[External assessment of analytic quality in hematology: a necessity in Latin America]

The assurance of analytical quality in a clinical laboratory is achieved through an internal system of quality control complemented by an external evaluation program. Quality assurance provides a foundation for the confidence that is placed in laboratory results and their use in the diagnosis and treatment of diseases. Many laboratories in Latin American countries do not have appropriate systems in place to evaluate and control quality. Given the importance of diagnoses based on hematologic data, the Pan American Health Organization sponsored a course in quality control in hematology during the XI Latin American Congress of Clinical Biochemistry (Mexico, 1993), in which representatives from Argentina, Chile, Cuba, Mexico, Paraguay, Dominican Republic, and Uruguay participated. As part of the course, the following control materials were produced: secondary standard solution of cyanmethemoglobin, stabilized concentrated hemoglobin solution, and preserved human whole blood with pseudoleukocytes. These materials were sent to laboratories in the seven participating countries for use in analytical procedures, and the results were then subjected to an external evaluation to assess individual performance and the comparability of results among the group. The specific tasks carried out were: (1) determination of values for hemoglobin, hematocrit, and red and white blood cell counts by the procedures normally used in each laboratory; (2) recording of the data on special reporting forms; and (3) transmittal of those forms to the coordinator in each country. The results were analyzed with regard to both the procedure used and the participating country. Reference values were established by consensus following application of a statistical method to eliminate outlying values. Comparative analysis of the results showed the coefficients of variation (CV) of the hematocrit (4.5%), red blood cell count (11.0%), and white blood cell count (22.2%) to be higher than those reported from the United States of America and Europe. With regard to analytical procedures, the manual methods yielded larger CV than the automated methods. When analysis of variance (ANOVA) was used on data broken down by country and by procedure, the only statistically significant result was for leukocyte count (P < 0.02). It was concluded that training in the preparation of quality control materials and the subsequent use of these materials in pilot surveys could provide a starting point for establishing continuous internal and external quality assessment systems in hematology. Such systems, together with continuing education for laboratory personnel and the availability of automated instrumentation, will lead to achievement of optimum laboratory quality.