NPC Archive Item: Current practice for monitoring and adjusting blood pressure treatment may be inappropriate

NOTE – This is an archive post from the NPC and has not been updated since first publication. Therefore, some hyperlinks may no longer be working.
MeReC Rapid Review NPC Logo

3rd June 2009

Two recent studies (Keenan et al 2009 and Law et al 2009) raise important concerns about current practice for routine monitoring of blood pressure. Firstly, using isolated blood pressure measurements for adjusting treatment can often be unreliable. Secondly, the relative reduction in cardiovascular risk achieved with antihypertensive drugs is related to the decrease in blood pressure and is largely independent of the class of antihypertensive drug used.

Level of evidence:
Law et al – Level 2 (limited-quality patient-oriented evidence) according to the SORT criteria.

The NICE guideline for the management of hypertension should continue to be followed. However, health professionals involved in the management of blood pressure need to be aware of the findings of these two recent studies, which may challenge current practice:

  • The practice of monitoring and adjusting blood pressure management on the basis of infrequent (e.g. every six months) measurements is an unreliable method of indicating underlying changes in blood pressure. Individual measurements that exceed treatment thresholds would seem inappropriate for adjusting antihypertensive treatment without further evaluation.
  • The relative decrease in cardiovascular (CV) risk achieved with antihypertensive drugs is related to the decrease in blood pressure they produce, and is independent of the pre-treatment blood pressure, the degree of CV risk, and with few exceptions, the type of antihypertensive drug used. As discussed in a recent blog, thiazide diuretics are an appropriate first-line choice for most people.

What is the background to these studies?
It is common practice to measure and titrate antihypertensive drug treatment in response to blood pressure measurements taken during clinic visits. However, individual measurements are prone to error, from inexact equipment, poor measurement technique, and normal day-to-day variations. Using data from a randomised controlled trial (RCT) of antihypertensive treatment (PROGRESS), Keenan et al investigated the reliability of individual blood pressure measurements to accurately predict changes in the true underlying changes in blood pressure that would necessitate alteration of antihypertensive medication.

For many years there has been considerable debate over which drug, or combination of drugs, should be used for controlling blood pressure. Law et al conducted a wide ranging meta-analysis (MA) that considered the efficacy of different classes of antihypertensive drugs for preventing CV disease in relation to their blood pressure lowering effects. It also considered the effects of blood pressure lowering in relation to CV disease status, age and pre-treatment blood pressure levels.

What do the studies claim?
Keenan et al demonstrated that the normal clinical approach of adjusting blood pressure lowering treatment on the basis of single blood pressure measurements made at clinic visits (e.g. every six months) is an unreliable method for identifying true changes in the underlying blood pressure and adjusting treatment. Because of variations in measurements resulting from normal day-to-day variations in blood pressure, and instrumental or operator error, readings in excess of values likely to prompt a change in drug treatment were more likely to be erroneous than a true reflection of the change in blood pressure since the last reading. For example, six months after blood pressure was stabilised on treatment at 130mmHg, for every seven recordings indicating a rise of 10mmHg in systolic blood pressure, only one of these readings would represent a true increase. A longer gap between start of treatment and follow up monitoring was shown to be more likely to reflect a true change in blood pressure, but even if the monitoring interval was left until nearly two years after the start of monitoring the false positive result was still around 50%. The authors call for evidence based guidelines to provide recommendations for monitoring treatment, in addition to providing advice about the choice of treatment.

Law et al carried out a meta-analysis of 147 RCTs of blood pressure lowering drugs (n=464,000). They found with few exceptions, that the five main classes of antihypertensive drugs (thiazide diuretics, beta-blockers, calcium channel blockers, ACE inhibitors, and angiotensin receptor blockers) were similarly effective in reducing the risk of coronary heart disease (CHD) events or stroke for a given reduction in blood pressure. Minor exceptions were calcium channel blockers, which had a greater preventative effect for strokes compared with other drugs (relative risk [RR] 0.91, 95% confidence interval [CI] 0.84 to 0.98; P=0.01), and beta-blockers which had a lesser effect compared with other drugs for preventing strokes (RR 1.18, 95%CI 1.03 to 1.36, P=0.02). Calcium channel blockers, though effective, had a lesser effect than other drugs in preventing heart failure. Beta-blockers had a special effect above that of blood pressure lowering in preventing CHD events in people with a history of CHD, although this extra effect was limited to a few years after myocardial infarction.

In people aged 60­–69 with a diastolic blood pressure of 90 mm Hg (or systolic blood pressure of 150 mm Hg), Law et al estimated that one drug at standard dose lowers the risk of CHD by about 25%, and of stroke by 35%. Law et al also extrapolated their results to predict the risk reduction in CHD obtained with two or three drugs at half standard dose. They estimated that three drugs at half standard dose would approximately double the effect of that achieved by one drug at standard dose.

Law et al also identified that the relative reduction in CV events was similar regardless of the pre-treatment blood pressure, in the presence or absence of existing CV disease. Of course the absolute benefit obtained from any intervention in the individual will be governed by their baseline risk. The authors question the need to measure blood pressure routinely in patients who are receiving antihypertensive drug treatment.

So what?
We emphasise in NPC materials on hypertension how important it is follow the NICE guideline recommendations when measuring blood pressure; immaculate technique and calibrated, maintained and validated measuring equipment are all essential for accurate measurement. However, even when recommended procedures are followed, isolated blood pressure measurements may give a false indication of changes in blood pressure with time because of normal individual intra- and inter-day variations. Even in the setting of a controlled clinical trial, Keenan et al demonstrate how unreliable isolated blood pressure measurements can be as a means of monitoring and adjusting antihypertensive treatment, and the results question the way the way that blood pressure is routinely monitored in people on antihypertensive treatment. Perhaps it is more appropriate to use such isolated measurements only as an indicator of a possible change and the need for more detailed assessment. Although an accompanying Editorial suggests intermittent self monitoring or ambulatory monitoring as alternatives, as pointed out in NICE guidance, the value of these approaches have yet to be adequately established in primary care, and these remain issues for further research. Strategies for monitoring blood pressure are given little attention in guidelines for managing blood pressure. Keenan et al point out that this is not surprising given the lack of quantitative data to support the use of any particular approach.

The MA by Law et al found little if any difference in the effectiveness of different classes of antihypertensive with regard to reducing the risk of CV events (with the possible exception of beta blockers) when blood pressure lowering effects were taken into account. It is the reduction in blood pressure achieved, not the class of drug used to achieve it that is important for reducing CV risk. Except, perhaps, for people with existing CHD, where beta-blockers are appropriate to include in treatment regimens, there is little to choose between the commonly used antihypertensive drugs and choice should be based on factors other than efficacy, including side-effect profiles and costs. As discussed in a recent blog, thiazide diuretics are an appropriate first-line choice for most people.

Another important finding of the Law et al study was that the relative reduction in risk of CV events achieved by blood pressure lowering was similar regardless of the pre-treatment blood pressure, the presence or absence of CV disease. The absolute risk reduction for the individual will, of course, depend on their baseline risk of CV disease. The results of this study support NICE’s advice to not treat hypertension in isolation, but to consider it in the context of a patient’s overall risk of CV disease.

There is very little evidence to support the predictions made by Law et al for the benefits of using combinations of drugs at half standard dose in comparison with a single drug at standard dose. Supportive evidence from a RCT is required before such an approach could be recommended.

More information on the management of hypertension can be found on the relevant section of NPC

Study details

Keenan K, et al. Long term monitoring in patients receiving treatment to lower blood pressure: analysis of data from placebo controlled randomised controlled trial. BMJ 2009;338:b1492

Design: Analysis of blood pressure measurements of patients in a randomised controlled trial (PROGRESS). The study evaluated the probability of measured changes in blood pressure being a true indication of the blood pressure change.

Blood pressure measurements: mean of two measurements recorded to nearest 2mmHg with a standard mercury sphygmomanometer at baseline and then at three months, six months, nine months, 15 months, and then every six months to 33 months.

Patients: 1709 patients with a history of stroke or transient ischaemic attack randomised to fixed doses of indapamide and perindopril.

Results: The likelihood that observed increases in blood pressure reflected true increases rose with time between measurements, such that the ratio of true positives to false positives reached parity at 21 months. Using a threshold for systolic blood pressure of 140mmHg, the estimated ratio of the number of false positives relative to true false positives for patients with a baseline measurement of 130mmHg reduced from 6 at 6 months, to 1.5 at 9 months and then to 0.8 at 33 months; the respective figures for a baseline measurement of 120mmHg were >200, 7.2 and 2.4. Using a threshold for diastolic blood pressure of 90mmHg, the estimated ratio of the number of false positives relative to true false positives for patients with a baseline measurement of 85mmHg reduced from 3.5 at 6 months, to 1.5 at 9 months and then to 0.8 at 33 months; the respective figures for a baseline measurement of 80mmHg were >39, 6.0 and 1.9.

Sponsorship: The study was sponsored by the Australian National Health and Medical Research Council Program.

Law MR, et al. Use of blood pressure lowering drugs in the prevention of cardiovascular disease: meta-analysis of 147 randomised trials in the context of expectations from prospective epidemiological studies. BMJ 2009;338:b1665

Design: Meta-analysis of 147 randomised trials, 108 considered blood pressure differences and 46 compared drug treatments. Results of two previous large MAs of 354 RCTs and 61 cohort studies were used to predict the effect of blood pressure lowering drugs in lowering blood pressure according to dose and for the reduction in disease events for a specified reduction in blood pressure, respectively.

Patients: 464,000 people defined into three categories: no history of CV disease; history of CHD, history of stroke.

Outcomes: Relative risk of CHD and strokes.

Results: For full results of this very wide-ranging analysis see the full paper and the appendices available online on the BMJ website. The following are some of the key findings:

Beta-blockers were more effective in reducing CHD events in people with a history of CHD (RR vs. placebo or active treatment 0.71, 95%CI 0.66 to 0.78) compared with a RR of 0.85 (95%CI 0.81 to 0.89) in trials of other drugs. There was no significant difference between beta-blockers and other drugs in the absence of a recent infarct (RR 0.99, 95%CI 0.82 to 1.20).

In drug comparison trials, for each of the five drug classes investigated excluding the special effect of beta-blockers in people with CHD, the relative risk estimates for reductions in CHD events or stroke, for each comparison, was close to 1.0 and the differences between classes of drug on average blood pressure reduction close to zero. The only exceptions were for prevention of stroke, where calcium channel blockers had a greater preventative effect (RR 0.91, 95%CI 0.84 to 0.98, P=0.01) and beta-blockers had a lesser preventative effect (RR 1.18, 95%CI 1.03 to 1.36, P=0.02) than other drugs.

Relative risk estimates for CHD events and stroke in blood pressure difference trials, standardised to a blood pressure reduction of 10mmHg systolic or 5mmHg diastolic, were 0.78 (95%CI 0.73 to 0.83) and 0.59 (95%CI 0.52 to 0.67), respectively.

The relative risk estimates of CHD events and stroke in the blood pressure difference trials were similar across all levels of pre-treatment blood pressure down to 110mmHg systolic and 70mmHg diastolic, and for patients with no history of CV disease, a history of CHD, or a history of stroke (RR 0.79, 0.76 and 0.79 for CHD, RR 0.54, 0.65, 0.66 for stroke, respectively).

For the trials reporting heart failure, there was considerable heterogeneity across the results of trials of beta-blockers, explained by a lack of a protective effect of beta-blockers without cardioselective or alpha blocking properties. However, when these were excluded, beta-blockers were shown to have a protective effect (RR 0.77, 95%CI 0.69 to 0.87, P=0.01). Although preventing heart failure by 19% in blood pressure difference trials (P=0.007), from drug comparison trials calcium channel blockers were significantly less effective than other drugs in preventing heart failure (RR 1.22, 95%CI 1.10 to 1.35, P<0.001).

Sponsorship: No funding source stated. The authors are from the Wolfsen institute of Preventative Medicine, Barts and The London School of Medicine, Queen Mary University of London. Two of the authors hold patents in the formulation of a combined pill to simultaneously reduce four cardiovascular risk factors, including blood pressure.

Please comment on this blog in the NPC discussion room, or using our feedback form.

Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you