In a recent article in the ARCHIVES, Shuler et al1 discuss a series of tests that were performed on the IVAC model 4200 noninvasive blood pressure monitor (IVAC Corporation, San Diego, Calif) in order to assess this monitor's accuracy. The authors used the Association for the Advancement of Medical Instrumentation (AAMI) SP-10 standard.2 Unfortunately, they chose to use the 1987 version of this standard rather than the current 1992 standard.3 The 1987 recommendation was to use the average of 3 or more recordings for each subject as the unit of comparison. The present, 1992, recommendations provide for the separate analysis of all readings. Pooling of data was eliminated. As one would expect, the use of the AAMI SP-10 standard calls for the following of a strictly prescribed measurement test protocol. Without following this protocol, no valid inferences can be made as to how the device undergoing testing performed. A test protocol calling for the same sphygmomanometer-cuff reference standard conditions when examining both "criterion standard" and "device in question" was not followed. The authors themselves have commented that testing was performed on patients with the use of inappropriate cuff sizes. The results obtained from these tests must be considered inaccurate owing to cuff selection. Thus, no comment as to the accuracy of the device being tested can be made.
Cohen ML, Berson AS, Alpert B, Avoy RP. Evaluation of a Blood Pressure Monitoring Device. Arch Intern Med. 1999;159(4):407. doi: