Is there something up your sleeve?

It is not necessary to emphasize for the readers of the Journal of Hypertension the importance of hypertension and its treatment. However, as we all know, measurement of blood pressure (BP) is of paramount importance, but frequently is ill performed. As a matter of fact, outside the research setting, proper BP measurements are probably the exception rather than the rule [1,2]. There are many reasons for that, BP measurement is not at all an obvious procedure and requires many steps to be taken both by the examinee and the examiner, these steps are not followed by many examiners for a variety of reasons, ignorance of the proper procedure not the least of them [2]. Previous [3] and current recommendations [2,4] for BP measurement stress the need for use of a proper sized cuff placement over the midst of the upper arm about 2.5 cm above the antecubital fossa. Nevertheless, there are many circumstances, which do not allow straightforward implementation of this directive. People do not come to clinics dressed in examination gowns, and depending on weather and social standards, they do not always come in short sleeves. Moreover, in different cultural groups of people, mainly women, may find it embarrassing, or even improper to expose their arm not to mention taking off their dress to free it… On top of it, taking off a coat, hanging it up, taking off one's dress or pulling one's arm out of it takes time, and guidelines require at least 5 min rest afterwards. Given that average primary care consultation may be rather short, ranging from 22 min in Sweden to less than 2 min in Bangladesh, and is somewhat inversely related to country's wealth and medical workforce, even in most European Union countries it is 10 min art best [5]. We are supposed to have the patient seated quietly for 5 min, then have BP measured several times, on first visit in both arms, without talking and sometimes also after one up to 3 min standing, that leaves very little time for important things like hearing medical history, discussion and sometime arguments. Therefore, the issue of the necessity to place the BP cuff over a sleeve or the bare arm is very relevant issue for everyday practice. Guidelines had emphasized to ‘Avoid rolling up the sleeve in such a manner that it forms a tight tourniquet around the upper arm’ [3,4]. Notwithstanding the tourniquet effect of a rolled-up sleeve, there are also other ways a rolled-up sleeve affects BP measurement. Unless the examinee has an unusually long forearm, it is not possible to place the cuff in the midpoint of the forearm and leave 2.5 cm above the antecubital fossa, so it is either placed too low, below the antecubital fossa, or over the rolled-up sleeve. The first position violates BP measurement guidelines and the latter may cause both miscuffing (the part over the sleeve is too short), requiring higher cuff pressure to occlude the brachial artery, as well as uneven distribution of oscillations that may obscure correct mean BP when measured with an oscillometric instrument (as most of use these days). Although two excellent reviews from 1995 [6], and 2017 [7] on inaccuracy in BP measurement, stated that BP measurement over a thin sleeve does not affect the measurement accuracy, these statements were supported by relatively little data. In recent years, much of the attention regarding BP measurement focused on issues, such as ‘white-coat hypertension’ and ‘masked hypertension’, even a recent chapter on new and old office BP approaches in the Manual of Hypertension of the European Society of Hypertension, mentioned it only in passing [8]. In this context, the systematic review and meta-analysis by Seguret et al.[9] in this issue of the Journal of Hypertension on blood pressure measurements on a bare arm, over a sleeve or below a rolled-up sleeve is timely indeed. They had extracted data from 12 studies and their main finding was that for the thinner sleeves, there was an overestimation of SBP of only 0.56 mmHg (upper 95% confidence interval bound of 1.30 mmHg), and 1.10 mmHg (upper 95% confidence interval bound of 2.40 mmHg) for the thickest sleeves. The authors wisely note that the ’…maximum bias induced by the sleeve, is 1.30 mmHg with the thinnest ones, and 2.40 mmHg with the thickest’, is significantly lower than the 3 mmHg, estimated inaccuracy of modern oscillometric BP devices and lower than other behavioral biases frequently associated with BP measurement. The authors state that the overestimation was minimal in the obese individuals who were ‘overrepresented’ in their meta-analysis (as well as in many other clinics, certainly hypertension clinics) lending support for measuring BP over a thin sleeve if necessary, and against rolling it up, which may create greater chance of both tourniquet effect or misplacement in obese subjects. Indeed, the greatest overestimation of BP measured over a sleeve came from two studies from the same research group in Japan of elderly subjects who were significantly thinner (BMI around 23 m2/kg) than the other, rather younger participants in the other studies [10,11]. The four studies comparing BP over a bare arm in comparison to under a rolled-up sleeve find an overestimation of SBP of 2.76 mmHg (upper 95% confidence interval bound of 6.47 mmHg). Being an overestimation, it does not support the tourniquet effect, or for exchange it does testify for proper precautions taken rolling the sleeve up, as the tourniquet effect would have produced an underestimation rather than overestimation of BP. Thus, it seems that it is best to measure BP over a bare arm, nevertheless, when not feasible, it is better to measure it over the sleeve than rolling the sleeve up, particularly if the sleeve is not thick. Having said that, it is needed to be mentioned that the single study who used a really thick multilayered sleeves (8 and 17 mm!), mimicking measurement during outdoor winter conditions, such as during trekking or mountaineering prehospital emergency care, did not find a significant bias, although measurement time was prolonged by up to 3.5 s [12]. It is generally accepted that a meta-analysis of different studies can provide results as reliable as are the included studies. None of the studies in this meta-analysis was devoid of caveats as very well depicted by the authors. Indeed, substantial heterogeneity was described, perhaps reducing the reliability of the estimated biases. However, that heterogeneity also provides greater confidence of generalizability of these results. The authors point out that the included studies did not include home BP nor ambulatory blood pressure monitoring. Nevertheless, I believe that the conclusion that BP measurement can be performed over a thin sleeve is probably valid also at home. In the case of ambulatory blood pressure monitoring, the situation is less certain as the tendency of the cuff to slip down and/or change position with activity may be substantial, and personally I have no experience with over the sleeve monitoring in this context, nevertheless, at least one validated and widely used instrument has a picture of a patient with the cuff over his sleeve for presentation (https://www.spacelabshealthcare.com/wp-content/uploads/2018/02/ABBM_October_2019.pdf). ACKNOWLEDGEMENTS Conflicts of interest There are no conflicts of interest.