[Evaluation of the Coulter STKS hematologic autoanalyzer]

To express the results of an evaluation of the Coulter STKS blood cell counter.

The following data were collected: inaccuracy, carry-over, linearity, comparison with other systems, suspect morphology alarms, stability of calibration, time-stability of samples and working velocity. The protocol recommended by the ICSH was followed for the first four of the above.

Good results were achieved for cell counts and derived parameters. The leucocyte differential count showed fair results for neutrophils, lymphocytes, monocytes and eosinophils; false basophilia was usually associated with the presence of abnormal lymphocytes. No differences were found between adult and paediatric samples. The usefulness of the suspect morphology alarms was also evaluated; they were found fairly sensitive for the presence of stabs, large amounts of atypical lymphocytes and blast cells, and moderately sensitive for erythroblasts and platelet clumps. The predictive value of alarms was found high in any type of pathology, but scarcely valuable when considering the pathology indicated by the alarm, such predictive value being noticeably increased upon associated alarms.

The Coulter STKS cell counter appears as adequate for the haematology laboratory running daily high number of samples with high pathologic rates.