In this subproject the analysis of S-calcium was used for evaluation of the specified system of internal quality control. Decisions on acceptance or rejection of series were based on a non-human pool with a calcium concentration near the upper reference limit. In the parallel system, specimens from patients and healthy individuals were used together with other non-human pools to disclose changes in accuracy. The results were examined for false rejections and false acceptances. In all the rejections based on a mean rule with 9 controls and a p(fr) of 0.01, the parallel system supported the decisions. On 18 occasions the series were rejected based on a range rule with a p(fr) of 0.05. Ten of the rejections were confirmed by the parallel system, and the remaining eight times of the 165 series were close to the expected five per cent of false rejections. The matrix effect estimated from the human and non-human serum pools was less than 15 per cent of the total variation. This supported the convenient use of non-human control materials in the analysis of S-calcium. The total variation, CV(t), was estimated to be in the range of 0.027 - 0.030. The median of the patient results could be a valuable tool in quality control, whereas the mean of patient results within the reference limits is of no value. In conclusion, the investigation confirmed the validity of a system for internal control of a specified quality.
|Bogserie||Scandinavian Journal of Clinical and Laboratory Investigation|
|Udgave nummer||SUPPL. 172|
|Status||Udgivet - 1. jan. 1984|