Fibrinogen determination in a series of proficiency studies

Fibrinogen estimations were performed in four quality control proficiency studies conducted by the National (UK) Reference Laboratory for Anticoagulant Reagents and Control for over five hundred hospitals in the UK and overseas during 1976--1978, to evaluate the various techniques in current practice. More than 30 different estimation techniques were employed by participants and these could be classified into eight groups. The success in detecting a fibrinogen level below the normal range ('sensitivity') and reliability in providing a normal result for a value within the normal range ('specificity') have been determined. No method failed in both aspects. The tyrosine, clot weight, immunological and Clauss technique appeared reasonably satisfactory. The best results were obtained with the Clauss technique although the difference between this and the other quantitative techniques did not achieve statistical significance. Some were less dependable, notably the turbidity techniques and the thrombin time. Although no technique appeared to be superior to others from the standpoint of precision, turbidity techniques appeared to be the least precise showing the largest co-efficient of variation.