Journal reference: Myers MG, Godwin M, Dawes M, et al. Conventional versus automated measurement of blood pressure in primary care patients with systolic hypertension: randomised parallel design controlled trial. BMJ 2011; 342: d286 ^[1]

Link: http://dx.doi.org/10.1136/bmj.d286

Published: 7 February 2011

Evidence cookie says...

Automated blood pressure measurement is significantly closer to waking ambulatory BP than conventional manual measurement in a primary care clinic setting. Mean over-estimation of systolic blood pressure: automated: 2.3 mmHg conventional: 6.5 mmHg Procedure: patient left alone, average of 5 consecutive blood pressure readings spaced every 2 minutes Caution: there is a narrow range of participants in this study. See below for more details

Article details:

Study design:

multi-site cluster randomised controlled trial

Study aim:

compare the quality and accuracy of manual office blood pressure (BP) and automated office BP using the awake ambulatory BP as a gold standard

Study conclusion:

In compliant, otherwise healthy, primary care patients with systolic hypertension, introduction of automated office blood pressure measurement into routine primary care significantly reduced the white coat response compared with the ongoing use of manual office blood pressure measurement. The quality and accuracy of automated office blood pressure in relation to the awake ambulatory blood pressure was also significantly better when compared with manual office blood pressure.

Critical appraisal:

Methodology (PICO):

Participants: who was studied?

• 555 patients
  • 88 primary care physicians in 67 Canadian community practices
• systolic hypertension (HTN):
  • untreated patients: systolic BP ≥ 160 mmHg, diastolic BP < 95 mmHg
  • existing patients treated for HTN: systolic BP ≥ 140 mmHg, diastolic BP < 90 mmHg
• age > 45 years
• no serious co-existing illness limiting participation
• no history of non-compliance
• not treated for diabetes mellitus
• serum creatinine less than twice normal
• not using home blood pressure measurements
• recruited by letter sent from the office of their family physician
• baseline characteristics:
  • 38% male
  • mean age: 65 years
  • not receiving antihypertensive medication: 4.5%
  • (approximate) last routine manual office BP: mmHg (standard deviation)
    • systolic BP: 150 (10.8)
    • diastolic BP: 81 (8.5)

Intervention: what was the exposure?

• office blood pressure measurement with the BpTRU automated sphygomomanometer
  • initial reading to verify cuff is properly positioned
  • patient then left alone; the device measures five more readings and displays the mean result as well as individual readings
  • readings are taken at 2 minute intervals
• 24 hour ambulatory BP monitoring
  • patients advised to engage in routine daily activities
  • device recorded blood pressure at 15 minute intervals between 0600 and 2200 and 30 minute intervals overnight
  • mean awake ambulatory BP calculated according to the awake period as recorded in patient’s diary

Comparator: what was the control/alternative?

• manual office blood pressure recorded as per usual method by physician
• no additional instructions on proper BP measurement technique
• 24 hour ambulatory BP monitoring also performed as previously described

Outcomes: what was measured?

• primary outcome:
  • difference in blood pressure between the 24 hour ambulatory BP monitor and the baseline blood pressure measurement (either manual or automated)

Are the trial results valid?

Internal validity: Wikipedia

Were there sufficient participants in the trial?

Yes.

• The study was designed with 90% power to detect a 5 mmHg difference between groups
• an estimated sample size of 276 participants was required in each group (assuming 20% attrition → 221 participants)
• there were 299 participants analysed in the intervention group and 249 participants analysed in the usual care group

Was the assignment of patients to treatment randomised?

Yes.

• Randomisation was clustered at the level of the practice rather than the patient.

Were the groups similar at the start of the trial?

Yes.

• There were only minor differences between participant groups at baseline (see Table 1 ^[1])
• There was an imbalance with more patients randomised to the automated BP group → this was the result of more sites with multiple physicians randomised to automated office BP group

Aside from the allocated treatment, were groups treated equally?

Yes.

Were all patients who entered the trial accounted for?

Yes.

Very few drop outs:

• automatic BP measurement group: 4 drop outs (withdrew consent) of 303 participants
• manual BP measurement group:  3 drop outs (2 withdrew consent, 1 died) of 252 participants

Were measures objective or were the patients and clinicians kept blind to which treatment was received?

Yes.

Neither patients or clinicians were blinded (impossible) but BP measurements are objective.

What were the results?

Primary outcome

Difference between awake ambulatory BP and post-enrolment office BP:

• automated BP group: estimated mean differences in mmHg (95% confidence interval)
  • systolic BP: -2.3 (95% CI -0.31 to -4.3), P = 0.02
  • diastolic BP: -3.3 (-2.2 to -4.4), P < 0.001
• conventional manual BP group: estimated mean difference in mmHg (95% confidence interval)
  • systolic BP: -6.5 (-4.3 to -8.6), P < 0.001
  • diastolic BP: -4.3 (-2.9 to -5.8), P < 0.001
• interpretative note: the closer the estimated mean difference is to 0.0, the less the difference between the office BP recording and the ambulatory BP recording (considered the gold standard in this study)

Other outcomes:

N/A

Will the results help me care for my patient?

External validity: Wikipedia

Are the participants different to my patient?

• this was a primary care population with the measurements performed in the usual practice
  • this “real world” setting enhances the external validity of the results
• older population (mean age 65 years) with no one under age 45 years
• substantially more women than men
• absence of any severe illness
• almost all patients were receiving antihypertensive medication treatment
• last office BP measurements (with standard deviations) suggest the participants had residual mild to moderate systolic hypertension only
  • participant characteristics limits the external validity of the results
  • caution should be taken in generalising these results to other hypertensive populations

Is the treatment feasible?

• the treatment is contingent on owning an automatic sphygmomanometer capable of recording multiple readings with a set program
• a private room available for the patient for up to 8 minutes (5 recordings at 2 minute intervals) is required
  • this can be difficult to arrange in many general practices on a routine basis

Were all the clinically important outcomes considered?

• the outcome was point blood pressure measurement rather than sequelae of hypertensive disease, or long term BP control, or physician or patient satisfaction
• however, those other measures would add substantial complexity and cost to a trial that may not be ethical or responsible to run

Are the treatment benefits worth the potential harms/costs?

• where an appropriate automatic sphygmomanometer is already owned, yes
• it would seem to be an appropriate initial office test to perform before 24 hour ambulatory BP monitoring given the time and expense

Study weaknesses (summary)

• although cluster randomisation at the practice level was a sensible choice (risk of treatment contamination if single community primary care practices administered both the intervention and comparator treatments), it introduces a potential bias:
  • possible confounding factor related to individual practices, or geographic locations
• known participant characteristics (older, no significant comorbidities, mild to moderate systolic hypertension only) may limit the generalisability of the results
• participant demographics have limited detail
  • no reported information on ethnicity or socioeconomic status
  • no detailed information on antihypertensive use

Study strengths (summary)

• study performed in a primary care setting
• generally good methodology
• intervention can be feasibly used in the primary care setting
• it is probable that any biases in the the study would favour a smaller difference between the groups, i.e., it is very likely that the detected difference is real

Biases and conflicts of interests

• study was supported by a grant from the Heart and Stroke Foundation of Ontario

Clinical relevance to primary health care

Hypertension is the most frequently managed condition by Australian general practitioners. White coat hypertension is a significant issue. It commonly leads to over-treatment but also can result in under-treatment when it is suspected but not present. Ambulatory blood pressure (BP) monitoring over 24 hours is the gold standard. It more accurately reflects actual BP than clinic measurements, is a better predictor of clinical outcomes and end-organ damage, and there is evidence that it is cost-effective by avoiding unnecessary drug therapy ^[2]. However, ambulatory blood pressure monitoring can be difficult to arrange, not to mention costly and cumbersome for the patient.

This study demonstrated that the use of an automatic sphygmomanometer to measure 5 sequential BP readings, spaced every 2 minutes, was able to report a systolic BP with a mean difference of 2.3 mmHg, as compared to ambulatory BP monitoring. This is compared to conventional BP monitoring; the mean difference in systolic BP was 6.5 mmHg higher than ambulatory BP readings.

General practitioners who: (i) own an automatic sphygmomanometer capable of automatic sequential BP measurements, and (ii) have the practice facilities, should consider using this function in patients who they suspect have white coat hypertension, or have borderline hypertension.

Tip: the Omron semi-automated digital blood pressure machines that were given to Australian general practitioners for free in 2007 as part of the Better Blood Pressure Measurement Initiative by the High Blood Pressure Research Council of Australia, are capable of automatically measuring up to three sequential blood pressures and reporting the average.

Note: the participants in this study were all older, mostly well, and had medicated mild to moderate systolic hypertension only. Caution should be taken in generalising the findings to other hypertensive populations.

References

1. Myers MG, Godwin M, Dawes M, et al. Conventional versus automated measurement of blood pressure in primary care patients with systolic hypertension: randomised parallel design controlled trial. BMJ 2011; 342: d286
2. Position statement: Ambulatory blood pressure monitoring (draft), December 2010. National Heart Foundation of Australia and the High Blood Pressure Research Council of Australia. December 2010