Do not just add new platelets, as these will undergo activation and apoptosis as well. I'm working on this problem. There is no drug on the market yet, that is allowed to use for this purpose.
The decrease in platelet count during ECMO may depend on a number of factors (e.g. baseline platelet count, bleeding during ECMO, temperature, ECMO duration..). In my experience, a correct management of the balance between pro-coagulation and anti-coagulation will result in bleeding control, thus eliminating one of the crucial factors for this drop down. We transfuse plt pools when the count is < 40000 with active bleeding during VA-ECMO
As said there are a lot of factor playing a role in platelet count decrease. If i can focus on two elements in the ECMO context i'll suggest to consider firstly an heprain-induced thrombocytopenia and secondly some clotting in the oxygenator membrane. Concerning the first point you have to consider that circuits are, most frequently, heparin-coated. It has been published recently by Pabst D et al (DOI: 10.1177/0267659119842056) that discountinuation of systemic heparin could be sufficient in this setting. About filter, each time you have signs of coagulopathy or thrombocytopenia an inspection of your filter could answer a lot of questions. An agressive policy of filter changing in this setting could (partially) relieve these problems. We consider to trasnfuse platelets if count is under 50.000 or under 100.000 with some bleeding.