Comparison of different methods of blood pressure measurements

Abstract

Objective: To compare different methods of measuring and averaging blood pressure readings, and see how they affect classification of patients as normotensive or hypertensive.

Methods: The comparisons were made in up to 145 suitable patients in the first and last week of a 6-week surveillance. Ambulatory blood pressure measurements were taken from the nondominant arm, as the average of up to 78 measurements over 24 h (24-h ambulatory blood pressure measurements), or 60 from 0700 h to 2200 h (daytime ambulatory blood pressure measurements), or 18 from 2200 h to 0700 h (night-time ambulatory blood pressure measurements). Office blood pressure measurements were taken by the outpatient department nurse in triplicate from both arms, and the averages were taken of the second and third of each triplicate. Home blood pressure measurements were taken in duplicate each morning and evening, entered by patients into diaries, and the available readings of up to four values per day were averaged for the first and sixth week. The classifications of normotensive versus hypertensive produced by criterion values of 24-h ambulatory blood pressure measurements of 125/80 and 130/80 mmHg were compared with the classifications produced by the other measurements with a range of criterion values.

Results: The home blood pressure measurements and office blood pressure measurements of systolic values underestimated the corresponding 24-h ambulatory blood pressure measurements values by 3-9 average (SD 9-18) mmHg, and the diastolic values overestimated them by averages of 3-6 (SD 6-13) mmHg. Daytime ambulatory blood pressure measurement systolic and diastolic values overestimated them by 2-4 (SD 2-4) mmHg and night-time ambulatory blood pressure measurement values underestimated them by 7-12 (SD 5-9) mmHg. In comparing the 24-h ambulatory blood pressure measurement classifications of hypertensives versus normotensives with those produced from the other types of measurements, it was easiest to detect criterion values for daytime ambulatory blood pressure measurement that gave the best agreement and they appeared different for the different criterion values of 24-h ambulatory blood pressure measurement. For the other types of measurement, the agreement was generally worse and it was harder to detect a best criterion value for agreement with either of the 24-h ambulatory blood pressure measurement classifications. A subsample of 63 patients identified as needing institution, maintenance or modification of antihypertensive treatment excluded about half of the patients classified as hypertensive by either 24-h ambulatory blood pressure measurement criterion.

Conclusion: Limited agreement existed between different ways of classifying patients. The utility of the classifications depends on the purpose to which they are put.