Automated blood pressure measurement may not improve efficiency if manual technique was suboptimal

Dr. Yarows makes a compelling case that primary care offices should replace manual sphygmomanometers with automated devices.1 We agree with this goal, but believe that he may overstate the likelihood that automated devices will lead to lower BP readings, or that medical offices will realize the time savings. His case is based on two principles: (1) oscillometric BP readings are often lower than those obtained manually, and (2) a health-care worker can obtain BP readings faster with automated devices than he or she could using auscultatory technique. That oscillometric BP readings are lower than those measured with auscultatory technique is well-documented, but it should be noted that studies often compare automated devices with research-quality manual techniques.2 Automated devices are also clearly faster than the auscultatory technique when performed correctly.3 Thus, one might expect primary care offices would improve hypertension control rates and patient flow simply by using automated devices. However, in our ongoing study involving 30 primary care offices,4 we have found that real-world health-care workers often do not perform the auscultatory technique correctly; specifically, many deflate the BP cuff too quickly. Therefore, some offices that adopt automated devices have seen hypertension control rates worsen, and experienced no or negative effect on efficiency. Guidelines recommend deflating BP cuffs no faster than 2-3 mm Hg/s.5 The underlying reason for this is, as the needle on the sphygmomanometer gauge descends, each tic mark it passes must be linked with a heartbeat in order to ensure that the needle is pointing at the correct BP when the 1st and 5th Korotkoff sounds appear and disappear. If a patient has a true BP of 140/70 mm Hg and pulse of 60 bpm, a health-care worker would have to allocate 60 seconds simply to deflate the blood pressure cuff from the recommended 30 mm Hg above the minimum inflation level to a level that ensures the physician does not miss an auscultatory gap (170-50 mm Hg). Deflating a BP cuff faster than this significantly increases the potential that one underestimates a patient's “true” BP.6 Dr. Yarows reported needing an average 58.6 seconds to perform all aspects of a single manual measurement, including cuff inflation and deflation. Thus, we were unsurprised to read that BP measurements in his practice were 10/6 mm Hg higher with the oscillometric technique. In our work, we have found that health-care workers deflate BP cuffs far faster, with average time from maximum cuff inflation to deflation completed in 13.2 seconds. As a result, several PCPs in our study have complained that the automated devices display BP readings higher than their patients' “usual” readings. We fully endorse widespread adoption of automated BP measurement in primary care, and indeed our experience suggests that quality metrics like the Health-care Effectiveness Data and Information Set (HEDIS) measures for hypertension control should only be considered valid if measured with an oscillometric device. But, we also share our experience because we made a case similar to Dr. Yarows's when introducing automated devices into our primary care offices. Subsequently, the difference between expected and actual effects on BP readings has led some PCPs to perceive automated devices as unreliable and invalid, and to reject them in favor of returning to manual technique. Were we to introduce automated devices into primary care settings again, we would describe the case made by Dr. Yarows, but also emphasize the potential disservice done to patients when their cardiovascular risk is misclassified due to sloppy manual technique. We hope, in doing so, that we would enlist PCPs in our efforts with more appropriate expectations. The authors declare that they have no competing interests. NHLBI/PCORI, contract number: UH3HL130688