A 54-year-old man presented for his first office visit looking for an alternative to prescription medications to treat his newly diagnosed hypertension. His blood pressure had always been normal in the past, but on a routine annual check-up with his internist two weeks previously he was found to have a blood pressure of 158/96 mm Hg. He had been advised to begin taking a beta-blocker and to have his blood pressure rechecked in two weeks. The man had not yet started the antihypertensive medication at the time of his consultation at my office.

Further questioning revealed that the man's blood pressure had been taken in a somewhat noisy environment, with no time taken to get calm, while casually chatting with the doctor, and with his arm resting on the arm of a chair. When I checked the patient's blood pressure after allowing him a quiet period in a dimly lit room, with no talking allowed during the measurement, and with the arm at heart level, it was 126/78 mm Hg. He was asked to return for a repeat measurement in two weeks, at which time it was again 126/78.

Comment: Inaccurate blood pressure measurements, which are all too common in clinical medicine, can lead to potentially dangerous errors in patient management. Proper technique is crucial for obtaining an accurate measurement. The cuff size should be based on the circumference of the midpoint of the upper arm: for 22 to 26.9 cm, use small adult size; for 27 to 32.9 cm, use adult size; and for 33 to 40.9 cm, use large adult size. The patient should be seated in a straight-back chair with both feet on the floor (when patients are sitting on an examining table, the contraction of muscles required to hold their position can increase blood pressure by 10/5 mm Hg. The arm should be at heart level; if it is allowed to hang lower, then the blood pressure reading will be falsely elevated. In addition, blood pressure can rise substantially if the patient is talking during the measurement. Two or more measurements, separated by two minutes, should be taken. If they differ by more than 5 mm Hg, additional measurements should be taken and averaged.

Worcester S. Blood pressure measurement frequently lacks accuracy. Fam Pract News 2004(September 1):16.