The influence of the method of data analysis on the reported accuracy of automated blood pressure measuring devices

Objective To show that different methods of data analysis affect the grading that blood pressure measuring devices achieve according to the British Hypertension Society (BHS)-protocol. Methods Based on the somewhat unclear description of the exact method of data analysis in the BHS-protocol four different methods can be discerned. The effect on the grading-results is calculated for these four different options. Results and Conclusions It is shown that using these four different options the achieved grade can range for diastolic blood pressure from C (option 1) to almost A (option 4) and for systolic blood pressure from D (option 1) to B (option 4). Different researchers may well have used different methods. Option 1 is the method that should be used. Also it is stated that the systematic error and the standard deviation of differences (SDD) are measures that give more insight to describe a device's performance. Calculating the grades after correction for the systematic error shows its influence and that of the SDD on the reported accuracy of a blood pressure measuring device. (C) 2001 Lippincott Williams & Wilkins.