Measuring a person's blood pressure has been a mainstay of medical treatment for over a century, but the way we currently go about it is not always accurate or consistent. We make a lot of assumptions, after all, that one reading at one site can tell us what's going on throughout the whole body.

A small new study now suggests that could be a fatal flaw. In a neuroscience intensive care unit (NSICU) blood pressure readings show marked differences depending on where they were taken and how.

Even in the same patient, blood pressure readings taken on opposite sides or at different places on the leg or arm could yield significantly different results, and this extreme variability might be putting already vulnerable lives at risk.

"For our patients in the NSICU, blood pressure often needs to be maintained in a very narrow range," says neurocritical care nurse and researcher Kathrina Siaron.

"Moving it one way or another could potentially harm the patient."

For decades now, physicians have largely assumed there's little difference in non-invasive blood pressure measurements taken at different points on a person's body. No matter where you cuff it, on the upper arm, wrist or thigh, the readings will lie in the same ball park.

Even when more invasive methods are used, like inserted catheters that read arterial blood pressure at great accuracy, the readings are predicted to generally match blood pressure in the upper arm. But is that really right?

To test these various assumptions, researchers at the University of Texas Southwestern worked with 80 patients in the NSICU, split almost equally between men and women who were around the age of 53.

These patients had been admitted to the unit for several serious neurological problems, including stroke, subarachnoid hemorrhages, and brain tumours.

Sitting upright in their hospital beds, the patients had their blood pressure taken on both upper arms and then on their wrists. At the same time, 29 of these patients were also given arterial blood pressure sensors.

Usually, the differences between blood pressure sites don't differ by more than a few points, but in some patients, the authors found dramatic 40 point discrepancies that could radically impact how the patient is cared for.

In total, the range of mean differences between systolic blood pressure readings - from site to site and method to method - were all 10 to 11 points apart.

"This is significant since even 5 mmHg on the SBP is enough to make a difference when deciding to intervene or withhold treatment," the authors write.

Comparing the systolic pressure in both left and right upper arms, the authors found an average difference of 8 points, while diastolic measures varied by 6 points.

"If we take pressure in one arm, a patient seems fine, but in the other arm, they're in a crisis," says neurologist DaiWai Olson.

"The values we collected were really all over the place."

And that's also true for the same arm and wrist.The upper arm had consistently lower measurements than the wrist, which, the authors argue, is probably because the wrist's tendons and bones are less reactive to external pressure than a muscle.

Whatever the cause, these two sites differed on average by up to 13 points in systolic pressure and about 5 points in diastolic measures. The authors say this "consistent mismatch" is worth considering in the clinical setting and in practice. In other words, avoid wrist cuffs if at all possible.

The authors aren't sure why these differences exist - blood pressure exists in a complicated web of physiological processes - only that their results indicate blood pressure is not the same throughout the body at any given time.

Even arterial pressure, which is considered a gold standard in medicine, was found to vary from the cuff measurements by as much as 15 points give or take.

Of course, the sample size in this study is quite small and only among a niche subset of patients. Nevertheless, previous research has also suggested blood pressure differs between the left and right arms.

"Therefore," the authors conclude, "location does matter when measuring BP; however, this study cannot conclusively recommend which site to use above others."

Further investigation is needed through larger trials, testing different sites of the body and blood pressure methods, but if the discrepancies keep popping up we might need to change our protocols.

Perhaps, for instance, taking several readings at different sites around the body and averaging them is a better way to predict how our blood is actually pumping at any given moment.

The study was published in Scientific Reports.