Estimation and predictive use of the corrected count increment--a proposed clinical guideline

Platelets are frequently given inappropriately. The accepted indication is symptomatic thrombocytopaenia due to bone marrow failure. In contrast replacement therapy is contraindicated in immune mediated peripheral sequestration because it is almost always ineffective and also aggravates the already rapid rate of the clearance due to superimposed isoimmunisation. Furthermore circulating levels, rather than the clinical situation, customarily trigger the request and would typically be at 20 x 10(9)/l. Usage is under evaluated and, accordingly, current practice has been scrutinised and a guideline proposed that is applicable to user and vendor alike.

Prospectively a consecutive series of eligible cases were selected and, with informed consent, all relevant information recorded. Specific observations were haematologic status, quality of the product and the impact of previous exposure to blood fractions on outcome. Additionally, factors that influence response were documented and included disseminated intravascular coagulation, splenomegaly and concurrent use of intravenous Amphotericin B. The absolute numbers infused, the increase in peripheral value recorded and this ratio which is designated as corrected count index generated in each instance.

The CCI was computed for each of the 85 megaunits given to 29 individuals and subsequently estimated for the entire sample population using regression analysis. Each of the variables was used to test the hypothesis that such a figure might differ depending on patient sub-population. This approach was accurate in 60% of the time and predicates a more rational use of donor characteristics. The cardinal measurement is that supplied by the transfusion service which can be used as a basis for clinically important predictors of anticipated benefit.

This data, in keeping with international practice, leads to the recommendation that commercial and other services routinely measure and specify absolute numbers. The clinicians, reciprocally, should always confirm this figure and then match anticipated to observed outcome. Failure to carry out these simple procedures lead to gross over utilisation of this expensive intervention, risks unnecessary sensitisation and furthers persistence with sub-optimum administration.