An easier alternative to orthogonal regression for calculation of International Sensitivity Indexes

To evaluate an easier alternative method to orthogonal regression analysis for calculating International Sensitivity Indexes (ISI).

ISI for 18 reagents were estimated from reference and test reagent prothrombin times using plasma from 60 stabilised patients undergoing anticoagulation therapy and 20 normal subjects. ISI were also derived for 12 systems (instrument/reagent combination) using lyophilised plasma calibrants. Orthogonal regression and the easier alternative log ratio method were evaluated by comparing resultant International Normalised Ratios (INR) for 58 patients using two test systems.

For the reagent calibrations, the differences in the two methods for the sensitivity slopes were very small. For the system calibrations, slope differences were still of little clinical importance. Parallel observations for INR on the 58 patients confirmed that the bias introduced by easier log ratio derivation of ISI was small and of minor clinical importance, although the bias increases for high INR.

The easier method for ISI determination is a useful alternative to orthogonal regression analysis, particularly when computer assistance is not available, for checking for gross errors in computer computation and for use when calculating ISI from INR calibrants.