MedSysEvidence-graded reference

MedSys / Heart & cholesterol / Measuring blood pressure

Measuring blood pressure properly

Most of the error in a blood pressure reading is put there before the cuff inflates — and the commonest mistake of all is measuring one arm.

What this tool does and does not do. It interprets one set of readings — the difference between your arms, and if you have ankle readings, the ankle-brachial index. It does not tell you whether your blood pressure is controlled or whether your medication is working; that needs readings across a day alongside your drugs and doses, and it is a different tool.

1. The errors, and what each one costs

These are not rounding errors. Several of them are the size of the difference between a normal reading and a diagnosis.

ErrorRoughly what it adds
Cuff over clothingUp to +50 mmHg in the worst case, and unpredictable. Bare arm, and do not roll a sleeve into a tourniquet above the cuff
Cuff too small+2 to +10 mmHg, and larger arms are affected most. The bladder should wrap about 80% of the arm. This is the commonest equipment error
Back unsupported+6 to +10 mmHg diastolic
Legs crossed+2 to +8 mmHg systolic
Feet dangling+6 mmHg. Flat on the floor
Arm unsupported, or below heart level+10 mmHg or more — holding the arm up is muscular work, and an arm below heart level reads high by hydrostatic pressure alone
Talking, or being talked to+10 mmHg
Full bladder+10 mmHg
Coffee, nicotine or exercise in the previous 30 minutesVariable and often substantial
No rest beforeSit quietly for five minutes. This is the instruction most often skipped and it is not optional

Stack three or four of those — which is easy — and you have manufactured a diagnosis. This is a large part of why home readings, taken properly, are more reliable than a rushed clinic reading, and why NICE prefers home or ambulatory measurement for diagnosis.

2. Doing it properly

  1. Set up before you start

    Empty bladder. No coffee, cigarette or exercise for 30 minutes. Upright chair with a back, feet flat, legs uncrossed, bare arm resting on a table so the cuff is level with your heart. Then sit still and quiet for five minutes.

  2. The first time, measure both arms

    This is the step almost everyone skips, including clinicians. Roughly one person in ten has a systolic difference of 10 mmHg or more, and using the lower arm moves about one in twelve people from above a diagnostic threshold to below it. Whichever arm reads higher is the arm you use from then on — not the convenient one, and not an average.

  3. Three readings, one minute apart

    Discard the first and average the second and third. The first is nearly always the highest. Most home monitors will do this for you if you let them.

  4. Twice a day for seven days, for a diagnosis

    Morning and evening, before food and before medication. Discard the whole of day one and average the rest — that is what NICE asks for from home monitoring.

  5. Use a validated upper-arm monitor

    The British and Irish Hypertension Society publishes a list of validated devices, and unvalidated ones are widely sold. Wrist monitors are far more position-sensitive and are a second choice at best. Check the cuff size against your arm circumference rather than assuming the one in the box fits.

3. Why both arms, specifically

  • A systolic difference of 10 mmHg or more is generally taken as the upper limit of normal, and is found in about 3.6% of the general adult population, 7.4% of people with diabetes and 11.2% of people with hypertension.
  • It carries information beyond the reading itself. In an individual participant data meta-analysis pooling more than 50,000 people from 24 cohorts with ten years of follow-up — the INTERPRESS-IPD collaboration — a difference of 10 mmHg or more was associated with higher cardiovascular events, cardiovascular mortality and all-cause mortality independently of standard risk scores.
  • NICE and the ESC act at 15 mmHg, treating a persistent difference at that level as a marker of additional cardiovascular risk and a reason to look for arterial disease on the lower side.
  • Confirm before concluding. Sequential readings exaggerate differences, because blood pressure falls with repetition — measure both arms simultaneously if you can, or alternate and repeat.

4. Arms and ankles — the ankle-brachial index

Comparing ankle pressure with arm pressure screens for narrowing in the leg arteries. Its value is less about the legs than about what it implies: peripheral arterial disease is a strong marker of disease in the coronary and cerebral arteries too, and finding it usually changes how aggressively cholesterol and antiplatelet treatment are pursued.

ABIInterpretation
Above 1.40Non-compressible arteries — not a good result. Usually stiff, calcified vessels, common in long-standing diabetes and kidney disease. The test is uninterpretable and the finding itself carries raised risk
1.00 – 1.40Normal
0.91 – 0.99Borderline
0.90 or belowPeripheral arterial disease
Below 0.40Severe. With rest pain or ulceration, urgent
Two things about ABI that get missed. A high number is not a good number — above 1.40 means the arteries would not compress, which is the opposite of reassuring, and it is the commonest misreading of this test. And an automatic upper-arm monitor is not validated at the ankle: a real ABI uses a manual cuff and a hand-held Doppler, taking the higher of the dorsalis pedis and posterior tibial pressures in each leg. A home attempt is a reason to ask, never a diagnosis and never an all-clear.

5. Sitting, then standing — for light-headedness

If you feel light-headed, unsteady or faint when you stand up, the measurement that answers it is a sitting reading followed by standing ones, and it is the third mode in the tool above. It is simple, it is under-done, and it is the single most useful thing on this page for anyone on blood pressure treatment who has been putting the symptom down to age.

  1. Sit quietly for five minutes, then measure

    Same arm throughout, cuff at heart level.

  2. Stand up, and measure again at one minute

    Have a chair behind you. If you feel faint, sit down — the test is not worth a fall.

  3. Measure again at three minutes

    This is the step almost everyone skips, and it is where a good proportion of cases are found. Stopping at one minute misses delayed orthostatic hypotension entirely.

A sustained fall of 20 mmHg systolic or 10 mmHg diastolic within three minutes of standing is the accepted definition of orthostatic hypotension. It affects roughly one in five independently living adults over 60 and around half of care home residents, and it is a fall risk before it is anything else.

Two cautions. A negative test in someone with real symptoms is not an answer — the drop varies with time of day, meals, heat and hydration, so repeat it in the morning and after eating, which is when it is most often positive. And a positive test is not a diagnosis of a cause: medication is a common one, but so are dehydration, anaemia, rhythm problems and neurological conditions. Nothing here is a reason to stop a tablet without your prescriber — and if you are on treatment, the treatment review tool is the companion to this.

6. When a difference between limbs is urgent

Almost never on its own. The exception is a large inter-arm difference appearing suddenly alongside severe chest, back or abdominal pain — classically tearing, often worst at onset — which can occur in acute aortic dissection. That is a 999 call, and worth saying explicitly because the combination is easy to rationalise as anxiety. It is uncommon; the point is that it is not something to measure repeatedly at home while deciding.

Everything about what the numbers mean — the bands, the targets, the drugs and their side effects — is in the blood pressure chapter. This page is only about getting a number worth interpreting.