Background Thrombocytopenia is a common finding in patients in the intensive care unit receiving continuous renal replacement therapy (CRRT). It is unknown if the hemofilter itself contributes to the platelet loss. Objective To measure the direct effect of the hemofilter on platelet counts during CRRT. Design Prospective, observational study Setting Intensive care unit of a University hospital Patients Critically ill patients with acute renal failure receiving CRRT. Methods Two samples of blood were drawn simultaneously, pre-filter and post-filter, and analyzed for platelet count. A correction factor was applied to the post-filter platelet count to adjust for the hemoconcentrating effect of net ultrafiltration. Results Forty-eight sets of paired data from 22 patients were studied. There was a small but significant decrease in mean platelet count across the hemofilter. The mean platelet count drop was 2.32 ×109/L (s.e. 1.06, p=0.0487, 95% CI (0.01,4.62)). Blood flow was strongly related to degree of platelet loss, with a decreased loss of 0.07 ×109/L for every ml/min increase in blood flow (p=0.015). There was no overall decrease in concurrently measured red cell counts across the hemofilter. However, there was a machine-specific affect on red cell loss (p<0.0001). The total calculated daily platelet loss across the filter was 625 ×109 cells. Conclusion The hemofilter may contribute to the thrombocytopenia seen during CRRT, by means of either destruction or retention of platelets during passage. This affect appears attenuated by higher blood flows. This information is useful in the assessment of a low platelet count in patients receiving CRRT.

1. The systemic plasminogen activator response has been examined after intravenous infusion of the following peptides related to neurohypophyseal hormones in approximately equimolar dosages into informed, consenting human volunteer subjects: (a) the natural nonapeptides: lysine-and argininevasopressin, oxytocin and arginine-vasotocin, (b) the N-and C-terminal tripeptide fragments of vasopressin and (c) four vasopressin analogues without pressor activity, altered at the jV-terminus, the disulphide bridge and/or sequence position 8 in the C-terminal tripeptide. In addition, angiotensin II and adrenaline were infused.2. It was' observed that some of the cyclic nonapeptides resulted in high and prolonged increases in amounts of plasminogen activator in venous blood, in the following order, both for amplitude and duration: l-desamino-6-monocarba-[8-D-arginine]-vasopressin > l-desamino-[8-D-arginine]-vasopressin i> arginine-vasopressin = lysinevasopressin > JV a -glycyl-glycyl-glycyl-lysine-vasopressin.3. No plasminogen activator responses followed infusions of vasopressin tripeptide fragments, oxy tocin, angiotension II, l-desamino-[8-iV-MeArg]-vasopressin or 9-desglycineamide-lysinevasopressin octapeptide.4. Mole for mole, the four most active substances were approximately two orders of ten * Deceased 29 April, 1977. Correspondence: Dr J. D. Cash, Regional Blood Transfusion Centre, Royal Infirmary, Edinburgh EH3 9HB, Scotland, U.K. more potent than adrenaline by amplitude com parison alone.5. Intra-arterial adrenaline, in one-tenth the systemic dose, stimulated a release of plasminogen activator from the infused local vascular bed only, This did not occur with equivalent doses of arginine-vasopressin and l-desamino-[8-Dargininel-vasopressin.6. It is concluded that plasminogen activator release arising from catecholamine-responding and vasopressin-responding receptors are both molecule-specific and different in anatomical location. The molecular structural requirements for triggering the latter hypothetical receptor type and potential clinical applications are discussed.

Two synthetic analogues of vasopressin, [N-α-triglycyl-8-lysine]-vasopressin (GVP) and [1-deamino-8-D-arginine]-vasopressin (DDAVP), with a pronounced and long-lasting pressor and antidiuretic action respectively, were given to healthy volunteers in order to investigate whether the ACTH-releasing ability of vasopressin, reflected by plasma cortisol increase, is related to its pressor or to its antidiuretic effect. The effecti'veness of 3 different intravenous doses of [8-lysine]-vasopressin (LVP) in causing release of ACTH was also studied, and the responses compared with that caused by LVP given intramuscularly, the usually recommended route of administration. The lowest dose of LVP used in this study, 1 μg (0.25 IU) intravenously, caused a significant change of plasma cortisol. Doses of GVP with a pressor effect comparable to that of 8 μg of LVP did not affect the plasma cortisol level significantly in spite of obvious pressor effects. Nor did DDAVP in doses which should exert a maximal antidiuretic effect (4 and 16 μg) increase the plasma cortisol level. In 2 out of 10 subjects, 40 μg of LVP (10 IU) given intramuscularly failed to give a rise in plasma cortisol exceeding 5μg/100 ml, which is generally required for regarding the response as normal, and also 16 μg (4 IU) of LVP intravenously failed to produce a response reaching this level in 1 out of 12 subjects. It is concluded that the ACTH-releasing ability of vasopressin is neither associated with its pressor nor with its antidiuretic action. The finding that normal subjects can have a low response to vasopressin reduces its value for testing the function of the hypothalamic-pituitary – adrenocortical system.

Reoperation, nephrological and pulmonary problems are strong predictors, obesity and age independent preoperative risk factors for sternal wound complications.