[A tool to predict the resources necessary for the whole hemodialysis population]

Predicting the resources necessary for the treatment of terminal renal insufficiency requires an understanding of the needs of the whole dialysis population. This study evaluates the advantages of a complete evaluation grid for care needs (CG, 47 items) compared with a simplified version (SG) and with the data obtained from REIN (R) as a tool for predicting the required resources. Compared to CG, the two other classifications under-estimate the level of care needed and the prevalence of patients who have at least one condition resulting in an 'excess' of needs. In a system with three types of structures, the theoretical distribution of patients according to the CG is: in structure A (permanent medical presence): 43.4% (vs 39 for SG and 21% for R); in structure B (intermediate): 34% (vs 31.5 and 24.7% respectively); and in structure C (independent): 22.3% (vs 29.5 and 53.5%) of patients. The care needs scores of populations assigned by R to home dialysis are incompatible with the resources allocated to it. With a certain number of adjustments and qualifications, a population study using a simplified needs grid could enable a prediction of the resources which need to be provided. The patients' records which take precedence in the future REIN should be modified in order to make possible a regular evaluation of the appropriateness of the resources allocated.