Cost-minimising analysis of acute normovolaemic haemodilution and hypervolaemic haemodilution in saving red blood cells as compared to transfusion ofallogeneic blood

Background: Analysing cost data with respect to savings of red blood cells (rbc) by acute normovolaemic haemodilution (ANH) / hypervolaemic haemodilut ion (HHD) versus purchasing homologous packed rbe (HPRBC). Methods: Using o ur previously published model for ANH/HHD we calculated saving of rbc mass for various initial (45 and 40%) and minimal (haematocrit) hct levels (27, 24, 21 and 18%). ANH was performed by isovolaemic exchange of 4 units of bl ood (each 500 ml) versus colloid with an intravascular volume effect of 1.0 . Intra-operative loss of colloid was only due to surgical blood loss. ANH units were re-transfused in reverse order; thereby maintaining isovolaemia and minimal hct level with ongoing blood loss by additional infusion of col oid. HHD was performed by pre-operative infusion of 1,000 ml colloid; assum ed intravascular volume effect is 1.0. Hypervolaemia during surgical blood loss was maintained by infusion of colloid until reaching minimal hct. Elim ination of the excessive colloid from the intravascular space lead to rise in hct (from minimal to final hct). This difference between final and minim al hct results in saving of rbc mass. Cost data refer to both hospital vari able/acquisition cost and data given in DKG-NT (variable/acquisition cost ( 'Sachkosten') and total cost ('Gesamtkosten')). Results: ANH at best allows for rbc mass savings of approximately 2 units (each 190 ml of rbc). Under less ideal conditions (initial hct less than or equal to 40%, minimal hct g reater than or equal to 21%), rbc savings amount to approximately 1 unit. W ith HHD, the corresponding rbe mass saved amounts up to 1 unit at best. Our cost-minimising analysis (CMA) shows that these rbe savings by ANH/HHD are less expensive than purchasing equal amounts of allogeneic rbc mass. With respect to the underlying cost base, these 2 units are between 25% (total c ost according to DKG-NT: ANH EUR 166.5 versus HPRBC EUR 221.7) and 60% (hos pital variable cost: ANH EUR 61.7 versus HPRBC EUR 157.2) cheaper than corr esponding units of HPRBC. rbc mass saved by HHD is even less expensive (tot al cost according to DKG-NT: HHD EUR 63.8 versus HPRBC EUR 135.1; hospital variable cost EUR 32.1 versus;HPRBC EUR 95.2). Thus, they are approximately 53% (total cost according to DKG-NT) to 62% (hospital variable cost) less expensive than HPRBC. Conclusions: Both ANH and HHD - under ideal condition s and within their limited extent - are more cost-effective in saving rbe t han transfusion of allogeneic blood. However, both the efficacy of these me asures and the resulting financial savings are limited at best to 2 units o f rbe for ANH and 1 unit for HHD.