The precision of the U. S. P. Assay for protamine zinc insulin

The precision of the U. S. P. assay of protamine zinc insulin has been tested statistically with the data from duplicate assays in different laboratories of 24 preparations of protamine zinc insulin. Both the mean difference of six periodic bleedings and the trend of these differences with time were significantly greater than the residual or sampling error. The mean differences from two independent assays of the same preparation were not correlated. For these reasons, within the tolerance limits provided by the U. S. P., the mean difference could be attributed to minor variations in laboratory procedure rather than real differences in potency. Since the trend with time was correlated significantly in parallel assays from different laboratories, the trend must represent, at least partially, a real characteristic of a given preparation of protamine zinc insulin. It should appear that the assay for this insulin form is capable of distinguishing finer differences than it is called upon to make in actual practice.