Controversy exists concerning the most accurate method for defining diastolic blood pressure in pregnancy. Both disappearance (phase V) and muffling (phase IV) of Korotkoff sounds have been advocated. We previously reported an objective noninvasive method for measuring blood pressure, called K2 analysis, which in nonpregnant subjects was not different from intra-arterial diastolic blood pressure and was more accurate than the auscultatory technique. For determination of the relation of diastolic blood pressure (using K2) in pregnancy with muffling and disappearance of Korotkoff sounds, 58 women (42 hypertensive, 16 normotensive) underwent 556 blood pressure evaluations in the supine position at various stages of pregnancy. K2 analysis was compared with simultaneous auscultation by two observers, A1 (n=461 observations; 364 hypertensive, 97 normotensive) and A2 (n=415; 316 hypertensive, 99 normotensive). Overall, muffling was detected by observer A1 52.9% (244/461) and by observer A2 44.3% (184/415) of the time. When evaluated by clinical classification, muffling was found by both observers to be present less often in the hypertensive group (A1: 47.5%; A2: 37.3%) compared with the normotensive group (A1: 73.2%; A2: 66.7%) (P<.00001). When both observers were present (n=348), they agreed that muffling was present only 112 times. Disappearance of sound was detected by both observers 98.3% (A1: 453/461; A2: 408/415) of the time. Muffling overestimated K2 diastolic pressure by 7 to 10 mm Hg, whereas there was no statistically significant difference between disappearance and K2 diastolic pressure for hypertensive subjects and a 2.5-mm Hg underestimation of KZ diastolic pressure for normotensive subjects. Both observers were able to obtain a greater percentage of readings within 5, 10, and 20 mm Hg of K2 diastolic pressure using disappearance rather than muffling. However, there were individual blood pressure determinations during which sounds were heard significantly below (or above) K2 diastolic blood pressure (7.9% of A1's and 18.9% of A2's disappearance determinations were greater than 10 mm Hg away from K2 diastolic pressure), including measurements in which muffling was closer to K2 diastolic pressure. In the 112 determinations in which muffling and disappearance were detected by both observers, interobserver variability of muffling was approximately equal to disappearance. For both observers, the differences between muffling and disappearance were large (A1: muffling 11.3 mm Hg above disappearance; A2: muffling 9.7 mm Hg above disappearance) and were not significantly affected by hypertensive status. We conclude that on average, the disappearance of sound (phase V) is a more accurate, reliable, and less variable measurement of diastolic blood pressure in pregnancy than muffling (phase IV).
