11 October 2011
A systematic review has found that clinic and home blood pressure (BP) measurement have insufficient accuracy to be used as a single diagnostic test and may result in substantial over diagnosis compared to ambulatory BP monitoring.
Level of evidence:
Level 3 (other evidence) according to the SORT criteria.
The updated NICE guidance on the management of hypertension recommends that ‘If the clinic blood pressure is 140/90 mmHg or higher, offer ambulatory blood pressure monitoring to confirm the diagnosis’. Health professionals need to be aware of this change in guidance, which is supported by the findings of this systematic review.
What is the background to this?
Measuring BP accurately is crucial to effective diagnosis and treatment of hypertension. NICE guidance indicates thresholds for initiating BP lowering treatment, and targets for appropriate BP control. It is common practice to measure and titrate BP lowering treatment in response to BP measurements taken during clinic visits. However, such measurements are prone to error from inaccurate equipment, observation bias, ‘white coat’ effect, and normal day-to-day variation. Furthermore, and as discussed in the NPC’s hypertension e-learning materials, there is evidence that BP measurement in clinic is not always done with immaculate technique (see part 2 of the <60minute eLearning event for more information).
What does this study claim?
The authors found that when compared to ambulatory monitoring, single clinic and home measurements lacked sensitivity or specificity and their use may result in substantial over diagnosis of hypertension. Compared with ambulatory monitoring thresholds of 135/85 mmHg, clinic measurements (over 140/90 mmHg) had mean sensitivity and specificity of 74.6% (95% confidence interval [CI] 60.7% to 84.8%) and 74.6% (95%CI 47.9% to 90.4%), respectively, whereas home measurements (over 135/85 mmHg) had mean sensitivity and specificity of 85.7% (95%CI 78.0% to 91.0%) and 62.4% (95% CI 48.0% to 75.0%).
How does this relate to other studies?
The difficulties of attaining sufficiently accurate BP measurements for the diagnosis and monitoring of hypertension has been a long-running problem. Issues regarding effective blood pressure measurement have been discussed in previous MeReC Rapid reviews, see 353, 644, 1149, and 1342 for more information.
A recent secondary analysis of the HINTS RCT, conducted with 444 US veterans with known poorly controlled hypertension, found that to improve correct classification of BP control the effect of within-patient variability could be greatly reduced if physicians average several measurements. Most benefit accrued at five or six measurements. Importantly, the authors state that single clinic measurements ‘potentially misclassify a large proportion of patients’.
This study shows clinic or home BP measurement results may differ substantially from those obtained by ambulatory monitoring. This appears to support more rigorous methods of measurement with ambulatory monitoring taken as their ‘gold standard’. They concluded that clinic and home measurement lack sufficient sensitivity or specificity to be recommended as a single diagnostic test.
Sensitivity is a measure of how good a test is at correctly identifying people who have the disease (e.g. hypertension), while specificity is a measure of how good a test is at correctly identifying people who do not have the disease. For example, compared with ambulatory BP monitoring as the gold standard, the mean sensitivity of clinic BP measurements was 74.6%. Therefore, using clinic BP measurements approximately 75% of people who actually had hypertension would be accurately diagnosed with hypertension, and 25% of people would be diagnosed with hypertension inaccurately (overdiagnosis). Similarly, the mean specificity of clinic BP measurements was 74.6%, so approximately 75% of people without hypertension would be correctly identified as not having the disease and 25% of people would not be diagnosed with hypertension, who actually had the disease (underdiagnosis).
If a diagnosis of hypertension is made without confirmation by ambulatory monitoring many people with a diagnosis may not in fact have hypertension. This has important implications including the mislabelling of otherwise healthy people; exposure to unnecessary risks of medication side effects; and avoidable prescribing costs. The degree of ‘overdiagnoses’ depends on the prevalence of hypertension in a population. The authors estimated if the prevalence of hypertension is 10% (for example, in the general population aged less than 40), then nearly three out of four diagnoses on clinic measurement would be incorrect when compared to ambulatory monitoring. If the prevalence is 50% (for example, in the general population over 65), then three out of four diagnoses will be correct (see Table). Where clinic measured BP is near the diagnostic threshold the sensitivity analysis suggests the rate of correct diagnoses drops to approximately 61% when compared to ambulatory monitoring.
Table. Probability (%) that test diagnosis with reference to ambulatory monitoring is correct given likelihood ratios and estimates of prevalence
|Sensitivity analysis 50%||67||61||80||76|
What are the limitations of this study?
There was a lack of comparable data with only ten of the twenty eligible studies useful for comparison against ambulatory monitoring and only one study directly compared all methods of measurement. The included studies varied in terms of mean baseline BP, their thresholds for the diagnosis of hypertension, and there was a wide variation in the number of measurements taken. The authors also noted methodological weaknesses in all the included studies with only 11 using validated devices and only six provided evidence of blinding. As a result only a small proportion of the studies were suitable for meta-analysis, and their planned sensitivity-analysis could not be performed.
In summary these findings challenge current practice. Updated NICE guidance on the management of hypertension recommends ‘If the clinic blood pressure is 140/90 mmHg or higher, offer ambulatory blood pressure monitoring to confirm the diagnosis of hypertension’. Home blood pressure monitoring may be used as an alternate means of confirmation where ambulatory monitoring is unavailable or not tolerated.
Hodgkinson J, et al. Relative effectiveness of clinic and home blood pressure monitoring compared with ambulatory blood pressure monitoring in diagnosis of hypertension: systematic review. BMJ 2011; 342:d3621
Design: A systematic review with meta-analysis and hierarchical summary receiver operating characteristic models. Methodological quality was appraised, including evidence of validation of BP measurement equipment. Studies with extractable data that looked at the diagnosis of hypertension in adults of all ages with various thresholds using home and/or clinic measurement, compared with those using ambulatory monitoring that clearly defined diagnostic thresholds.
Patients: A total of 5863 individuals with a mean age of 48.8 years, 57% women, from twenty studies were used in the analysis. People who were pregnant, in hospital, or receiving treatment at the time of the comparison (unless these groups could be excluded from other data within a paper) were excluded.
Intervention and comparison
Home and/or clinic BP measurement vs. ambulatory BP monitoring for the diagnosis of hypertension. Of the twenty eligible studies, only seven (clinic) and three (home) could be directly compared with ambulatory monitoring. Only one compared all three methods of measurement.
Outcomes and results
See ‘What does this study claim?’
Sponsorship: No funding source stated.
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