This study applies only to CVC insertion, not to other invasive procedures. The authors conclude that haematology patients with platelet counts <30 × 109/L should have a platelet transfusion pre-CVC insertion, and that ICU patients having non-tunnelled lines may be able to omit platelet transfusion. I consider that this study confirms that thrombocytopenia is a risk factor for bleeding post-CVC insertion, and that platelet transfusions may help to mitigate this in some patients, but it does not tell us what the threshold should be for platelet transfusion in order to prevent bleeding. Ultrasound use and patient, operator and clinical site selection remain vital for safe CVC insertion. The author has no competing interests. None.