[Standardization Committee for Haematology. External Quality Assurance Program for General Hematology. Evaluation of the 1994 results].

UNLABELLED

Since 1994, the Standardisation Committee for Haematology publishes yearly the results of its external quality assurance programme (EQAPH), after it has been improved and even integrated in the Health Services of some autonomous communities.

METHODS

Four hundred and seventy-three laboratories take part in EQAPH. The evaluation of the haematological records is carried out every year from the results of two whole blood samples. Two samples of lyophillised plasma are sent every month to participant laboratories in order to assess prothrombin time (PT), partial thromboplastin time (PTT) and fibrinogen. Samples are sent every three months for antithrombin III (AT-III) determination. According to the types of autoanalysers used, 9 groups have been established: Group I (Technicon H* and H-6000), group II (Technicon H2 and H3), group III (Coulter MaxM, STKS, Cobas Argos and Cell-Dyn 3000/3500), group IV (Coulter STKR), group V (Coulter JS/JT), group VI (Coulter S/T, Cobas Helios/Minos), group VII (Sysmex E/NE), group VIII (Sysmex K-1000), and group IX (other semi-automated counters). Three groups are established for general coagulation tests and two others for AT-III. The value of the mean of all results (consensus mean) and of each particular group (group mean) are used as statistical methods. The results are expressed as: (1) coefficient of variation, % (CV); (2) deviation index (DI); DI values attained with respect to the same group (or method) or all groups (or methods) allowed us to classify the results in four categories: excellent (0 < DI < 0.5), good (0.5 < DI < 1.0), satisfactory (1.0 < DI < 2.0), and out of acceptable limits (ID > 2.0); (3) Youden graphs or graphic representations of DI calculated from the analysis of the control specimens.

RESULTS

Group I yielded lesser values for leucocyte count (0.221) and higher for MCH (0.833). Group II showed high DI values for PCV (0.933) and MCV (1.146). Group III had lower values for red cell count (0.097), haemoglobin concentration (0.133) and MCH (0.91). Group IV showed lower platelet count. Group V had higher haemoglobin concentration values. Group VI yielded higher DI in the platelet count and group VII in the white cell count. Group VII showed the lowest DI for MCV and MHC. Group IX had high values for red cell count and MCH. In the coagulation field, group I had higher values for PT and AT-III. Group II showed higher DI values for PTT and fibrinogen. The highest CV values were seen in groups VIII and IX. The lowest values were present in group II for haemoglobin, IV for MCH, in V for PCV and MCV, in VII for red cell count and MCH, in VII for white cell count and in VII for platelet count. The coagulation tests showed lower CV values than cytometry, the lowest being for PT in group I and PTT and fibrinogenein group II. With regard to the influence of acceptable results on adverse ones within this group, the percentage of laboratories with mean DI over 2 were calculated. Thus, the laboratories achieved 43.5% of values within acceptable limits for platelet count and 56.7% for white cell count. Regarding the PCV, out of 47.9% of the laboratories, only 1.0% showed high deviation of some results. In the coagulation parameters, of the 37.2 in this group for AT-III, 32.4% showed a mean DI over 2.

CONCLUSIONS

Participation in External Quality Assurance programmes contributes to the comparability of the results provided by different laboratories, thus increasing their accuracy and improving the quality of patient care.