 |
 |
31 October 2011
The cost-effectiveness modelling study underpinning the recommendations on blood pressure measurement in the updated NICE hypertension guidance, has recently been published. It concludes that, when used to diagnose hypertension after an initial raised reading in the clinic, ambulatory blood pressure monitoring (ABPM) reduces misdiagnosis and allows better targeted treatment, thereby reducing overall cost (saving between about £50-320 per patient over their lifetime).
Level of evidence:
Level 3 (other evidence) according to the SORT criteria.
Action
Healthcare professionals should follow NICE guidance on the management of hypertension. This advises that ABPM should be used to confirm a diagnosis of hypertension if the clinic-based blood pressure (BP) measurement is 140/90mmHg or higher on two or more readings. Home BP monitoring is an alternative for those who can’t tolerate ABPM.
Active plans need to be made locally to move practice towards the new NICE hypertension guidance. During this transitional period, and until arrangements to introduce ambulatory and/or home BP monitoring have been put in place, it seems reasonable to continue to follow the previous, 2006 guidance and ask patients (who do not have signs/symptoms of malignant hypertension) to return at least twice for further measurement in clinic, normally at monthly intervals.
What is the background to this?
The diagnosis of hypertension has traditionally been based on clinic BP measurements. However, ambulatory and home BP measurements better correlate with cardiovascular (CV) outcomes. Also, ABPM is more accurate than the other two methods for the diagnosis of hypertension. A systematic review recently found that clinic and home BP measurements have insufficient accuracy to be used as single diagnostic tests for hypertension and may result in substantial over diagnosis compared to ABPM (see MeReC Rapid Review No. 4501). NICE considered this systematic review when they updated the hypertension guideline and noted that, although ABPM may be more accurate at diagnosing hypertension than clinic or home BP monitoring, it is also more expensive in terms of monitor costs.
This cost-effectiveness analysis was therefore funded by NICE in order to inform the hypertension guideline. It compared three diagnostic strategies after a raised initial clinic BP reading: further clinic BP measurements at monthly intervals over three months, home BP measurements over a week, and 24-hour ABPM. Various factors that could potentially impact on the costs of using these strategies were taken into consideration. For example, the number and type of healthcare appointments required to confirm a diagnosis with each method, the failure rate, and the number of uses of the devices each year. As well as initial diagnosis costs, the analysis took into account downstream costs including hypertension treatment, checkups and development of CV disease.
What does this study claim?
In terms of costs, home BP monitoring was similar to clinic BP monitoring, whereas ABPM was cost saving. It dominated both of the other strategies and was the most cost-effective strategy for confirming a diagnosis of hypertension for both men and women, and for all age subgroups. Compared with further clinic monitoring, the mean lifetime cost saving per patient with APBM was between £56 (95% CI £10 to £105) in men aged 75 years and £323 (95% CI £222 to £389) in women aged 40 years. It also resulted in more quality-adjusted life years (QALYs), for men and women older than 50 years: from 0.006 (95% CI 0.000 to 0.015) for women aged 60 years to 0.022 (95% CI 0.012 to 0.035) for men aged 70 years. The findings were robust when assessed with a wide range of sensitivity analyses.
The authors state that the additional costs of ABPM, compared with further clinic or home BP monitoring, are counterbalanced by cost savings from the better targeting of treatment, and recommend that ABPM should be seriously considered for most people before antihypertensive treatment is started.
So what?
As stated in the NICE full hypertension guideline, contrary to what might have been expected, and mindful of the higher costs of ABPM devices, this cost-effectiveness analysis found using ABPM to confirm diagnosis of hypertension to be the most cost effective option across a range of age groups in both men and women. Remarkably, in most groups ABPM was found to actually improve health (increased QALYs) and reduce costs, suggesting that use of ABPM as recommended by NICE has the potential to be cost saving for the NHS. The NICE guideline development group (GDG) noted that this conclusion was robust to a wide range of sensitivity analyses including those varying the cost of ABPM, the failure rate for ABPM, the level of CV risk and the prevalence of true hypertension in the population. The conclusion was sensitive to assumptions regarding the accuracy of diagnosis with each method, e.g. when the other methods (clinic or home BP measurement) were assumed to be as accurate as ABPM – which the effectiveness analysis suggests they are not. The conclusion was also sensitive to the assumption that people who did not have hypertension but were treated did not receive benefits from treatment, which they might.
The analysis did not model the impact of unnecessarily treating people who did not have hypertension and the costs, inconvenience, adverse effects of treatment and impact disease labelling may have on individual patients incorrectly diagnosed as having hypertension. In addition, the conclusions based on the model were sensitive to variations in sensitivity and specificity values, which is important because the studies included in the systematic review underpinning the model did not conform precisely to the guideline recommendations modelled in the analysis. The paucity of studies and lack of direct comparison of all three diagnostic methods with sufficient follow-up is another limitation of this analysis.
What did NICE conclude?
NICE concluded that the use of ABPM for the routine diagnosis of hypertension, using a daytime average threshold of 135/85mmHg, in people who have previously been identified as potentially hypertensive at a threshold of 140/90mmHg using a clinic BP measurement, would be both cost-effective and in almost all cases, cost saving for the NHS, as well as improving the accuracy of diagnosis for patients. They therefore recommended that using ABPM in this way should be implemented for the routine diagnosis of hypertension in primary care.
The GDG recognised and discussed the considerable challenges for implementation of this recommendation. Sufficient numbers of validated ABPM devices need to be procured and adequately maintained. Staff need to be trained in their use and the interpretation of data generated by the ABPM reports. The existing recommendations on use of appropriate cuff size and recognition that automated measurements may be unreliable or impossible in people with significant pulse irregularity (e.g. atrial fibrillation) still apply. Some people will not tolerate ABPM and in others the procedure will fail. In those unable to tolerate or unwilling to undergo ABPM, the GDG recommends home BP measurement as an alternative means of confirming the diagnosis of hypertension with emphasis that ABPM is the preferred method. For those with significant pulse irregularity, ABPM and home BP measurement are likely to be unreliable methods for BP measurement and a series of clinic BP readings via manual auscultation remains the only suitable option.
Finally, the GDG discussed the practicalities of implementing this strategy for the diagnosis of hypertension. That implementation of this strategy is a challenge is acknowledged. Presently, some but not all primary care practices have access to ABPM devices. Some practices access ABPM through referral to secondary care. Few practices presently have sufficient numbers of devices to increase their use as required by this guideline recommendation. The GDG discussed the fact that models of future care cannot just be based on what we do now and considered it likely that alternative models of service provision would emerge, reflecting first and foremost what was best and most convenient for patients and local demand. The GDG considered it inevitable that the costs of ABPM devices will fall as demand for their use increases and that different models of ABPM provision will evolve over time to meet local demand.
Design: Markov model-based probabilistic cost-effectiveness analysis
Patients: A hypothetical primary-care population aged 40 years or older with a screening BP measurement greater than 140/90 mmHg and risk-factor prevalence equivalent to the general population.
Intervention and comparison: The study compared three diagnostic strategies: further clinic BP measurements at monthly intervals over three months, home BP measurements over a week, and 24-hour ABPM. The analysis was based on sensitivity and specificity data from the systematic review discussed above. It accounted for both costs and health outcomes from a UK perspective.
Outcomes: Findings were expressed in terms of lifetime costs, QALYs, and cost-effectiveness.
Results: ABPM was the most cost-effective strategy for confirming a diagnosis of hypertension in men and women of all ages. Compared with further clinic monitoring, the mean lifetime cost saving per patient with APBM was between £56 (95% CI £10 to £105) in men aged 75 years and £323 (95% CI £222 to £389) in women aged 40 years. It also resulted in more QALYs, for men and women older than 50 years: from 0.006 (95% CI 0.000 to 0.015) for women aged 60 years to 0.022 (95% CI 0.012 to 0.035) for men aged 70 years. This finding was robust when assessed with a wide range of deterministic sensitivity analyses around the base case, but was sensitive if home monitoring was judged to have equal test performance to ambulatory monitoring or if treatment was judged effective irrespective of whether an individual was hypertensive.
The savings with ABPM, compared with home and clinic BP monitoring, were primarily because of the costs of hypertensive treatment that were avoided because of the higher specificity of ABPM. For example, the discounted treatment costs for men aged 60 years were £776 for clinic BP measurement compared with £744 for home BP measurement and £631 for ABPM (incremental costs compared with clinic BP measurement of −£32 [95% CI −£99 to £19] and −£144 [95% CI −£246 to −£22] respectively). The incremental diagnosis costs were £6 (95% CI £0 to £13) for home monitoring and £42 (95% CI £22 to £57) for ABPM with similar costs in terms of subsequent events and follow-up. Similar treatment savings offset the additional diagnostic costs associated with ABPM in all age and sex subgroups.
Sponsorship: National Institute for Health Research and the National Institute for Health and Clinical Excellence
Further information can be found on NHS Evidence and in the Cardiovascular disease – hypertension e-learning section of the NPC website.
Please comment on this rapid review 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