Summary
A standard clotting system has been analysed according to the average clotting time
and variations due to thrombin, fibrinogen, day-to-day preparation of reagents and
the clotting assay itself. The variations, εT, εF and ετ are reproducible and normally distributed with zero mean.
Transfer of thrombin into a vessel produces adsorption losses at the solution-air
(S/A) and solution-wall (S/W) interfaces. Specific loss at τ = 30 sec = (% loss in
τ) (thrombin volume)/surface area. For clean non-polar surfaces such as polyethylene
or paraffin the average S/W loss of 3.8%-cm (0.77 mg/m2) has the large coefficient of variation of 3.2%. Vessel surfaces may be saturated
by prior conditioning using a thrombin concentration tenfold that required to give
τ = 50 sec. Emptying and immediately refilling now gives an S/W loss of 0.19%-cm (0.04 mg/m2). The specific S/A loss, always present, is 2.7%-cm (0.54 mg/m2). The combined average loss for conditioned vessels is 1.0% at τ = 30 sec and is
reproducible within sτ
2 for single vessels. A set of vessels gives a coefficient of variation of 0.53%.
Stock fibrinogen aliquots are stored frozen at —95° and are thawed and diluted under
standard conditions. If a syringe is used to withdraw clotting tube aliquots, SF2 is significantly increased over that obtained by using pipettes. Since pipettes do
not contribute, the observed coefficient of variation of 1.7% then applies to the
variation between tubes of stock-F. Six different lots of Fraction I gave a coefficient
of variation of 12% between lot numbers.
Experiments revealed a variance between equivalent experiments performed on different
days (coefficient of variation 1.9%). Within this variance pairs of stock reagents
were stable within the maximum limits of ± 0.12% of τ per day. There is no reason
to suspect that either stock reagent has an intrinsic instability. The day-to-day
variance given above arises from non-reproducibilities in the preparation of buffer
diluents. Variances were reduced when buffers were prepared in deaerated water at
room temperature.
The total day-to-day coefficient of variation of fibrinogen of 1.5% and the variation
between fibrinogen lots effectively eliminates this reagent as a reference standard.
Thrombin appeared otherwise : by selecting and reusing reservoirs the maximum total
thrombin day-to-day coefficient of variance is 0.36%. Thus, 95% of the thrombin concentrations
prepared on different days will lie within 0.72% of the true value. By making use
of known purification procedures it is possible to reproduce a stock thrombin reagent.
This would serve as a universal standard.
