[Standardization of the methods of lipid determination according to WHO in the regional project of prevention of cardiovascular diseases in Friuli-Venezia Giulia].

Background: The recognition of dyslipidemias as a major modifiable risk factor for atherosclerosis and coronary heart disease underlines the need to obtain precise and accurate assay results of plasma lipids. Today the use of automatic laboratory methods and of internal quality control favours the precision of the results but does not guarantee accuracy. The efficiency of a laboratory can be ensured by a standardization programme, systematically monitoring precision and accuracy by means of independent internal and external quality control, international reference standards (e.g. those of CDC-NHLBI and WHO) and protocols to identify and reduce the errors due to biological variability and pre-analytical factors. After the foundation of the Regional Project for Prevention of Cardiovascular Diseases in Friuli-Venezia Giulia, a lipid standardization programme was set-up, covering the 20 chemico-clinical laboratories of the Region. The programme was directed by the International WHO-MONICA-Lipid Reference Centres of Prague-Udine.

Methods: During the years 1993-1994, three sets of lyophilized human serum samples were dispatched to each laboratory for the blind evaluation of total cholesterol, triglycerides and HDL-cholesterol. The samples were obtained by the combination of three serum pools at least at different lipid concentration. The first set included 20 samples to be tested in 5 weeks, the second set included 30 samples to be tested in 8 weeks and the third set included 21 samples to be tested in 9 weeks. The assay results were sent to the Prague-Udine WHO-MONICA Centres where they were computerized and evaluated, particularly considering precision for each set, estimated by the variation coefficient (i.e. standard deviation/mean value of the measurements per cent) and accuracy (the bias was computed as mean of the measurement minus the reference value/reference value per cent).

Results: In the three assay series for total cholesterol, almost all the laboratories showed the variation co-efficient (precision) to be less than the WHO-MONICA limit of 3.7% (for a cholesterol level of 250 mg) (Tab. II) and in 8 cases out of 20, less than the CDC limit of 3%; the accuracy bias was less than the WHO-MONICA limit of 5% in 17 laboratories out of 20 and less than the CDC limit of 3% in 11 cases out of 20. For the HDL-cholesterol standardization programme the reference values were based upon the phosphotungstate method. However, the pools were also controlled by the other precipitation methods used in the 20 participating laboratories: 11 laboratories worked within the WHO-MONICA limits of precision and accuracy (respectively 6.5% and 7.5%) in at least two of the three sets. Concerning triglycerides, the regional laboratories showed a greater variability and, though most of the variation coefficients were within the WHO-MONICA limit of 5%, half of the accuracy biases were greater than the limit of 10%. The bias of the measurement average of all the laboratories was excellent for total and HDL-cholesterol, not quite good but acceptable for triglycerides. Laboratory performance improved progressively from the first to the last set, on more than one occasion.

Conclusions: The lipid standardization experience carried out in the framework of the Regional Project for Prevention of Cardiovascular Disease demonstrates that it is possible to set up a wide and co-ordinated collaboration with laboratories of an entire region with positive and improving results. For this global quality control system, the resource allocated is limited but widely rewarded by the community benefits in terms of assay reliability and savings at medical care level, basic research and population studies.